├── CONTRIBUTING.md ├── LICENSE ├── README.md ├── docs └── .gitkeep ├── subtitle en+zh ├── Game Theory I-Week 1 │ ├── GTO-1-01 - The TCP Backoff Game, and an Introduction to Game Theory.srt │ ├── GTO-1-02 - Self Interest and Utility Theory.srt │ ├── GTO-1-03 - Defining Games.srt │ ├── GTO-1-04 - Examples of Games from Game Theory.srt │ ├── GTO-1-05 - Nash Equilibrium Introduction, and the Keynes Beauty Contest.srt │ ├── GTO-1-06 - Strategic Reasoning and the Keynes Beauty Contest Game.srt │ ├── GTO-1-07 - Best Response and Nash Equilibrium.srt │ ├── GTO-1-08 - Examples.srt │ ├── GTO-1-09 - Dominant Strategies.srt │ └── GTO-1-10 - Pareto Optimality.srt ├── Game Theory I-Week 2 │ ├── GTO-2-01 - Mixed Strategies and Nash Equilibrium - A High-Level Taste.srt │ ├── GTO-2-02 - MixedNash.srt │ ├── GTO-2-03 - Computing Mixed-Strategy Nash Equilibria.srt │ ├── GTO-2-04a - Hardness of Computing a Nash Equilibrium (Basic).srt │ ├── GTO-2-04b - Hardness of Computing a Nash Equilibrium (Advanced).srt │ ├── GTO-2-05 - Mixed-Strategy Nash Equilibrium in Practice.srt │ └── GTO-2-06 - Data on Mixed Strategies from Tennis and Soccer.srt ├── Game Theory I-Week 3 │ ├── GTO-3-01 - Other solution concepts - A High-Level Taste.srt │ ├── GTO-3-02 - Strictly Dominated Strategies and Iterative Removal.srt │ ├── GTO-3-03 - Dominated Strategies and Iterative Removal - An Application.srt │ ├── GTO-3-04a - Maxmin Strategies (Basic).srt │ ├── GTO-3-04b - Maxmin Strategies (Advanced).srt │ └── GTO-3-05 - Correlated Equilibrium - Intuition.srt ├── Game Theory I-Week 4 │ ├── GTO-4-01 - Perfect Information Extensive Form - Taste.srt │ ├── GTO-4-02 - Formalizing Perfect Information Extensive Form Games.srt │ ├── GTO-4-03 - Perfect Information Extensive Form - Strategies, Best Response, Nash Equilibrium.srt │ ├── GTO-4-04 - Subgame Perfection.srt │ ├── GTO-4-05 - Backward Induction.srt │ ├── GTO-4-06 - Subgame Perfect Application - Ultimatum Bargaining.srt │ ├── GTO-4-07 - Imperfect Information Extensive Form - Poker.srt │ ├── GTO-4-08 - Imperfect Information Extensive Form - Definition, Strategies.srt │ ├── GTO-4-09 - Mixed and Behavioral Strategies.srt │ └── GTO-4-10 - Incomplete Information in the Extensive Form - Beyond Subgame Perfection.srt ├── Game Theory I-Week 5 │ ├── GTO-5-01 - Repeated Games.srt │ ├── GTO-5-02 - Infinitely Repeated Games & Utility.srt │ ├── GTO-5-03 - Stochastic Games.srt │ ├── GTO-5-04 - Learning in Repeated Games.srt │ ├── GTO-5-05 - Equilibria of Infinitely Repeated Games.srt │ ├── GTO-5-06 - Discounted Repeated Games.srt │ └── GTO-5-07 - A Folk Theorem for Discounted Repeated Games.srt ├── Game Theory I-Week 6 │ ├── GTO-6-01 - Bayesian Games - Taste.srt │ ├── GTO-6-02 - Bayesian Games - First Definition.srt │ ├── GTO-6-03 - Bayesian Games - Second Definition.srt │ ├── GTO-6-04 - Analyzing Bayesian Games.srt │ └── GTO-6-05 - Bayesian Games - Another Example.srt ├── Game Theory I-Week 7 │ ├── GTO-7-01 - Coalitional Game Theory - Taste.srt │ ├── GTO-7-02 - Coalitional Game Theory - Definitions.srt │ ├── GTO-7-03 - The Shapley Value.srt │ ├── GTO-7-04 - The Core.srt │ └── GTO-7-05 - Comparing the Core and the Shapley Value in an Example.srt ├── Game Theory II-Week 1 │ ├── GTO2-1-01 - Social Choice - Taste.srt │ ├── GTO2-1-02 - Social Choice - Voting Schemes.srt │ ├── GTO2-1-03 - Social Choice - Paradoxical Outcomes.srt │ ├── GTO2-1-04 - Social Choice - Impossibility of Non-Paradoxical Social Welfare Functions.srt │ ├── GTO2-1-05 - Social Choice - Arrow's Theorem.srt │ ├── GTO2-1-06 - Impossible of Non-paradoxical Social Choice Functions.srt │ └── GTO2-1-07 - Single-Peaked Preferences.srt ├── Game Theory II-Week 2 │ ├── GTO2-2-01 - Mechanism Design - Taste.srt │ ├── GTO2-2-02 - Mechanism Design - Implementation.srt │ ├── GTO2-2-03 - Revelation Principle.srt │ ├── GTO2-2-04 - Impossibility of General, Dominant-Strategy Implementation.srt │ ├── GTO2-2-05 - Transferable Utility.srt │ └── GTO2-2-06 - Mechanism Design as an Optimization Problem.srt ├── Game Theory II-Week 3 │ ├── GTO2-3-01 - VCG - Taste.srt │ ├── GTO2-3-02 - Vickrey-Clarke-Groves Mechanisms - Definitions.srt │ ├── GTO2-3-03 - VCG Example.srt │ ├── GTO2-3-04 - Limitations of VCG.srt │ ├── GTO2-3-05 - Individual Rationality and Budget Balance in VCG.srt │ └── GTO2-3-06 - Myerson-Satterthwaite Theorem.srt └── Game Theory II-Week 4 │ ├── GTO2-4-01 - Auctions - Taste.srt │ ├── GTO2-4-02 - Auctions - Taxonomy.srt │ ├── GTO2-4-03 - Bidding in Second-Price Auctions.srt │ ├── GTO2-4-04 - Bidding in First-Price Auctions.srt │ ├── GTO2-4-05 - Revenue Equivalence.srt │ ├── GTO2-4-06 - Optimal Auctions.srt │ └── GTO2-4-07 - More Advanced Auctions.srt └── update.sh /CONTRIBUTING.md: -------------------------------------------------------------------------------- 1 | # 贡献指南 2 | 3 | > 请您勇敢地去翻译和改进翻译。虽然我们追求卓越，但我们并不要求您做到十全十美，因此请不要担心因为翻译上犯错——在大部分情况下，我们的服务器已经记录所有的翻译，因此您不必担心会因为您的失误遭到无法挽回的破坏。（改编自维基百科） 4 | 5 | 课程视频： 6 | 7 | + [斯坦福博弈论课程官网](http://www.game-theory-class.org/) 8 | + [Cousera 博弈论 1](https://www.coursera.org/learn/game-theory-1) 9 | + [Cousera 博弈论 2](https://www.coursera.org/learn/game-theory-2) 10 | 11 | 负责人： 12 | 13 | + [viviwong](https://github.com/viviwong) 14 | 15 | ## 章节列表 16 | 17 | + 博弈论 I 18 | + 1-1 Game Theory Intro - TCP Backoff 19 | + 1-2 Self-Interested Agents and Utility Theory 20 | + 1-3 Defining Games 21 | + 1-4 Examples of Games 22 | + 1-5 Nash Equilibrium Intro 23 | + 1-6 Strategic Reasoning 24 | + 1-7 Best Response and Nash Equilibrium 25 | + 1-8 Nash Equilibrium of Example Games 26 | + 1-9 Dominant Strategies 27 | + 1-10 Pareto Optimality 28 | + 2-1 Mixed Strategies and Nash Equilibrium (I) 29 | + 2-2 Mixed Strategies and Nash Equilibrium (II) 30 | + 2-3 Computing Mixed Nash Equilibrium 31 | + 2-4 Hardness Beyond 2x2 Games - Basic 32 | + 2-4 Hardness Beyond 2x2 Games - Advanced 33 | + 2-5 Example: Mixed Strategy Nash 34 | + 2-6 Data: Professional Sports and Mixed Strategies 35 | + 3-1 Beyond the Nash Equilibrium 36 | + 3-2 Strictly Dominated Strategies & Iterative Removal 37 | + 3-3 Dominated Strategies & Iterative Removal: An Application 38 | + 3-4 Maxmin Strategies 39 | + 3-4 Maxmin Strategies - Advanced 40 | + 3-5 Correlated Equilibrium: Intuition 41 | + 4-1 Perfect Information Extensive Form: Taste 42 | + 4-2 Formalizing Perfect Information Extensive Form Games 43 | + 4-3 Perfect Information Extensive Form: Strategies, BR, NE 44 | + 4-4 Subgame Perfection 45 | + 4-5 Backward Induction 46 | + 4-6 Subgame Perfect Application: Ultimatum Bargaining 47 | + 4-7 Imperfect Information Extensive Form: Poker 48 | + 4-8 Imperfect Information Extensive Form: Definition, Strategies 49 | + 4-9 Mixed and Behavioral Strategies 50 | + 4-10 Incomplete Information in the Extensive Form: Beyond Subgame Perfection 51 | + 博弈论 II 52 | + 1.1 Social Choice: Taste 53 | + 1.2 Social Choice: Voting Scheme 54 | + 1.3 Paradoxical Outcomes 55 | + 1.4 Impossibility of Non-Paradoxical Social Welfare Functions 56 | + 1.5 Arrow's Theorem 57 | + 1.6 Impossibility of Non-Pardoxical Social Choice Functions 58 | + 1.7 Single-Peaked Preferences 59 | + 2.1 Mechanism Design: Taste 60 | + 2.2 Implementation 61 | + 2.3 Mechanism Design: Examples 62 | + 2.4 Revelation Principle 63 | + 2.5 Revelation Principle: Examples 64 | + 2.6 Impossibility of General Dominant-Strategy Implementation 65 | + 2.7 Transferable Utility 66 | + 2.8 Transferable Utility Example 67 | + 2.9 Mechanism Design as an Optimization Problem 68 | + 3.1 VCG: Taste 69 | + 3.2 VCG: Definitions 70 | + 3.3 VCG: Examples 71 | + 3.4 VCG: Limitations 72 | + 3.5 VCG: Individual Rationality and Budget Balance in VCG 73 | + 3.6 VCG: The Myerson-Satterthwaite Theorem 74 | + 4.1 Auctions: Taste 75 | + 4.2 Auctions: Taxonomy 76 | + 4.3 Bidding in Second-Price Auctions 77 | + 4.4 Bidding in First-Price Auctions 78 | + 4.5 Revenue Equivalence 79 | + 4.6 Optimal Auctions 80 | + 4.7 More Advanced Auctions 81 | 82 | ## 流程 83 | 84 | ### 一、认领 85 | 86 | 首先查看[整体进度](https://github.com/apachecn/stanford-game-theory-notes-zh/issues/1)，确认没有人认领了你想认领的章节。 87 | 88 | 然后回复 ISSUE，注明“章节 + QQ 号”。 89 | 90 | ### 二、整理笔记 91 | 92 | + 翻译 Coursera 课程页面的字幕，如果已有中文翻译，跳过此步 93 | 94 | （可以利用[谷歌翻译](https://translate.google.cn)，但一定要把它变得可读） 95 | 96 | + 排版成段落，并添加视频截图 97 | 98 | ### 三、提交 99 | 100 | + `fork` Github 项目 101 | + 将文档（**Markdown 格式**）放在`docs`中。 102 | + `push` 103 | + `pull request` 104 | 105 | 请见 [Github 入门指南](https://github.com/apachecn/kaggle/blob/master/docs/GitHub)。 106 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License (CC BY-NC-SA 4.0) 2 | 3 | Copyright © 2020 ApacheCN(apachecn@163.com) 4 | 5 | By exercising the Licensed Rights (defined below), You accept and agree to be bound by the terms and conditions of this Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License ("Public License"). To the extent this Public License may be interpreted as a contract, You are granted the Licensed Rights in consideration of Your acceptance of these terms and conditions, and the Licensor grants You such rights in consideration of benefits the Licensor receives from making the Licensed Material available under these terms and conditions. 6 | 7 | Section 1 – Definitions. 8 | 9 | a. Adapted Material means material subject to Copyright and Similar Rights that is derived from or based upon the Licensed Material and in which the Licensed Material is translated, altered, arranged, transformed, or otherwise modified in a manner requiring permission under the Copyright and Similar Rights held by the Licensor. For purposes of this Public License, where the Licensed Material is a musical work, performance, or sound recording, Adapted Material is always produced where the Licensed Material is synched in timed relation with a moving image. 10 | b. Adapter's License means the license You apply to Your Copyright and Similar Rights in Your contributions to Adapted Material in accordance with the terms and conditions of this Public License. 11 | c. BY-NC-SA Compatible License means a license listed at creativecommons.org/compatiblelicenses, approved by Creative Commons as essentially the equivalent of this Public License. 12 | d. Copyright and Similar Rights means copyright and/or similar rights closely related to copyright including, without limitation, performance, broadcast, sound recording, and Sui Generis Database Rights, without regard to how the rights are labeled or categorized. For purposes of this Public License, the rights specified in Section 2(b)(1)-(2) are not Copyright and Similar Rights. 13 | e. Effective Technological Measures means those measures that, in the absence of proper authority, may not be circumvented under laws fulfilling obligations under Article 11 of the WIPO Copyright Treaty adopted on December 20, 1996, and/or similar international agreements. 14 | f. Exceptions and Limitations means fair use, fair dealing, and/or any other exception or limitation to Copyright and Similar Rights that applies to Your use of the Licensed Material. 15 | g. License Elements means the license attributes listed in the name of a Creative Commons Public License. The License Elements of this Public License are Attribution, NonCommercial, and ShareAlike. 16 | h. Licensed Material means the artistic or literary work, database, or other material to which the Licensor applied this Public License. 17 | i. Licensed Rights means the rights granted to You subject to the terms and conditions of this Public License, which are limited to all Copyright and Similar Rights that apply to Your use of the Licensed Material and that the Licensor has authority to license. 18 | j. Licensor means the individual(s) or entity(ies) granting rights under this Public License. 19 | k. NonCommercial means not primarily intended for or directed towards commercial advantage or monetary compensation. For purposes of this Public License, the exchange of the Licensed Material for other material subject to Copyright and Similar Rights by digital file-sharing or similar means is NonCommercial provided there is no payment of monetary compensation in connection with the exchange. 20 | l. Share means to provide material to the public by any means or process that requires permission under the Licensed Rights, such as reproduction, public display, public performance, distribution, dissemination, communication, or importation, and to make material available to the public including in ways that members of the public may access the material from a place and at a time individually chosen by them. 21 | m. Sui Generis Database Rights means rights other than copyright resulting from Directive 96/9/EC of the European Parliament and of the Council of 11 March 1996 on the legal protection of databases, as amended and/or succeeded, as well as other essentially equivalent rights anywhere in the world. 22 | n. You means the individual or entity exercising the Licensed Rights under this Public License. Your has a corresponding meaning. 23 | 24 | Section 2 – Scope. 25 | 26 | a. License grant. 27 | 1. Subject to the terms and conditions of this Public License, the Licensor hereby grants You a worldwide, royalty-free, non-sublicensable, non-exclusive, irrevocable license to exercise the Licensed Rights in the Licensed Material to: 28 | A. reproduce and Share the Licensed Material, in whole or in part, for NonCommercial purposes only; and 29 | B. produce, reproduce, and Share Adapted Material for NonCommercial purposes only. 30 | 2. Exceptions and Limitations. For the avoidance of doubt, where Exceptions and Limitations apply to Your use, this Public License does not apply, and You do not need to comply with its terms and conditions. 31 | 3. Term. The term of this Public License is specified in Section 6(a). 32 | 4. Media and formats; technical modifications allowed. The Licensor authorizes You to exercise the Licensed Rights in all media and formats whether now known or hereafter created, and to make technical modifications necessary to do so. The Licensor waives and/or agrees not to assert any right or authority to forbid You from making technical modifications necessary to exercise the Licensed Rights, including technical modifications necessary to circumvent Effective Technological Measures. For purposes of this Public License, simply making modifications authorized by this Section 2(a)(4) never produces Adapted Material. 33 | 5. Downstream recipients. 34 | A. Offer from the Licensor – Licensed Material. Every recipient of the Licensed Material automatically receives an offer from the Licensor to exercise the Licensed Rights under the terms and conditions of this Public License. 35 | B. Additional offer from the Licensor – Adapted Material. Every recipient of Adapted Material from You automatically receives an offer from the Licensor to exercise the Licensed Rights in the Adapted Material under the conditions of the Adapter’s License You apply. 36 | C. No downstream restrictions. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material. 37 | 6. No endorsement. Nothing in this Public License constitutes or may be construed as permission to assert or imply that You are, or that Your use of the Licensed Material is, connected with, or sponsored, endorsed, or granted official status by, the Licensor or others designated to receive attribution as provided in Section 3(a)(1)(A)(i). 38 | b. Other rights. 39 | 1. Moral rights, such as the right of integrity, are not licensed under this Public License, nor are publicity, privacy, and/or other similar personality rights; however, to the extent possible, the Licensor waives and/or agrees not to assert any such rights held by the Licensor to the limited extent necessary to allow You to exercise the Licensed Rights, but not otherwise. 40 | 2. Patent and trademark rights are not licensed under this Public License. 41 | 3. To the extent possible, the Licensor waives any right to collect royalties from You for the exercise of the Licensed Rights, whether directly or through a collecting society under any voluntary or waivable statutory or compulsory licensing scheme. In all other cases the Licensor expressly reserves any right to collect such royalties, including when the Licensed Material is used other than for NonCommercial purposes. 42 | 43 | Section 3 – License Conditions. 44 | 45 | Your exercise of the Licensed Rights is expressly made subject to the following conditions. 46 | 47 | a. Attribution. 48 | 1. If You Share the Licensed Material (including in modified form), You must: 49 | A. retain the following if it is supplied by the Licensor with the Licensed Material: 50 | i. identification of the creator(s) of the Licensed Material and any others designated to receive attribution, in any reasonable manner requested by the Licensor (including by pseudonym if designated); 51 | ii. a copyright notice; 52 | iii. a notice that refers to this Public License; 53 | iv. a notice that refers to the disclaimer of warranties; 54 | v. a URI or hyperlink to the Licensed Material to the extent reasonably practicable; 55 | B. indicate if You modified the Licensed Material and retain an indication of any previous modifications; and 56 | C. indicate the Licensed Material is licensed under this Public License, and include the text of, or the URI or hyperlink to, this Public License. 57 | 2. You may satisfy the conditions in Section 3(a)(1) in any reasonable manner based on the medium, means, and context in which You Share the Licensed Material. For example, it may be reasonable to satisfy the conditions by providing a URI or hyperlink to a resource that includes the required information. 58 | 3. If requested by the Licensor, You must remove any of the information required by Section 3(a)(1)(A) to the extent reasonably practicable. 59 | b. ShareAlike. 60 | In addition to the conditions in Section 3(a), if You Share Adapted Material You produce, the following conditions also apply. 61 | 1. The Adapter’s License You apply must be a Creative Commons license with the same License Elements, this version or later, or a BY-NC-SA Compatible License. 62 | 2. You must include the text of, or the URI or hyperlink to, the Adapter's License You apply. You may satisfy this condition in any reasonable manner based on the medium, means, and context in which You Share Adapted Material. 63 | 3. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, Adapted Material that restrict exercise of the rights granted under the Adapter's License You apply. 64 | 65 | Section 4 – Sui Generis Database Rights. 66 | 67 | Where the Licensed Rights include Sui Generis Database Rights that apply to Your use of the Licensed Material: 68 | 69 | a. for the avoidance of doubt, Section 2(a)(1) grants You the right to extract, reuse, reproduce, and Share all or a substantial portion of the contents of the database for NonCommercial purposes only; 70 | b. if You include all or a substantial portion of the database contents in a database in which You have Sui Generis Database Rights, then the database in which You have Sui Generis Database Rights (but not its individual contents) is Adapted Material, including for purposes of Section 3(b); and 71 | c. You must comply with the conditions in Section 3(a) if You Share all or a substantial portion of the contents of the database. 72 | 73 | For the avoidance of doubt, this Section 4 supplements and does not replace Your obligations under this Public License where the Licensed Rights include other Copyright and Similar Rights. 74 | 75 | Section 5 – Disclaimer of Warranties and Limitation of Liability. 76 | 77 | a. Unless otherwise separately undertaken by the Licensor, to the extent possible, the Licensor offers the Licensed Material as-is and as-available, and makes no representations or warranties of any kind concerning the Licensed Material, whether express, implied, statutory, or other. This includes, without limitation, warranties of title, merchantability, fitness for a particular purpose, non-infringement, absence of latent or other defects, accuracy, or the presence or absence of errors, whether or not known or discoverable. Where disclaimers of warranties are not allowed in full or in part, this disclaimer may not apply to You. 78 | b. To the extent possible, in no event will the Licensor be liable to You on any legal theory (including, without limitation, negligence) or otherwise for any direct, special, indirect, incidental, consequential, punitive, exemplary, or other losses, costs, expenses, or damages arising out of this Public License or use of the Licensed Material, even if the Licensor has been advised of the possibility of such losses, costs, expenses, or damages. Where a limitation of liability is not allowed in full or in part, this limitation may not apply to You. 79 | c. The disclaimer of warranties and limitation of liability provided above shall be interpreted in a manner that, to the extent possible, most closely approximates an absolute disclaimer and waiver of all liability. 80 | 81 | Section 6 – Term and Termination. 82 | 83 | a. This Public License applies for the term of the Copyright and Similar Rights licensed here. However, if You fail to comply with this Public License, then Your rights under this Public License terminate automatically. 84 | b. Where Your right to use the Licensed Material has terminated under Section 6(a), it reinstates: 85 | 1. automatically as of the date the violation is cured, provided it is cured within 30 days of Your discovery of the violation; or 86 | 2. upon express reinstatement by the Licensor. 87 | For the avoidance of doubt, this Section 6(b) does not affect any right the Licensor may have to seek remedies for Your violations of this Public License. 88 | c. For the avoidance of doubt, the Licensor may also offer the Licensed Material under separate terms or conditions or stop distributing the Licensed Material at any time; however, doing so will not terminate this Public License. 89 | d. Sections 1, 5, 6, 7, and 8 survive termination of this Public License. 90 | 91 | Section 7 – Other Terms and Conditions. 92 | 93 | a. The Licensor shall not be bound by any additional or different terms or conditions communicated by You unless expressly agreed. 94 | b. Any arrangements, understandings, or agreements regarding the Licensed Material not stated herein are separate from and independent of the terms and conditions of this Public License. 95 | 96 | Section 8 – Interpretation. 97 | 98 | a. For the avoidance of doubt, this Public License does not, and shall not be interpreted to, reduce, limit, restrict, or impose conditions on any use of the Licensed Material that could lawfully be made without permission under this Public License. 99 | b. To the extent possible, if any provision of this Public License is deemed unenforceable, it shall be automatically reformed to the minimum extent necessary to make it enforceable. If the provision cannot be reformed, it shall be severed from this Public License without affecting the enforceability of the remaining terms and conditions. 100 | c. No term or condition of this Public License will be waived and no failure to comply consented to unless expressly agreed to by the Licensor. 101 | d. Nothing in this Public License constitutes or may be interpreted as a limitation upon, or waiver of, any privileges and immunities that apply to the Licensor or You, including from the legal processes of any jurisdiction or authority. -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # 斯坦福 **Game Theory** 博弈论中文笔记 2 | 3 | > 课程：[斯坦福博弈论课程官网](http://www.game-theory-class.org/) 4 | > 5 | > [斯坦福博弈论课程视频youtube版](https://www.youtube.com/user/gametheoryonline) 6 | > 7 | > [斯坦福博弈论课程视频coursera版 **Game Theory I** ](https://www.coursera.org/learn/game-theory-1) 8 | > 9 | > [斯坦福博弈论课程视频coursera版 **Game Theory-2** ](https://www.coursera.org/learn/game-theory-2) 10 | > 11 | > [斯坦福博弈论课程视频bilibili版 **Game Theory-I**待上传 ]() 12 | > 13 | > [斯坦福博弈论课程视频bilibili版 **Game Theory-2** ](https://www.bilibili.com/video/av63545278) 14 | > 15 | > * 由于不可知的原因 [GT02-1-01](https://www.bilibili.com/video/av63543124/)在这个链接看 16 | > * GT02-2-01需要fq观看,点击youtube和coursera地址观看 17 | > 18 | > 协议：[CC BY-NC-SA 4.0](http://creativecommons.org/licenses/by-nc-sa/4.0/) 19 | > 20 | > 欢迎任何人参与和完善：一个人可以走的很快，但是一群人却可以走的更远。 21 | 22 | + [斯坦福 CS229 机器学习中文笔记](http://ai-start.com/ml2014/) 23 | + [DeepLearning.ai 深度学习中文笔记](http://ai-start.com/dl2017/) 24 | + [斯坦福 CS234 强化学习中文讲义](https://github.com/apachecn/stanford-cs234-notes-zh) 25 | + [斯坦福 CS224n 自然语言处理中文笔记](https://github.com/apachecn/stanford-cs224n-notes-zh) 26 | + [ApacheCN 机器学习交流群 629470233](http://shang.qq.com/wpa/qunwpa?idkey=30e5f1123a79867570f665aa3a483ca404b1c3f77737bc01ec520ed5f078ddef) 27 | + [ApacheCN 学习资源](https://www.ibooker.org.cn/) 28 | 29 | ## 课程说明 30 | 31 | ### 博弈论 32 | 33 | *博弈论（英语：game theory），又译为对策论，或者赛局理论，应用数学的一个分支，1944年冯·诺伊曼与奥斯卡·摩根斯特恩合著《博弈论与经济行为》，标志着现代系统博弈理论的的初步形成，因此他被称为“博弈论之父”。博弈论被认为是20世纪经济学最伟大的成果之一。目前在生物学、经济学、国际关系、计算机科学、政治学、军事战略和其他很多学科都有广泛的应用。主要研究公式化了的激励结构（游戏或者博弈）间的相互作用。是研究具有斗争或竞争性质现象的数学理论和方法。也是运筹学的一个重要学科。* 34 | 35 | 支撑Google, 百度等互联网巨头广告业务的竞价排名机制的核心之一就是博弈论。另外经济学中有很多博弈论的影子，电影“美丽心灵”中的主角数学家约翰纳什，由于他与另外两位数学家在非合作博弈的均衡分析理论方面做出了开创性的贡献，对博弈论和经济学产生了重大影响，而获得1994年诺贝尔经济学奖，纳什均衡则是博弈论课程中不可或缺的一节课。 36 | 37 | ### 课程信息 38 | 39 | 斯坦福 **Game Theory** 为斯坦福博弈论课程，课程分两部分，**Game Theory I: An Introduction** 博弈论介绍，一共八周课程20个小时，[斯坦福博弈论1课程信息](http://www.game-theory-class.org/game-theory-I.html)；**Game Theory II: Advanced Applications** 博弈论高级应用，一共五次课程15个小时，[斯坦福博弈论2课程信息](http://www.game-theory-class.org/game-theory-II.html)，[coursera](https://www.coursera.org/learn/game-theory-1)中的最新课程为2016年8月10号录制 40 | 41 | ### 讲师 42 | 43 | ![]() 44 | 45 | [**Matthew O. Jackson**](https://web.stanford.edu/~jacksonm/) 46 | 47 | Stanford University 斯坦福大学 48 | 49 | Professor 教授 50 | 51 | Economics 经济学 52 | 53 | ![]() 54 | 55 | [**Kevin Leyton-Brown**](http://cs.ubc.ca/~kevinlb) 56 | 57 | University of British Columbia 英属哥伦比亚大学 58 | 59 | Professor 教授 60 | 61 | Computer Science 计算机科学 62 | 63 | ![](https://d3njjcbhbojbot.cloudfront.net/api/utilities/v1/imageproxy/https://coursera-instructor-photos.s3.amazonaws.com/25/002de5447b361051438576f26cf14f/Yoav-cropped-1.jpg?auto=format%2Ccompress&dpr=2&w=200&h=200) 64 | 65 | [**Yoav Shoham**](http://robotics.stanford.edu/~shoham/) 66 | 67 | Stanford University 斯坦福大学 68 | 69 | Professor 教授 70 | 71 | Computer Science 计算机科学 72 | 73 | ## 贡献指南 74 | 75 | 笔记整理活动正在进行，请查看[贡献指南](CONTRIBUTING.md)，并在[整体进度](https://github.com/apachecn/stanford-cs224n-notes-zh/issues/1)中领取任务。 76 | 77 | > 请您勇敢地去翻译和改进翻译。虽然我们追求卓越，但我们并不要求您做到十全十美，因此请不要担心因为翻译上犯错——在大部分情况下，我们的服务器已经记录所有的翻译，因此您不必担心会因为您的失误遭到无法挽回的破坏。（改编自维基百科） 78 | 79 | ## 联系方式 80 | 81 | ### 负责人 82 | 83 | + [viviwong](https://github.com/viviwong) 84 | 85 | ### 其他 86 | 87 | * 认领翻译和项目进度-地址: https://github.com/apachecn/stanford-game-theory-notes-zh/issues/1 88 | * 在我们的 [apachecn/stanford-game-theory-notes-zh](https://github.com/apachecn/stanford-game-theory-notes-zh) github 上提 issue. 89 | * 发邮件到 Email: `861787932@qq.com`. 90 | * 在我们的 [组织学习交流群](https://www.ibooker.org.cn/organization/) 中联系群主/管理员即可. 91 | 92 | ## 赞助我们 93 | 94 | ![](http://data.apachecn.org/img/about/donate.jpg) -------------------------------------------------------------------------------- /docs/.gitkeep: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/apachecn/stanford-game-theory-notes-zh/302105d046d9bc1f8b960f2c2473ee42f4cfeefe/docs/.gitkeep -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 1/GTO-1-02 - Self Interest and Utility Theory.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,210 --> 00:00:03,009 3 | we will be talking about self-interested 4 | 我们将谈论自私的 5 | 6 | 2 7 | 00:00:03,209 --> 00:00:04,658 8 | agents and their interactions 9 | 代理及其互动 10 | 11 | 3 12 | 00:00:04,859 --> 00:00:07,149 13 | so let's first speak about what we mean 14 | 所以我们先说说我们的意思 15 | 16 | 4 17 | 00:00:07,349 --> 00:00:14,290 18 | by self-interested agents we don't mean 19 | 对于自私的经纪人，我们并不是说 20 | 21 | 5 22 | 00:00:14,490 --> 00:00:17,829 23 | necessarily that agents are adversarial 24 | 必然是代理商具有对抗性 25 | 26 | 6 27 | 00:00:18,028 --> 00:00:21,010 28 | or don't care about what happened to 29 | 或不在乎发生了什么 30 | 31 | 7 32 | 00:00:21,210 --> 00:00:24,220 33 | other agents what we mean by that is 34 | 其他代理商，我们的意思是 35 | 36 | 8 37 | 00:00:24,420 --> 00:00:26,318 38 | that agents have opinions have 39 | 代理商有意见 40 | 41 | 9 42 | 00:00:26,518 --> 00:00:28,949 43 | preferences and so there's some 44 | 偏好，所以有一些 45 | 46 | 10 47 | 00:00:29,149 --> 00:00:31,568 48 | description of the world how the world 49 | 世界的描述世界如何 50 | 51 | 11 52 | 00:00:31,768 --> 00:00:33,518 53 | could be and in different descriptions 54 | 可能是并且在不同的描述中 55 | 56 | 12 57 | 00:00:33,719 --> 00:00:36,669 58 | the agents have different preferences 59 | 代理商有不同的偏好 60 | 61 | 13 62 | 00:00:36,869 --> 00:00:39,608 63 | and different utilities as we'll say and 64 | 以及我们将要说的不同的实用程序 65 | 66 | 14 67 | 00:00:39,808 --> 00:00:43,329 68 | so what we mean by utility function is a 69 | 所以我们所说的效用函数是 70 | 71 | 15 72 | 00:00:43,530 --> 00:00:47,619 73 | mathematical measure that tells you how 74 | 告诉你如何的数学测量 75 | 76 | 16 77 | 00:00:47,820 --> 00:00:50,709 78 | much the agents likes or does not like a 79 | 代理商非常喜欢或不喜欢 80 | 81 | 17 82 | 00:00:50,909 --> 00:00:52,119 83 | given situation 84 | 给定情况 85 | 86 | 18 87 | 00:00:52,320 --> 00:00:57,428 88 | it describes not only their attitude 89 | 它不仅描述了他们的态度 90 | 91 | 19 92 | 00:00:57,628 --> 00:01:00,279 93 | towards a definite of events so for 94 | 对于一定的事件，所以 95 | 96 | 20 97 | 00:01:00,479 --> 00:01:02,198 98 | example tomorrow the temperature will be 99 | 明天的温度将是 100 | 101 | 21 102 | 00:01:02,399 --> 00:01:07,988 103 | exactly 25 degrees centigrade but but in 104 | 正好25摄氏度，但 105 | 106 | 22 107 | 00:01:08,188 --> 00:01:09,730 108 | fact it will describe the preferences 109 | 实际上它将描述偏好 110 | 111 | 23 112 | 00:01:09,930 --> 00:01:15,129 113 | towards a distribution of such outcomes 114 | 分配这样的结果 115 | 116 | 24 117 | 00:01:15,329 --> 00:01:18,039 118 | so it really captures their attitude 119 | 所以这真的抓住了他们的态度 120 | 121 | 25 122 | 00:01:18,239 --> 00:01:21,399 123 | towards uncertainty about events so for 124 | 对事件的不确定性 125 | 126 | 26 127 | 00:01:21,599 --> 00:01:26,200 128 | example if I tell you that it will be 25 129 | 例如，如果我告诉你它将是25 130 | 131 | 27 132 | 00:01:26,400 --> 00:01:30,069 133 | degrees with probability 0.7 and 24 134 | 度为0.7和24 135 | 136 | 28 137 | 00:01:30,269 --> 00:01:33,308 138 | degrees with probability 0.3 you might 139 | 度，概率为0.3，您可能 140 | 141 | 29 142 | 00:01:33,509 --> 00:01:35,500 143 | have an opinion about how much you like 144 | 对你喜欢多少有意见 145 | 146 | 30 147 | 00:01:35,700 --> 00:01:38,399 148 | that versus some other distribution and 149 | 与其他发行版相比， 150 | 151 | 31 152 | 00:01:38,599 --> 00:01:42,429 153 | the decision theoretic approach which is 154 | 决策理论方法是 155 | 156 | 32 157 | 00:01:42,629 --> 00:01:45,759 158 | what underlies modern game theory says 159 | 现代博弈论的基础是什么 160 | 161 | 33 162 | 00:01:45,959 --> 00:01:47,469 163 | that you're going to try to act in the 164 | 你将要尝试采取行动 165 | 166 | 34 167 | 00:01:47,670 --> 00:01:50,528 168 | way that maximizes your expected or 169 | 最大化您的预期或 170 | 171 | 35 172 | 00:01:50,728 --> 00:01:54,640 173 | average utility and so this is a concept 174 | 平均效用，所以这是一个概念 175 | 176 | 36 177 | 00:01:54,840 --> 00:01:59,439 178 | we need to get comfortable with and and 179 | 我们需要对and和 180 | 181 | 37 182 | 00:01:59,640 --> 00:02:02,230 183 | it's not obvious that one would want to 184 | 人们是否愿意 185 | 186 | 38 187 | 00:02:02,430 --> 00:02:06,128 188 | use such an approach so for example we 189 | 使用这种方法，例如 190 | 191 | 39 192 | 00:02:06,328 --> 00:02:08,800 193 | are going to look at a single dimension 194 | 将要看一个单一的维度 195 | 196 | 40 197 | 00:02:09,000 --> 00:02:11,170 198 | so your preferences will 199 | 所以你的喜好会 200 | 201 | 41 202 | 00:02:11,370 --> 00:02:15,849 203 | all beyond a scale as we'll see the 204 | 一切都超出规模，我们将看到 205 | 206 | 42 207 | 00:02:16,049 --> 00:02:17,920 208 | scale that is not that important unlike 209 | 规模不那么重要 210 | 211 | 43 212 | 00:02:18,120 --> 00:02:19,750 213 | probabilities utilities don't have to 214 | 概率实用程序不必 215 | 216 | 44 217 | 00:02:19,949 --> 00:02:22,810 218 | lie in the 0-1 scale but they will lie 219 | 说谎在0-1规模，但他们会说谎 220 | 221 | 45 222 | 00:02:23,009 --> 00:02:26,830 223 | on a linear dimension and maybe maybe 224 | 在线性维度上，也许 225 | 226 | 46 227 | 00:02:27,030 --> 00:02:29,860 228 | that's inappropriate or for example you 229 | 那是不合适的，例如你 230 | 231 | 47 232 | 00:02:30,060 --> 00:02:32,950 233 | might have some level of wealth and some 234 | 可能有一定程度的财富和一些 235 | 236 | 48 237 | 00:02:33,150 --> 00:02:37,270 238 | degree of health and for a certain level 239 | 健康程度和一定水平 240 | 241 | 49 242 | 00:02:37,469 --> 00:02:41,679 243 | of of each one you'll have some some 244 | 每一个中，您都会有一些 245 | 246 | 50 247 | 00:02:41,878 --> 00:02:46,179 248 | some some notion of well-being but is it 249 | 一些关于幸福的概念，但这是 250 | 251 | 51 252 | 00:02:46,378 --> 00:02:48,670 253 | appropriate to put the two together and 254 | 适合将两者放在一起 255 | 256 | 52 257 | 00:02:48,870 --> 00:02:50,439 258 | have a single scale you might you might 259 | 有一个单一的规模，你可能会 260 | 261 | 53 262 | 00:02:50,639 --> 00:02:54,819 263 | question that similarly why is looking 264 | 同样为什么要看的问题 265 | 266 | 54 267 | 00:02:55,019 --> 00:02:56,649 268 | at the expected value when you're 269 | 当您处于期望值时 270 | 271 | 55 272 | 00:02:56,848 --> 00:02:58,659 273 | looking at your uncertainty while 274 | 看着你的不确定性 275 | 276 | 56 277 | 00:02:58,859 --> 00:03:00,939 278 | looking at the expected value an 279 | 看期望值 280 | 281 | 57 282 | 00:03:01,139 --> 00:03:02,860 283 | appropriate way to capture your attitude 284 | 捕捉态度的适当方法 285 | 286 | 58 287 | 00:03:03,060 --> 00:03:07,330 288 | and so these are not trivial statements 289 | 所以这些不是琐碎的陈述 290 | 291 | 59 292 | 00:03:07,530 --> 00:03:10,209 293 | and in fact are not tautological they 294 | 而且实际上不是重言式的 295 | 296 | 60 297 | 00:03:10,408 --> 00:03:13,569 298 | make a subjective claim but there's a 299 | 提出主观主张，但有一个 300 | 301 | 61 302 | 00:03:13,769 --> 00:03:16,599 303 | very long tradition and maybe the most 304 | 悠久的传统，也许是最悠久的 305 | 306 | 62 307 | 00:03:16,799 --> 00:03:20,319 308 | famous reference is to phenomena 309 | 著名的参考是现象 310 | 311 | 63 312 | 00:03:20,519 --> 00:03:26,560 313 | Morgenstern simul book on on which is 314 | Morgenstern simul上的书是 315 | 316 | 64 317 | 00:03:26,759 --> 00:03:28,179 318 | really in some ways the introduction to 319 | 确实在某种程度上介绍了 320 | 321 | 65 322 | 00:03:28,378 --> 00:03:31,000 323 | modern day in game theory that derives 324 | 博弈论中的现代派生 325 | 326 | 66 327 | 00:03:31,199 --> 00:03:34,990 328 | the these utility function from more 329 | 这些实用功能更多 330 | 331 | 67 332 | 00:03:35,189 --> 00:03:38,649 333 | basic assumptions one makes and we won't 334 | 一个基本假设，我们不会 335 | 336 | 68 337 | 00:03:38,848 --> 00:03:40,990 338 | go into that but we just wanted to flag 339 | 进入那个，但是我们只想标记 340 | 341 | 69 342 | 00:03:41,189 --> 00:03:43,840 343 | this issue as something that will 344 | 这个问题会 345 | 346 | 70 347 | 00:03:44,039 --> 00:03:48,550 348 | underlie everything we say about game 349 | 成为我们谈论游戏的基础 350 | 351 | 71 352 | 00:03:48,750 --> 00:03:51,340 353 | theory and which really underlies one 354 | 理论，这实际上是一个基础 355 | 356 | 72 357 | 00:03:51,539 --> 00:03:56,539 358 | game theory 359 | 博弈论 360 | 361 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 1/GTO-1-04 - Examples of Games from Game Theory.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,030 --> 00:00:03,969 3 | in the celebrated example of prisoner's 4 | 在囚犯的著名例子中 5 | 6 | 2 7 | 00:00:04,169 --> 00:00:05,859 8 | dilemma we have this general scheme 9 | 困境，我们有这个总体方案 10 | 11 | 3 12 | 00:00:06,059 --> 00:00:08,530 13 | where the two prisoners can either 14 | 两个囚犯可以 15 | 16 | 4 17 | 00:00:08,730 --> 00:00:12,550 18 | cooperate or not defect as is called if 19 | 合作或不存在缺陷 20 | 21 | 5 22 | 00:00:12,750 --> 00:00:15,460 23 | they both cooperate they get some payoff 24 | 他们俩都合作，他们得到了一些回报 25 | 26 | 6 27 | 00:00:15,660 --> 00:00:17,530 28 | a-and if they both defect they get a 29 | -如果他们俩都叛逆，他们会得到一个 30 | 31 | 7 32 | 00:00:17,730 --> 00:00:20,199 33 | different payoff D where a is greater 34 | a较大时的不同收益D 35 | 36 | 8 37 | 00:00:20,399 --> 00:00:24,669 38 | than D however if they miss coordinate 39 | 比D但是如果他们错过坐标 40 | 41 | 9 42 | 00:00:24,868 --> 00:00:26,980 43 | and one of them cooperates and the other 44 | 其中一个合作，另一个 45 | 46 | 10 47 | 00:00:27,179 --> 00:00:30,460 48 | defects then the cooperator gets the 49 | 缺陷，然后合作者得到 50 | 51 | 11 52 | 00:00:30,660 --> 00:00:32,768 53 | lowest possible payoff and the defector 54 | 最低的回报和叛逃者 55 | 56 | 12 57 | 00:00:32,969 --> 00:00:34,899 58 | gets the largest possible path and 59 | 得到最大可能的路径 60 | 61 | 13 62 | 00:00:35,100 --> 00:00:37,329 63 | that's true symmetrically here as well 64 | 在这里对称也是如此 65 | 66 | 14 67 | 00:00:37,530 --> 00:00:41,320 68 | and this a very well-known example that 69 | 这是一个非常著名的例子 70 | 71 | 15 72 | 00:00:41,520 --> 00:00:44,099 73 | has a rather counterintuitive 74 | 有点违反直觉 75 | 76 | 16 77 | 00:00:44,299 --> 00:00:49,088 78 | paradoxical properties most games are 79 | 大多数游戏的悖论性质 80 | 81 | 17 82 | 00:00:49,289 --> 00:00:55,689 83 | not as conceptually confusing here's an 84 | 没有概念上的混乱，这是 85 | 86 | 18 87 | 00:00:55,890 --> 00:00:58,149 88 | example that conceptually very clear and 89 | 这个例子在概念上非常清楚， 90 | 91 | 19 92 | 00:00:58,350 --> 00:01:00,788 93 | these are gains of pure competition the 94 | 这些都是纯竞争的收益 95 | 96 | 20 97 | 00:01:00,988 --> 00:01:03,070 98 | situation here is limited to two players 99 | 这种情况仅限于两名球员 100 | 101 | 21 102 | 00:01:03,270 --> 00:01:07,140 103 | where one players payoff is exactly the 104 | 一个玩家的收益正好是 105 | 106 | 22 107 | 00:01:07,340 --> 00:01:10,869 108 | complement of another players payoff so 109 | 补充另一个玩家的收益，所以 110 | 111 | 23 112 | 00:01:11,069 --> 00:01:13,988 113 | they always sum to some constant C often 114 | 他们经常总和一些常数C 115 | 116 | 24 117 | 00:01:14,188 --> 00:01:18,009 118 | that concept is that we use zero and we 119 | 这个概念是我们使用零，我们 120 | 121 | 25 122 | 00:01:18,209 --> 00:01:20,230 123 | call them for that reason zero-sum games 124 | 称他们为零和游戏 125 | 126 | 26 127 | 00:01:20,430 --> 00:01:22,119 128 | as opposed to constant sum games 129 | 与固定和博弈相反 130 | 131 | 27 132 | 00:01:22,319 --> 00:01:27,849 133 | and since they do sum to zero or to 134 | 由于它们的总和为零或 135 | 136 | 28 137 | 00:01:28,049 --> 00:01:31,119 138 | constant we only need to remember one 139 | 不变，我们只需要记住一个 140 | 141 | 29 142 | 00:01:31,319 --> 00:01:33,278 143 | number the path to one of the players 144 | 编号玩家之一的路径 145 | 146 | 30 147 | 00:01:33,478 --> 00:01:36,668 148 | and we can infer the payoff to the other 149 | 我们可以推断其他人的收益 150 | 151 | 31 152 | 00:01:36,868 --> 00:01:41,528 153 | player from that here's the most simple 154 | 玩家从这里最简单 155 | 156 | 32 157 | 00:01:41,728 --> 00:01:43,808 158 | version of it this is a games of 159 | 它的版本是 160 | 161 | 33 162 | 00:01:44,009 --> 00:01:46,599 163 | matching pennies so you and I each need 164 | 匹配便士，所以您和我都需要 165 | 166 | 34 167 | 00:01:46,799 --> 00:01:49,840 168 | to pick either heads or tail for the for 169 | 为之挑选头或尾 170 | 171 | 35 172 | 00:01:50,040 --> 00:01:54,189 173 | the coin if we pick the same side either 174 | 如果我们选择同一面的话 175 | 176 | 36 177 | 00:01:54,390 --> 00:01:58,149 178 | heads or tails I win which means that I 179 | 头或尾我赢了，这意味着我 180 | 181 | 37 182 | 00:01:58,349 --> 00:02:01,599 183 | get a payoff of 1 and U of minus 1 if we 184 | 如果我们得到1的收益和-1的U 185 | 186 | 38 187 | 00:02:01,799 --> 00:02:04,569 188 | miss coordinate and so I pick heads and 189 | 想念坐标，所以我挑了头 190 | 191 | 39 192 | 00:02:04,769 --> 00:02:06,878 193 | you tailed or the other way around then 194 | 你拖尾或反过来 195 | 196 | 40 197 | 00:02:07,078 --> 00:02:10,000 198 | you win a very straightforward game of 199 | 您会赢得一个非常简单的游戏 200 | 201 | 41 202 | 00:02:10,199 --> 00:02:12,080 203 | pure competition 204 | 纯竞争 205 | 206 | 42 207 | 00:02:12,280 --> 00:02:14,210 208 | here's another word our very well-known 209 | 这是我们非常知名的另一个词 210 | 211 | 43 212 | 00:02:14,409 --> 00:02:17,090 213 | similar games with three actions both of 214 | 具有三个动作的相似游戏 215 | 216 | 44 217 | 00:02:17,289 --> 00:02:20,060 218 | us and that's the game of rock papers 219 | 我们就是摇滚游戏 220 | 221 | 45 222 | 00:02:20,259 --> 00:02:21,130 223 | and scissors 224 | 和剪刀 225 | 226 | 46 227 | 00:02:21,330 --> 00:02:25,670 228 | also known as Rochambeau and so if we 229 | 也被称为Rochambeau，所以如果我们 230 | 231 | 47 232 | 00:02:25,870 --> 00:02:29,360 233 | pick the same action then it's a draw 234 | 选择相同的动作然后平局 235 | 236 | 48 237 | 00:02:29,560 --> 00:02:33,500 238 | and otherwise there are rules for who 239 | 否则有谁的规则 240 | 241 | 49 242 | 00:02:33,699 --> 00:02:36,020 243 | wins for example if I pick Rock and you 244 | 例如，如果我选择Rock而您获胜 245 | 246 | 50 247 | 00:02:36,219 --> 00:02:38,210 248 | paper the new one if I picked Rock and 249 | 如果我选择了Rock， 250 | 251 | 51 252 | 00:02:38,409 --> 00:02:42,800 253 | use scissors then I win and so on again 254 | 用剪刀，然后我赢了，依此类推 255 | 256 | 52 257 | 00:02:43,000 --> 00:02:46,600 258 | the payoffs in both cases sum to zero 259 | 两种情况下的收益总和为零 260 | 261 | 53 262 | 00:02:46,800 --> 00:02:52,490 263 | this parenthetically this very simple 264 | 这很简单 265 | 266 | 54 267 | 00:02:52,689 --> 00:02:55,700 268 | children game actually has an annual 269 | 儿童游戏实际上每年都有 270 | 271 | 55 272 | 00:02:55,900 --> 00:02:58,430 273 | competition that carries a non-trivial 274 | 不平凡的竞争 275 | 276 | 56 277 | 00:02:58,629 --> 00:03:01,580 278 | prize of $10,000 and it's actually a 279 | 奖金$ 10,000，实际上是 280 | 281 | 57 282 | 00:03:01,780 --> 00:03:02,540 283 | sobering thought 284 | 清醒的思想 285 | 286 | 58 287 | 00:03:02,740 --> 00:03:05,150 288 | that when we look at this trivial game 289 | 当我们看这个琐碎的游戏时 290 | 291 | 59 292 | 00:03:05,349 --> 00:03:08,330 293 | then perhaps check a little bit if we 294 | 然后也许检查一下 295 | 296 | 60 297 | 00:03:08,530 --> 00:03:10,759 298 | actually participate in this competition 299 | 参加比赛 300 | 301 | 61 302 | 00:03:10,959 --> 00:03:12,230 303 | we'd actually think hard about how to 304 | 我们实际上会认真思考如何 305 | 306 | 62 307 | 00:03:12,430 --> 00:03:18,890 308 | play it is the other extreme of games of 309 | 玩是游戏的另一极端 310 | 311 | 63 312 | 00:03:19,090 --> 00:03:22,670 313 | pure coordination or pure cooperation in 314 | 纯粹的协调或纯粹的合作 315 | 316 | 64 317 | 00:03:22,870 --> 00:03:25,939 318 | this case all agents have exactly the 319 | 在这种情况下，所有代理商都有 320 | 321 | 65 322 | 00:03:26,139 --> 00:03:28,039 323 | same interest in other words their 324 | 换句话说，他们的兴趣相同 325 | 326 | 66 327 | 00:03:28,239 --> 00:03:32,569 328 | payoffs for every action vector that 329 | 每个动作向量的收益 330 | 331 | 67 332 | 00:03:32,769 --> 00:03:37,610 333 | they take is the same and so the utility 334 | 他们采取的是相同的，所以实用程序 335 | 336 | 68 337 | 00:03:37,810 --> 00:03:40,039 338 | for play I is always the same of utility 339 | 对于玩游戏，我永远都是相同的工具 340 | 341 | 69 342 | 00:03:40,239 --> 00:03:44,539 343 | for pair J for every action sequence 344 | 对于J对每个动作序列 345 | 346 | 70 347 | 00:03:44,739 --> 00:03:47,990 348 | that they choose and so again we here - 349 | 他们选择，所以我们再次在这里- 350 | 351 | 71 352 | 00:03:48,189 --> 00:03:50,450 353 | we only need to write in each cell of 354 | 我们只需要在 355 | 356 | 72 357 | 00:03:50,650 --> 00:03:52,160 358 | the matrix only one number because it's 359 | 矩阵只有一个数字，因为 360 | 361 | 73 362 | 00:03:52,360 --> 00:03:55,099 363 | common to all the players it drives home 364 | 它带回家的所有球员都是共同的 365 | 366 | 74 367 | 00:03:55,299 --> 00:03:58,240 368 | the perhaps the unfortunate term 369 | 也许是不幸的术语 370 | 371 | 75 372 | 00:03:58,439 --> 00:04:01,250 373 | noncorporeal game theory that describes 374 | 非物质博弈论 375 | 376 | 76 377 | 00:04:01,449 --> 00:04:03,259 378 | this dominant strand of game theory 379 | 博弈论的主导 380 | 381 | 77 382 | 00:04:03,459 --> 00:04:06,439 383 | we're discussing for now it's the name 384 | 我们现在正在讨论的是名字 385 | 386 | 78 387 | 00:04:06,639 --> 00:04:08,710 388 | would suggest that these are games for 389 | 会建议这些是 390 | 391 | 79 392 | 00:04:08,909 --> 00:04:10,910 393 | that describe situations that are 394 | 描述情况 395 | 396 | 80 397 | 00:04:11,110 --> 00:04:14,180 398 | inherent inherently conflictual but as 399 | 内在固有的冲突，但作为 400 | 401 | 81 402 | 00:04:14,379 --> 00:04:17,598 403 | we see they apply also to games in which 404 | 我们看到它们也适用于其中 405 | 406 | 82 407 | 00:04:17,798 --> 00:04:20,750 408 | the interests of the players coincide so 409 | 玩家的利益是一致的 410 | 411 | 83 412 | 00:04:20,949 --> 00:04:23,350 413 | here's a here's a game that 414 | 这是一个游戏 415 | 416 | 84 417 | 00:04:23,550 --> 00:04:28,600 418 | describes the purely court cooperative 419 | 描述纯粹的法院合作社 420 | 421 | 85 422 | 00:04:28,800 --> 00:04:31,259 423 | situation you and I walk each other 424 | 你和我互相走动的情况 425 | 426 | 86 427 | 00:04:31,459 --> 00:04:34,120 428 | towards each other on the sidewalk we 429 | 在人行道上彼此接近 430 | 431 | 87 432 | 00:04:34,319 --> 00:04:36,129 433 | can each decide whether to go to our 434 | 每个人都可以决定是否去我们的 435 | 436 | 88 437 | 00:04:36,329 --> 00:04:37,840 438 | respective lay lifts our respective 439 | 各自的升降机 440 | 441 | 89 442 | 00:04:38,040 --> 00:04:40,900 443 | right and if we pick the same side then 444 | 对，如果我们选择同一边，那么 445 | 446 | 90 447 | 00:04:41,100 --> 00:04:44,079 448 | all is good we avoid a collision if we 449 | 一切都很好，如果我们避免碰撞 450 | 451 | 91 452 | 00:04:44,279 --> 00:04:48,550 453 | don't then the then we do collide and 454 | 然后，那么我们就发生了碰撞， 455 | 456 | 92 457 | 00:04:48,750 --> 00:04:52,810 458 | that's equally bad for both of us of 459 | 这对我们两个人都同样有害 460 | 461 | 93 462 | 00:04:53,009 --> 00:04:55,110 463 | course in general gains will be neither 464 | 当然，总的收益不会 465 | 466 | 94 467 | 00:04:55,310 --> 00:04:58,020 468 | purely cooperative nor purely 469 | 纯粹合作也不是纯粹 470 | 471 | 95 472 | 00:04:58,220 --> 00:05:02,460 473 | conflictual and here's a game that 474 | 冲突，这是一个游戏 475 | 476 | 96 477 | 00:05:02,660 --> 00:05:05,050 478 | exemplifies that this is a game that 479 | 举例说明这是一款 480 | 481 | 97 482 | 00:05:05,250 --> 00:05:07,900 483 | called battle of the sexes imagine a 484 | 所谓的性别之战想象 485 | 486 | 98 487 | 00:05:08,100 --> 00:05:10,629 488 | husband and a wife we want to go out to 489 | 丈夫和妻子，我们想出去 490 | 491 | 99 492 | 00:05:10,829 --> 00:05:13,300 493 | a movie there are two movies they could 494 | 电影可以有两部电影 495 | 496 | 100 497 | 00:05:13,500 --> 00:05:16,870 498 | choose from let's say Battle of 499 | 从说“战役”中选择 500 | 501 | 101 502 | 00:05:17,069 --> 00:05:21,759 503 | Armageddon and flower child the one a 504 | 大决战和花童一个 505 | 506 | 102 507 | 00:05:21,959 --> 00:05:25,900 508 | violent war movie and the other is a is 509 | 暴力战争电影，另一个是 510 | 511 | 103 512 | 00:05:26,100 --> 00:05:30,100 513 | a romantic comedy above all they want to 514 | 他们想要的浪漫喜剧 515 | 516 | 104 517 | 00:05:30,300 --> 00:05:31,660 518 | go together to the movie if they go to 519 | 一起去看电影，如果他们去 520 | 521 | 105 522 | 00:05:31,860 --> 00:05:34,840 523 | different movies then they are equally 524 | 不同的电影，他们都是一样的 525 | 526 | 106 527 | 00:05:35,040 --> 00:05:36,910 528 | unhappy so they want to go to the same 529 | 不开心，所以他们想去同一个 530 | 531 | 107 532 | 00:05:37,110 --> 00:05:38,530 533 | movie but they have conflicting 534 | 电影，但他们有冲突 535 | 536 | 108 537 | 00:05:38,730 --> 00:05:41,980 538 | preferences the wife clearly would 539 | 妻子的喜好显然会 540 | 541 | 109 542 | 00:05:42,180 --> 00:05:47,250 543 | prefer to go to battle of Armageddon and 544 | 宁愿去参加世界末日之战， 545 | 546 | 110 547 | 00:05:47,449 --> 00:05:50,379 548 | husband romantic as he is would like to 549 | 丈夫浪漫，因为他想 550 | 551 | 111 552 | 00:05:50,579 --> 00:05:51,460 553 | go to flower child 554 | 去花童 555 | 556 | 112 557 | 00:05:51,660 --> 00:05:56,860 558 | so both cooperation and competition in 559 | 因此，在 560 | 561 | 113 562 | 00:05:57,060 --> 00:06:02,060 563 | this game 564 | 这个游戏 565 | 566 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 1/GTO-1-05 - Nash Equilibrium Introduction, and the Keynes Beauty Contest.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,000 --> 00:00:02,889 3 | hi folks its Matt again and now we're 4 | 嗨，大家好，Matt，现在我们 5 | 6 | 2 7 | 00:00:03,089 --> 00:00:04,509 8 | ready to start solving games and making 9 | 准备开始解决游戏和制作 10 | 11 | 3 12 | 00:00:04,710 --> 00:00:06,519 13 | some predictions of how people will play 14 | 关于人们如何玩的一些预测 15 | 16 | 4 17 | 00:00:06,719 --> 00:00:08,500 18 | in different settings and so we're 19 | 在不同的设置中，所以我们 20 | 21 | 5 22 | 00:00:08,699 --> 00:00:11,859 23 | talking right now about Nash equilibrium 24 | 现在谈论纳什均衡 25 | 26 | 6 27 | 00:00:12,058 --> 00:00:14,950 28 | which is one probably the most basic and 29 | 这可能是最基本的 30 | 31 | 7 32 | 00:00:15,150 --> 00:00:17,260 33 | standard solution concept of all in all 34 | 全部的标准解决方案概念 35 | 36 | 8 37 | 00:00:17,460 --> 00:00:19,600 38 | of game theory it's named after John 39 | 博弈论的名字，以约翰命名 40 | 41 | 9 42 | 00:00:19,800 --> 00:00:21,310 43 | Nash who was a mathematician at 44 | 纳什（Nash）是数学家 45 | 46 | 10 47 | 00:00:21,510 --> 00:00:23,470 48 | Princeton and actually some years back 49 | 普林斯顿，实际上几年前 50 | 51 | 11 52 | 00:00:23,670 --> 00:00:25,329 53 | won the Nobel Prize for his work on this 54 | 他在这方面的工作获得了诺贝尔奖 55 | 56 | 12 57 | 00:00:25,528 --> 00:00:28,510 58 | subject and it's a very basic and 59 | 主题，这是一个非常基本的 60 | 61 | 13 62 | 00:00:28,710 --> 00:00:31,329 63 | fundamental concept and in order to sort 64 | 基本概念和排序 65 | 66 | 14 67 | 00:00:31,528 --> 00:00:33,070 68 | of motivate it let's let's start by just 69 | 激励它，让我们从一开始就开始 70 | 71 | 15 72 | 00:00:33,270 --> 00:00:38,349 73 | talking through some a particular game 74 | 通过特定游戏进行交谈 75 | 76 | 16 77 | 00:00:38,549 --> 00:00:41,320 78 | that was described and invented by 79 | 被描述和发明的 80 | 81 | 17 82 | 00:00:41,520 --> 00:00:45,189 83 | another famous person so this is John 84 | 另一个名人，所以这是约翰 85 | 86 | 18 87 | 00:00:45,390 --> 00:00:48,219 88 | Kane John Maynard Keynes beauty contest 89 | 凯恩·约翰·梅纳德·凯恩斯选美大赛 90 | 91 | 19 92 | 00:00:48,420 --> 00:00:51,998 93 | game so what's the idea here so let's 94 | 游戏，这是什么主意，让我们 95 | 96 | 20 97 | 00:00:52,198 --> 00:00:53,919 98 | let's think of a basic situation that 99 | 让我们考虑一个基本情况 100 | 101 | 21 102 | 00:00:54,119 --> 00:00:57,038 103 | you might be interested in and this was 104 | 您可能对此感兴趣，这是 105 | 106 | 22 107 | 00:00:57,238 --> 00:00:58,658 108 | one that Keynes described in some detail 109 | 凯恩斯对此进行了详细描述 110 | 111 | 23 112 | 00:00:58,859 --> 00:01:01,419 113 | so the idea was you have a stock you're 114 | 所以想法是你有一只股票 115 | 116 | 24 117 | 00:01:01,619 --> 00:01:04,388 118 | holding on to it and the stock price is 119 | 坚持下去，股价就是 120 | 121 | 25 122 | 00:01:04,588 --> 00:01:07,090 123 | rising that's great you're an investor 124 | 上升那太好了，您是投资者 125 | 126 | 26 127 | 00:01:07,290 --> 00:01:08,769 128 | you're trying to make profits off of 129 | 你试图从中获利 130 | 131 | 27 132 | 00:01:08,969 --> 00:01:12,278 133 | your your stock holdings and you begin 134 | 您的库存量，然后开始 135 | 136 | 28 137 | 00:01:12,478 --> 00:01:13,869 138 | to believe that maybe the stock is too 139 | 相信也许股票也是如此 140 | 141 | 29 142 | 00:01:14,069 --> 00:01:15,849 143 | high to be justified by the value of the 144 | 高要由价值来证明 145 | 146 | 30 147 | 00:01:16,049 --> 00:01:17,799 148 | company so you're thinking that it's 149 | 公司，所以您认为这是 150 | 151 | 31 152 | 00:01:18,000 --> 00:01:19,599 153 | possible that this stock is overvalued 154 | 这只股票可能被高估了 155 | 156 | 32 157 | 00:01:19,799 --> 00:01:22,209 158 | maybe there's a bubble in the market and 159 | 也许市场上有泡沫， 160 | 161 | 33 162 | 00:01:22,409 --> 00:01:24,629 163 | you're starting to think about selling 164 | 您开始考虑出售 165 | 166 | 34 167 | 00:01:24,829 --> 00:01:27,308 168 | okay well you'd like to sell it but 169 | 好吧，你想卖掉它，但是 170 | 171 | 35 172 | 00:01:27,509 --> 00:01:28,840 173 | you'd like to wait until the price is at 174 | 您想等到价格在 175 | 176 | 36 177 | 00:01:29,040 --> 00:01:31,090 178 | its peak right so you'd want to wait 179 | 它的峰值合适，所以您想等待 180 | 181 | 37 182 | 00:01:31,290 --> 00:01:33,009 183 | until the price was just where it's 184 | 直到价格刚好在 185 | 186 | 38 187 | 00:01:33,209 --> 00:01:35,909 188 | going to hit its maximum before you sold 189 | 在您卖出之前将达到最高 190 | 191 | 39 192 | 00:01:36,109 --> 00:01:37,959 193 | so you wanted to get out of the market 194 | 所以你想退出市场 195 | 196 | 40 197 | 00:01:38,159 --> 00:01:40,539 198 | just before the other investors do so 199 | 就在其他投资者这样做之前 200 | 201 | 41 202 | 00:01:40,739 --> 00:01:41,950 203 | this is a game where now you have to 204 | 这是一个游戏，现在您必须 205 | 206 | 42 207 | 00:01:42,150 --> 00:01:43,390 208 | predict what other people think about 209 | 预测别人的想法 210 | 211 | 43 212 | 00:01:43,590 --> 00:01:45,189 213 | the stock price and and what they're 214 | 股票价格以及它们的含义 215 | 216 | 44 217 | 00:01:45,390 --> 00:01:46,778 218 | going to do and when they want to get 219 | 去做，什么时候想得到 220 | 221 | 45 222 | 00:01:46,978 --> 00:01:50,168 223 | out so how will they act how should you 224 | 出来，他们将如何行动，你应该如何 225 | 226 | 46 227 | 00:01:50,368 --> 00:01:53,049 228 | respond to that so this is the basic 229 | 对此做出回应，这是最基本的 230 | 231 | 47 232 | 00:01:53,250 --> 00:01:55,149 233 | ingredients of Nash equilibrium are 234 | 纳什均衡的成分是 235 | 236 | 48 237 | 00:01:55,349 --> 00:01:57,009 238 | going to be having some prediction of 239 | 会有一些预测 240 | 241 | 49 242 | 00:01:57,209 --> 00:01:58,778 243 | what other players are doing and then 244 | 其他玩家在做什么，然后 245 | 246 | 50 247 | 00:01:58,978 --> 00:02:00,579 248 | choosing the optimal strategy in 249 | 选择最佳策略 250 | 251 | 51 252 | 00:02:00,780 --> 00:02:01,959 253 | response to that so these are they're 254 | 回应，所以这些是 255 | 256 | 52 257 | 00:02:02,159 --> 00:02:03,579 258 | going to be two key ingredients that we 259 | 将成为我们的两个关键要素 260 | 261 | 53 262 | 00:02:03,780 --> 00:02:07,000 263 | have so there's a very stylized version 264 | 有一个非常程式化的版本 265 | 266 | 54 267 | 00:02:07,200 --> 00:02:09,880 268 | of this which is known as the the Keynes 269 | 这就是凯恩斯 270 | 271 | 55 272 | 00:02:10,080 --> 00:02:12,850 273 | beauty contest game where did it come 274 | 选美比赛是哪里来的 275 | 276 | 56 277 | 00:02:13,050 --> 00:02:13,600 278 | from 279 | 从 280 | 281 | 57 282 | 00:02:13,800 --> 00:02:17,140 283 | actually Keynes described the there was 284 | 实际上凯恩斯描述了 285 | 286 | 58 287 | 00:02:17,340 --> 00:02:22,950 288 | a newspaper in the in England that had a 289 | 英格兰的一家报纸 290 | 291 | 59 292 | 00:02:23,150 --> 00:02:26,500 293 | contest where players had to guess which 294 | 比赛中玩家必须猜测哪个 295 | 296 | 60 297 | 00:02:26,699 --> 00:02:32,290 298 | picture of several women other readers 299 | 其他读者的几个女人的照片 300 | 301 | 61 302 | 00:02:32,490 --> 00:02:36,460 303 | would think was the the the most 304 | 会认为是最 305 | 306 | 62 307 | 00:02:36,659 --> 00:02:38,290 308 | attractive one so it wasn't to guess 309 | 有吸引力的一个，所以不用猜测 310 | 311 | 63 312 | 00:02:38,490 --> 00:02:39,910 313 | what you thought but what you thought 314 | 你在想什么，但你在想什么 315 | 316 | 64 317 | 00:02:40,110 --> 00:02:43,620 318 | other people were thinking so Keynes 319 | 其他人在想，所以凯恩斯 320 | 321 | 65 322 | 00:02:43,819 --> 00:02:47,290 323 | likened investing to this you it's not 324 | 比作投资，这不是 325 | 326 | 66 327 | 00:02:47,490 --> 00:02:49,900 328 | only what you think of this the stock 329 | 只有您对此的看法 330 | 331 | 67 332 | 00:02:50,099 --> 00:02:51,340 333 | but what you think other people are 334 | 但是你认为别人是什么 335 | 336 | 68 337 | 00:02:51,539 --> 00:02:52,540 338 | thinking about the stock that's 339 | 考虑那只股票 340 | 341 | 69 342 | 00:02:52,740 --> 00:02:54,130 343 | important and driving your decisions 344 | 重要并推动您的决策 345 | 346 | 70 347 | 00:02:54,330 --> 00:02:57,280 348 | okay so this now is represented by a 349 | 好吧，现在以 350 | 351 | 71 352 | 00:02:57,479 --> 00:02:59,530 353 | very simple game that is is played by 354 | 玩的非常简单的游戏 355 | 356 | 72 357 | 00:02:59,729 --> 00:03:02,230 358 | many people so what's this game look 359 | 很多人，所以这个游戏看起来像什么 360 | 361 | 73 362 | 00:03:02,430 --> 00:03:06,400 363 | like each person gets to name an integer 364 | 就像每个人都用一个整数命名 365 | 366 | 74 367 | 00:03:06,599 --> 00:03:08,890 368 | between 1 and 100 ok so you get to pick 369 | 在1到100 ok之间，所以您可以选择 370 | 371 | 75 372 | 00:03:09,090 --> 00:03:11,230 373 | a number between 1 and 100 s to be an 374 | 1到100秒之间的数字是一个 375 | 376 | 76 377 | 00:03:11,430 --> 00:03:15,340 378 | integer so 1 2 3 etc players are going 379 | 整数，所以1 2 3等玩家进行 380 | 381 | 77 382 | 00:03:15,539 --> 00:03:17,800 383 | to move simultaneously and the player 384 | 与玩家同时移动 385 | 386 | 78 387 | 00:03:18,000 --> 00:03:19,900 388 | who names the integer that's closest to 389 | 谁命名最接近的整数 390 | 391 | 79 392 | 00:03:20,099 --> 00:03:23,020 393 | two-thirds of the average integer wins a 394 | 平均整数的三分之二赢得 395 | 396 | 80 397 | 00:03:23,219 --> 00:03:25,900 398 | prize and the other players get nothing 399 | 奖，其他玩家一无所获 400 | 401 | 81 402 | 00:03:26,099 --> 00:03:28,509 403 | so to win this game you have to get you 404 | 所以要赢得这场比赛，你必须让你 405 | 406 | 82 407 | 00:03:28,709 --> 00:03:32,050 408 | have to guess the average and then 2/3 409 | 必须猜测平均值，然后再2/3 410 | 411 | 83 412 | 00:03:32,250 --> 00:03:34,120 413 | of it right so you want to be right at 414 | 它是正确的，所以你想在正确 415 | 416 | 84 417 | 00:03:34,319 --> 00:03:36,280 418 | 2/3 of whatever the average guess is so 419 | 平均猜测的2/3 420 | 421 | 85 422 | 00:03:36,479 --> 00:03:39,520 423 | a little bit below the average guess if 424 | 比平均猜测低一点 425 | 426 | 86 427 | 00:03:39,719 --> 00:03:41,289 428 | there's two people that happen to hit 429 | 有两个人碰巧 430 | 431 | 87 432 | 00:03:41,489 --> 00:03:43,240 433 | the same integer that it's the the right 434 | 正确的整数 435 | 436 | 88 437 | 00:03:43,439 --> 00:03:46,090 438 | one then ties are going to be broken 439 | 一个然后领带将被打破 440 | 441 | 89 442 | 00:03:46,289 --> 00:03:47,890 443 | uniformly at random so we'll just flip a 444 | 均匀地随机分布，所以我们只需要翻转一个 445 | 446 | 90 447 | 00:03:48,090 --> 00:03:49,710 448 | coin or if there's three people will 449 | 硬币，或者如果有三个人 450 | 451 | 91 452 | 00:03:49,909 --> 00:03:54,520 453 | roll a dice a three-sided die etc okay 454 | 掷骰子，三面骰子等 455 | 456 | 92 457 | 00:03:54,719 --> 00:03:57,789 458 | so how would you play this game you have 459 | 所以你会怎么玩这个游戏 460 | 461 | 93 462 | 00:03:57,989 --> 00:03:58,930 463 | to think about what other players are 464 | 考虑其他玩家 465 | 466 | 94 467 | 00:03:59,129 --> 00:04:01,600 468 | going to do and then forecast what you 469 | 去做，然后预测你 470 | 471 | 95 472 | 00:04:01,800 --> 00:04:04,120 473 | think the best integer is in a response 474 | 认为最好的整数在响应中 475 | 476 | 96 477 | 00:04:04,319 --> 00:04:09,319 478 | to that 479 | 那个 480 | 481 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 1/GTO-1-07 - Best Response and Nash Equilibrium.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,000 --> 00:00:03,099 3 | I will now speak about how we predict 4 | 我现在将谈论我们如何预测 5 | 6 | 2 7 | 00:00:03,299 --> 00:00:05,139 8 | that people will play in a game and 9 | 人们会玩游戏， 10 | 11 | 3 12 | 00:00:05,339 --> 00:00:07,780 13 | specifically about the concepts of best 14 | 特别是关于最佳概念 15 | 16 | 4 17 | 00:00:07,980 --> 00:00:12,100 18 | response and Nash equilibrium now let's 19 | 反应和纳什均衡现在让我们 20 | 21 | 5 22 | 00:00:12,300 --> 00:00:14,319 23 | go home were to assume that you are one 24 | 回家假设你是一个 25 | 26 | 6 27 | 00:00:14,519 --> 00:00:17,530 28 | of the players and the other players are 29 | 的球员和其他球员是 30 | 31 | 7 32 | 00:00:17,730 --> 00:00:19,109 33 | going to play in a way that you know 34 | 会以您知道的方式演奏 35 | 36 | 8 37 | 00:00:19,309 --> 00:00:22,239 38 | then we can speak about what is your 39 | 那我们可以说说你是什么 40 | 41 | 9 42 | 00:00:22,439 --> 00:00:24,039 43 | best thing to do and we'll call that the 44 | 最好的事情，我们称之为 45 | 46 | 10 47 | 00:00:24,239 --> 00:00:27,550 48 | best response and technically speaking 49 | 最佳回应，从技术上讲 50 | 51 | 11 52 | 00:00:27,750 --> 00:00:31,149 53 | let's use the following notation let's 54 | 让我们使用以下表示法 55 | 56 | 12 57 | 00:00:31,349 --> 00:00:36,099 58 | call AI the action sequence of everybody 59 | 把AI称为每个人的行动顺序 60 | 61 | 13 62 | 00:00:36,299 --> 00:00:39,878 63 | except you player I so it's player 1 and 64 | 除了你我是玩家1 65 | 66 | 14 67 | 00:00:40,079 --> 00:00:42,909 68 | so on internal player I minus 1 I plus 1 69 | 所以内部玩家我减去1我加1 70 | 71 | 15 72 | 00:00:43,109 --> 00:00:46,619 73 | and all the rest so the entire action 74 | 其余的一切，整个动作 75 | 76 | 16 77 | 00:00:46,820 --> 00:00:49,649 78 | vector we call it the action profile a 79 | 向量，我们将其称为行动概况 80 | 81 | 17 82 | 00:00:49,850 --> 00:00:53,320 83 | is really made up of all those a minus I 84 | 真的是由所有减去我的人组成 85 | 86 | 18 87 | 00:00:53,520 --> 00:00:58,329 88 | and your own action with that notation 89 | 以及您使用该符号的操作 90 | 91 | 19 92 | 00:00:58,530 --> 00:01:01,149 93 | here's the definition we will say that 94 | 这是我们要说的定义 95 | 96 | 20 97 | 00:01:01,350 --> 00:01:03,849 98 | your best response which we'll call AI 99 | 您最好的回应，我们称之为AI 100 | 101 | 21 102 | 00:01:04,049 --> 00:01:07,689 103 | star it may not be unique but it's got 104 | 明星可能不是唯一的但有 105 | 106 | 22 107 | 00:01:07,890 --> 00:01:10,599 108 | to be part of the set it will call BR 109 | 作为集合的一部分，它将调用BR 110 | 111 | 23 112 | 00:01:10,799 --> 00:01:13,269 113 | the best response assuming that all the 114 | 假设所有 115 | 116 | 24 117 | 00:01:13,469 --> 00:01:17,709 118 | other agents play is a a - I if it's the 119 | 其他特工打的是aa-如果是 120 | 121 | 25 122 | 00:01:17,909 --> 00:01:20,890 123 | case that for any anything you might be 124 | 不管你做什么， 125 | 126 | 26 127 | 00:01:21,090 --> 00:01:24,250 128 | thinking of doing a sub I for all a sub 129 | 想为所有子项目做一个子项目 130 | 131 | 27 132 | 00:01:24,450 --> 00:01:27,399 133 | I it's got to be the case that your 134 | 我一定要这样 135 | 136 | 28 137 | 00:01:27,599 --> 00:01:31,028 138 | utility for playing that in other words 139 | 换句话说，实用程序 140 | 141 | 29 142 | 00:01:31,228 --> 00:01:34,899 143 | when you playing AI star and the others 144 | 当你玩AI之星和其他人 145 | 146 | 30 147 | 00:01:35,099 --> 00:01:38,500 148 | are playing AI a - I for whatever it is 149 | 在玩AI a-我不管 150 | 151 | 31 152 | 00:01:38,700 --> 00:01:40,778 153 | has got to be at least as great as 154 | 必须至少与 155 | 156 | 32 157 | 00:01:40,978 --> 00:01:44,278 158 | anything else you might choose a sub I 159 | 还有其他你可以选择的子 160 | 161 | 33 162 | 00:01:44,478 --> 00:01:47,439 163 | that's the case we'll say that a AI star 164 | 就是这样，我们会说一个AI明星 165 | 166 | 34 167 | 00:01:47,640 --> 00:01:51,069 168 | is a best response very intuitive notion 169 | 最好的回应是非常直观的概念 170 | 171 | 35 172 | 00:01:51,269 --> 00:01:55,119 173 | and that would bring us all almost all 174 | 那将带给我们几乎所有人 175 | 176 | 36 177 | 00:01:55,319 --> 00:01:57,369 178 | the way there the problem of course is 179 | 那里的问题当然是 180 | 181 | 37 182 | 00:01:57,569 --> 00:01:59,319 183 | that you don't know what the others will 184 | 你不知道别人会怎么做 185 | 186 | 38 187 | 00:01:59,519 --> 00:02:02,378 188 | do but that's ok because when we use the 189 | 可以，但是没关系，因为当我们使用 190 | 191 | 39 192 | 00:02:02,578 --> 00:02:04,778 193 | notion of best response as a building 194 | 最佳响应的概念 195 | 196 | 40 197 | 00:02:04,978 --> 00:02:07,028 198 | block to define what we call the Nash 199 | 定义我们称为Nash的块 200 | 201 | 41 202 | 00:02:07,228 --> 00:02:09,850 203 | equilibrium and the National Caribbean 204 | 均衡与加勒比国家联盟 205 | 206 | 42 207 | 00:02:10,050 --> 00:02:12,520 208 | briefly is a set of acts 209 | 简单来说是一套行为 210 | 211 | 43 212 | 00:02:12,719 --> 00:02:14,950 213 | one for each of the agents such that 214 | 每个代理一个 215 | 216 | 44 217 | 00:02:15,150 --> 00:02:17,730 218 | each is the best response to the others 219 | 每个都是对其他人的最佳回应 220 | 221 | 45 222 | 00:02:17,930 --> 00:02:20,469 223 | specifically we'll look at an action 224 | 具体来说，我们来看看一个动作 225 | 226 | 46 227 | 00:02:20,669 --> 00:02:24,850 228 | profile here a a one through a n and 229 | 在这里通过一个和一个配置文件 230 | 231 | 47 232 | 00:02:25,050 --> 00:02:26,710 233 | we'll say that it's a Nash equilibrium 234 | 我们会说那是纳什均衡 235 | 236 | 48 237 | 00:02:26,909 --> 00:02:29,560 238 | and later on we'll tell you why we call 239 | 稍后我们会告诉您为什么我们打电话 240 | 241 | 49 242 | 00:02:29,759 --> 00:02:32,140 243 | it specifically a pure strategy Nash 244 | 它专门是一种纯粹的策略纳什 245 | 246 | 50 247 | 00:02:32,340 --> 00:02:35,020 248 | equilibrium if it's the case that for 249 | 如果是这样的话 250 | 251 | 51 252 | 00:02:35,219 --> 00:02:38,650 253 | every agent that action a sub I 254 | 每个行动一个子我 255 | 256 | 52 257 | 00:02:38,849 --> 00:02:40,120 258 | associate with agent is your best 259 | 与代理商合作是您最好的选择 260 | 261 | 53 262 | 00:02:40,319 --> 00:02:45,319 263 | response all the rest 264 | 其余所有回应 265 | 266 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 1/GTO-1-08 - Examples.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,149 --> 00:00:03,188 3 | let us now look at some examples and of 4 | 现在让我们来看一些例子 5 | 6 | 2 7 | 00:00:03,388 --> 00:00:07,609 8 | games and Nash equilibria in those games 9 | 游戏和那些游戏中的纳什均衡 10 | 11 | 3 12 | 00:00:08,000 --> 00:00:10,390 13 | so here's the first game a familiar game 14 | 所以这是第一个熟悉的游戏 15 | 16 | 4 17 | 00:00:10,589 --> 00:00:14,189 18 | this is of course the prisoner's dilemma 19 | 这当然是囚犯的困境 20 | 21 | 5 22 | 00:00:14,388 --> 00:00:18,190 23 | the if both prisoners cooperate and 24 | 如果两个囚犯合作， 25 | 26 | 6 27 | 00:00:18,390 --> 00:00:20,859 28 | content and then they get a light 29 | 内容，然后他们就亮了 30 | 31 | 7 32 | 00:00:21,059 --> 00:00:25,568 33 | punishment and if they do not cooperate 34 | 惩罚，如果他们不配合 35 | 36 | 8 37 | 00:00:25,768 --> 00:00:28,839 38 | they get a more severe punishment if the 39 | 他们会受到更严厉的惩罚 40 | 41 | 9 42 | 00:00:29,039 --> 00:00:30,550 43 | one cooperates and the others does not 44 | 一个合作而其他不合作 45 | 46 | 10 47 | 00:00:30,750 --> 00:00:33,309 48 | then the co-op irrigated terrible 49 | 然后合作社灌溉很糟糕 50 | 51 | 11 52 | 00:00:33,509 --> 00:00:36,608 53 | punishment and the one that does not 54 | 惩罚与不惩罚 55 | 56 | 12 57 | 00:00:36,808 --> 00:00:38,500 58 | cooperate gets off scot it gets off 59 | 合作下车苏格兰人下车 60 | 61 | 13 62 | 00:00:38,700 --> 00:00:41,378 63 | scot-free and of course this game has a 64 | 无苏格兰人，当然这个游戏有 65 | 66 | 14 67 | 00:00:41,579 --> 00:00:45,038 68 | dominant strategy to defect no matter 69 | 主导缺陷策略无论如何 70 | 71 | 15 72 | 00:00:45,238 --> 00:00:47,649 73 | what the other agent does you're better 74 | 另一个代理商你做得更好 75 | 76 | 16 77 | 00:00:47,850 --> 00:00:52,570 78 | off not cooperating and so of course the 79 | 不合作，所以当然 80 | 81 | 17 82 | 00:00:52,770 --> 00:00:56,018 83 | only dominant strategy outcome is this 84 | 唯一的主要策略结果是 85 | 86 | 18 87 | 00:00:56,219 --> 00:00:58,689 88 | one of both defecting and indeed that is 89 | 既有缺陷，又有缺陷的是 90 | 91 | 19 92 | 00:00:58,890 --> 00:01:00,788 93 | the only Nash equilibrium in this game 94 | 这场比赛中唯一的纳什均衡 95 | 96 | 20 97 | 00:01:00,988 --> 00:01:03,729 98 | so it's the Nash equilibrium it's the 99 | 所以这是纳什均衡 100 | 101 | 21 102 | 00:01:03,929 --> 00:01:05,560 103 | best response if the other person 104 | 如果对方最佳反应 105 | 106 | 22 107 | 00:01:05,760 --> 00:01:07,659 108 | defects then it's the best respond to 109 | 缺陷，那么它是最好的应对方法 110 | 111 | 23 112 | 00:01:07,859 --> 00:01:09,969 113 | defect but in fact it's much stronger 114 | 缺陷，但实际上要强大得多 115 | 116 | 24 117 | 00:01:10,170 --> 00:01:11,709 118 | than that it's best to defect no matter 119 | 最好不要有任何缺点 120 | 121 | 25 122 | 00:01:11,909 --> 00:01:14,890 123 | what the other the other agent does so 124 | 另一个代理人这样做 125 | 126 | 26 127 | 00:01:15,090 --> 00:01:17,439 128 | this is an example of one unique Nash 129 | 这是一个独特的纳什的例子 130 | 131 | 27 132 | 00:01:17,640 --> 00:01:19,090 133 | equilibrium that happened to be very 134 | 恰好是非常均衡 135 | 136 | 28 137 | 00:01:19,290 --> 00:01:21,278 138 | strong one a dominant strategy Nash 139 | 强势策略Nash 140 | 141 | 29 142 | 00:01:21,478 --> 00:01:25,659 143 | equilibrium so so here's another game 144 | 平衡，所以这是另一个游戏 145 | 146 | 30 147 | 00:01:25,859 --> 00:01:27,549 148 | this is the game of pure coordination I 149 | 这是纯粹协调的游戏 150 | 151 | 31 152 | 00:01:27,750 --> 00:01:31,299 153 | think of it as walking towards each 154 | 认为它是走向每个 155 | 156 | 32 157 | 00:01:31,500 --> 00:01:33,789 158 | other on the sidewalk and you both can 159 | 其他人行道上，你们俩都可以 160 | 161 | 33 162 | 00:01:33,989 --> 00:01:35,950 163 | decide whether to go to your respective 164 | 决定是否去你各自的 165 | 166 | 34 167 | 00:01:36,150 --> 00:01:40,769 168 | lifts or respective rights in both cases 169 | 两种情况下的升降机或各自的权利 170 | 171 | 35 172 | 00:01:40,969 --> 00:01:43,659 173 | you will do fine and you will not 174 | 你会做得很好，你不会 175 | 176 | 36 177 | 00:01:43,859 --> 00:01:45,549 178 | collide and of course if you miss 179 | 碰撞，当然，如果您想念 180 | 181 | 37 182 | 00:01:45,750 --> 00:01:47,619 183 | coordinate if you one goes to the left 184 | 协调，如果你去左边 185 | 186 | 38 187 | 00:01:47,819 --> 00:01:48,609 188 | and the other to the right you will 189 | 另一个在右边 190 | 191 | 39 192 | 00:01:48,810 --> 00:01:52,628 193 | collide so this is a natural game and in 194 | 碰撞，所以这是自然的游戏 195 | 196 | 40 197 | 00:01:52,828 --> 00:01:55,359 198 | fact you see that you have two Nash 199 | 事实上你看到你有两个纳什 200 | 201 | 41 202 | 00:01:55,560 --> 00:01:58,028 203 | equilibria the one that I wrote down 204 | 平衡我写下的那一个 205 | 206 | 42 207 | 00:01:58,228 --> 00:02:01,869 208 | here if one one of the players goes to 209 | 如果一位玩家去 210 | 211 | 43 212 | 00:02:02,069 --> 00:02:03,730 213 | the left it's the best respond to go to 214 | 左边是最好的回应 215 | 216 | 44 217 | 00:02:03,930 --> 00:02:06,609 218 | the left and conversely if the other 219 | 左边，反之亦然 220 | 221 | 45 222 | 00:02:06,810 --> 00:02:08,469 223 | player goes to the right you're best off 224 | 玩家向右走，你最好 225 | 226 | 46 227 | 00:02:08,669 --> 00:02:10,838 228 | going to the right as well and the 229 | 也会向右走， 230 | 231 | 47 232 | 00:02:11,038 --> 00:02:12,550 233 | others are not Nash equilibria so here's 234 | 其他人不是纳什均衡，所以这里的 235 | 236 | 48 237 | 00:02:12,750 --> 00:02:13,090 238 | an exam 239 | 考试 240 | 241 | 49 242 | 00:02:13,289 --> 00:02:17,649 243 | of a game where there are two Nash 244 | 一个游戏中有两个纳什 245 | 246 | 50 247 | 00:02:17,848 --> 00:02:20,560 248 | equilibria or to specifically pure 249 | 均衡或特纯 250 | 251 | 51 252 | 00:02:20,759 --> 00:02:23,200 253 | strategy Nash equilibria again we'll 254 | 策略纳什均衡再一次 255 | 256 | 52 257 | 00:02:23,400 --> 00:02:25,929 258 | discuss why we call these pure strategy 259 | 讨论为什么我们称这些为纯策略 260 | 261 | 53 262 | 00:02:26,128 --> 00:02:31,110 263 | later on here's a very different game 264 | 稍后这是一个非常不同的游戏 265 | 266 | 54 267 | 00:02:31,310 --> 00:02:33,879 268 | this is often called the game of battle 269 | 这通常被称为战斗游戏 270 | 271 | 55 272 | 00:02:34,079 --> 00:02:37,480 273 | of the sexes imagine a couple and they 274 | 的性别想象一对夫妇，他们 275 | 276 | 56 277 | 00:02:37,680 --> 00:02:39,039 278 | want to go together to a movie and 279 | 想一起去看电影， 280 | 281 | 57 282 | 00:02:39,239 --> 00:02:41,379 283 | they're considering two movies one of 284 | 他们正在考虑两部电影之一 285 | 286 | 58 287 | 00:02:41,579 --> 00:02:45,399 288 | them a very violent movie Battle of the 289 | 他们是一部非常暴力的电影 290 | 291 | 59 292 | 00:02:45,598 --> 00:02:49,569 293 | Titans and the other very relaxed movie 294 | 泰坦和另一部非常轻松的电影 295 | 296 | 60 297 | 00:02:49,769 --> 00:02:53,640 298 | about flower growing called Lee's B&F 299 | 关于花的成长，叫做李的B＆F 300 | 301 | 61 302 | 00:02:53,840 --> 00:02:56,409 303 | the wife of course would prefer to go to 304 | 妻子当然更愿意去 305 | 306 | 62 307 | 00:02:56,609 --> 00:02:58,899 308 | battle of the Titans and the the husband 309 | 泰坦与丈夫之战 310 | 311 | 63 312 | 00:02:59,098 --> 00:03:01,599 313 | would prefer to watch flower growing but 314 | 宁愿看花开，但 315 | 316 | 64 317 | 00:03:01,799 --> 00:03:03,219 318 | more than anything else they would want 319 | 比他们想要的更重要 320 | 321 | 65 322 | 00:03:03,419 --> 00:03:05,080 323 | to go together and so here are the 324 | 一起去，所以这里是 325 | 326 | 66 327 | 00:03:05,280 --> 00:03:09,520 328 | payoffs if they both go to battle of the 329 | 如果他们都参加战斗的话 330 | 331 | 67 332 | 00:03:09,719 --> 00:03:12,849 333 | Titans then they're both positively 334 | 泰坦然后他们都是积极的 335 | 336 | 68 337 | 00:03:13,049 --> 00:03:15,580 338 | happy their wife more than the husband 339 | 妻子比丈夫更快乐 340 | 341 | 69 342 | 00:03:15,780 --> 00:03:18,670 343 | if the go both go to the flower growing 344 | 如果两者都去花开 345 | 346 | 70 347 | 00:03:18,870 --> 00:03:22,599 348 | movie then the husband is slightly happy 349 | 电影，然后丈夫有点高兴 350 | 351 | 71 352 | 00:03:22,799 --> 00:03:24,340 353 | and the wife but if they go to different 354 | 和妻子，但如果他们去不同的地方 355 | 356 | 72 357 | 00:03:24,539 --> 00:03:26,409 358 | movies neither of them is happy that's 359 | 电影他们俩都不开心 360 | 361 | 73 362 | 00:03:26,609 --> 00:03:30,340 363 | that's that's the that's that's the the 364 | 那就是那那就是那 365 | 366 | 74 367 | 00:03:30,539 --> 00:03:36,999 368 | game how many how many equilibria we 369 | 游戏我们多少平衡 370 | 371 | 75 372 | 00:03:37,199 --> 00:03:40,629 373 | have here well again we have two pure 374 | 再来一遍，我们有两个纯 375 | 376 | 76 377 | 00:03:40,829 --> 00:03:43,390 378 | strategy Nash equilibria why is that 379 | 策略纳什均衡为什么 380 | 381 | 77 382 | 00:03:43,590 --> 00:03:49,030 383 | well if either them goes to the Battle 384 | 好吧，如果他们去参加战斗 385 | 386 | 78 387 | 00:03:49,229 --> 00:03:50,439 388 | of the Titans then the other one would 389 | 的泰坦，然后另一个 390 | 391 | 79 392 | 00:03:50,639 --> 00:03:52,659 393 | want to go there too because if they go 394 | 也想去那里，因为如果他们去 395 | 396 | 80 397 | 00:03:52,859 --> 00:03:54,969 398 | to a different one they would get zero 399 | 换成另一种，他们将得到零 400 | 401 | 81 402 | 00:03:55,169 --> 00:03:57,969 403 | rather than whatever they get here one 404 | 而不是他们来到这里 405 | 406 | 82 407 | 00:03:58,169 --> 00:03:59,499 408 | or two depending on whether the husband 409 | 一两个取决于丈夫 410 | 411 | 83 412 | 00:03:59,699 --> 00:04:02,140 413 | of the wife and conversely on the on the 414 | 的妻子，相反在 415 | 416 | 84 417 | 00:04:02,340 --> 00:04:04,300 418 | flower watching movie flower grow movie 419 | 花看电影花种电影 420 | 421 | 85 422 | 00:04:04,500 --> 00:04:07,090 423 | and so in both cases they were the best 424 | 所以在两种情况下它们都是最好的 425 | 426 | 86 427 | 00:04:07,289 --> 00:04:09,999 428 | response is to go to the movie selected 429 | 回应是去看电影 430 | 431 | 87 432 | 00:04:10,199 --> 00:04:13,360 433 | by the other party so on the face of it 434 | 对方如此面对 435 | 436 | 88 437 | 00:04:13,560 --> 00:04:15,159 438 | it looks very similar to the game of 439 | 它看起来非常类似于 440 | 441 | 89 442 | 00:04:15,359 --> 00:04:17,499 443 | pure coordination that we have here but 444 | 我们在这里有纯粹的协调，但是 445 | 446 | 90 447 | 00:04:17,699 --> 00:04:19,180 448 | we do see a slight difference here and 449 | 我们在这里确实看到了细微的差别， 450 | 451 | 91 452 | 00:04:19,379 --> 00:04:21,310 453 | it will revisit that later on when we 454 | 稍后，当我们 455 | 456 | 92 457 | 00:04:21,509 --> 00:04:23,860 458 | speak about not pure strategies but 459 | 谈论的不是纯粹的策略，而是 460 | 461 | 93 462 | 00:04:24,060 --> 00:04:25,670 463 | mixed strategies 464 | 混合策略 465 | 466 | 94 467 | 00:04:25,870 --> 00:04:28,259 468 | here's a here's another example the last 469 | 这是一个这是另一个例子 470 | 471 | 95 472 | 00:04:28,459 --> 00:04:30,090 473 | one we'll look at and this is the game 474 | 我们来看一看，这就是游戏 475 | 476 | 96 477 | 00:04:30,290 --> 00:04:32,939 478 | called matching pennies imagine each of 479 | 所谓的匹配便士想象每个 480 | 481 | 97 482 | 00:04:33,139 --> 00:04:35,218 483 | us two players needing to just need to 484 | 我们两个玩家只需要 485 | 486 | 98 487 | 00:04:35,418 --> 00:04:38,430 488 | decide on some side of a of a coin heads 489 | 在硬币头的a的某一侧决定 490 | 491 | 99 492 | 00:04:38,629 --> 00:04:41,879 493 | or tail if we decide on the same size 494 | 还是尾巴，如果我们决定相同的尺寸 495 | 496 | 100 497 | 00:04:42,079 --> 00:04:44,730 498 | heads or tail but we decide on the same 499 | 头或尾，但我们决定相同 500 | 501 | 101 502 | 00:04:44,930 --> 00:04:48,960 503 | one then then I win if we decide on 504 | 一个然后我赢了，如果我们决定 505 | 506 | 102 507 | 00:04:49,160 --> 00:04:51,090 508 | different sides you heads and me tails 509 | 你的头和我的尾巴的不同侧面 510 | 511 | 103 512 | 00:04:51,290 --> 00:04:54,449 513 | or vice versa the new win and so we see 514 | 反之亦然，新的胜利，所以我们看到 515 | 516 | 104 517 | 00:04:54,649 --> 00:04:59,999 518 | this here if we both decides on heads or 519 | 如果我们俩都决定要这样做，还是在这里 520 | 521 | 105 522 | 00:05:00,199 --> 00:05:01,949 523 | we both decides on tails I win and 524 | 我们都决定我赢的尾巴， 525 | 526 | 106 527 | 00:05:02,149 --> 00:05:03,900 528 | otherwise you win by winning I mean I 529 | 不然你赢了就是我 530 | 531 | 107 532 | 00:05:04,100 --> 00:05:05,759 533 | get 1 you get minus 1 so this is a 534 | 得到1你得到负1所以这是一个 535 | 536 | 108 537 | 00:05:05,959 --> 00:05:10,338 538 | zero-sum game the sum of our payoff is 0 539 | 零和游戏，我们的收益之和为0 540 | 541 | 109 542 | 00:05:10,538 --> 00:05:13,439 543 | what is a pure strategy Nash equilibrium 544 | 什么是纯策略纳什均衡 545 | 546 | 110 547 | 00:05:13,639 --> 00:05:16,710 548 | here let's think about it suppose I pick 549 | 让我们考虑一下，假设我选择了 550 | 551 | 111 552 | 00:05:16,910 --> 00:05:20,400 553 | head what is your best response well 554 | 领导什么是你最好的反应 555 | 556 | 112 557 | 00:05:20,600 --> 00:05:22,740 558 | your best response then is to pick tails 559 | 你最好的选择就是挑尾巴 560 | 561 | 113 562 | 00:05:22,939 --> 00:05:26,100 563 | because you get one rather than minus 564 | 因为你得到一个而不是减 565 | 566 | 114 567 | 00:05:26,300 --> 00:05:31,350 568 | one but if you pick I tails then my best 569 | 一个，但是如果你选择我尾巴，那我最好 570 | 571 | 115 572 | 00:05:31,550 --> 00:05:33,540 573 | response is not to play tail because I 574 | 回应不是打尾巴，因为我 575 | 576 | 116 577 | 00:05:33,740 --> 00:05:36,000 578 | want to coordinate with you because then 579 | 想和你协调，因为 580 | 581 | 117 582 | 00:05:36,199 --> 00:05:39,660 583 | I will get one rather than minus one 584 | 我会得到一个而不是减一 585 | 586 | 118 587 | 00:05:39,860 --> 00:05:42,360 588 | that I would be getting here but now if 589 | 我会到这里，但现在如果 590 | 591 | 119 592 | 00:05:42,560 --> 00:05:47,009 593 | I play tails you'd rather play heads 594 | 我打尾巴你宁可打头 595 | 596 | 120 597 | 00:05:47,209 --> 00:05:50,189 598 | because you'd get one rather than the 599 | 因为你会得到一个而不是 600 | 601 | 121 602 | 00:05:50,389 --> 00:05:53,009 603 | minus one you're getting here but again 604 | 减一，你要来这里，但是再来一次 605 | 606 | 122 607 | 00:05:53,209 --> 00:05:55,528 608 | if you're playing tails I want to if you 609 | 如果你在打尾巴，我想 610 | 611 | 123 612 | 00:05:55,728 --> 00:05:57,300 613 | playing heads I want to play heads to 614 | 我要打头 615 | 616 | 124 617 | 00:05:57,500 --> 00:06:01,528 618 | match so we have this cycle where the 619 | 匹配，所以我们有这个周期 620 | 621 | 125 622 | 00:06:01,728 --> 00:06:07,230 623 | best responses are leading us in the 624 | 最好的回应正在引领我们 625 | 626 | 126 627 | 00:06:07,430 --> 00:06:10,800 628 | cycle and so there is no pure strategy 629 | 周期，所以没有纯粹的策略 630 | 631 | 127 632 | 00:06:11,000 --> 00:06:12,629 633 | Nash equilibrium in this game of mashing 634 | 这场捣蛋游戏中的纳什均衡 635 | 636 | 128 637 | 00:06:12,829 --> 00:06:17,829 638 | pennies 639 | 便士 640 | 641 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 2/GTO-2-01 - Mixed Strategies and Nash Equilibrium - A High-Level Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,000 --> 00:00:02,139 3 | this video is going to introduce the 4 | 该视频将介绍 5 | 6 | 2 7 | 00:00:02,339 --> 00:00:04,359 8 | idea of mixed strategies and explain why 9 | 混合策略的想法并解释原因 10 | 11 | 3 12 | 00:00:04,559 --> 00:00:08,770 13 | they're important to Nash equilibria so 14 | 它们对纳什均衡很重要，所以 15 | 16 | 4 17 | 00:00:08,970 --> 00:00:10,180 18 | the example that I want to think about 19 | 我想想的例子 20 | 21 | 5 22 | 00:00:10,380 --> 00:00:14,500 23 | here is the United Nations setting up 24 | 这是联合国的设置 25 | 26 | 6 27 | 00:00:14,699 --> 00:00:17,890 28 | checkpoints to defend against terrorist 29 | 防御恐怖分子的检查站 30 | 31 | 7 32 | 00:00:18,089 --> 00:00:21,940 33 | attacks at a port in Somalia and you can 34 | 袭击索马里的一个港口，你可以 35 | 36 | 8 37 | 00:00:22,140 --> 00:00:26,649 38 | see that what happens is they they place 39 | 看看发生什么事了 40 | 41 | 9 42 | 00:00:26,849 --> 00:00:28,749 43 | a checkpoint on the road they stop all 44 | 他们在路上阻止所有的检查站 45 | 46 | 10 47 | 00:00:28,949 --> 00:00:31,749 48 | cars like this taxi here and they go 49 | 像这样的出租车在这里的车，他们去 50 | 51 | 11 52 | 00:00:31,949 --> 00:00:33,219 53 | through the contents of the car to make 54 | 通过汽车的内容来制作 55 | 56 | 12 57 | 00:00:33,420 --> 00:00:34,538 58 | sure that it doesn't contain any 59 | 确保它不包含任何 60 | 61 | 13 62 | 00:00:34,738 --> 00:00:36,928 63 | explosives or other dangerous materials 64 | 炸药或其他危险材料 65 | 66 | 14 67 | 00:00:37,128 --> 00:00:39,820 68 | now let's think about this situation as 69 | 现在让我们考虑这种情况 70 | 71 | 15 72 | 00:00:40,020 --> 00:00:42,820 73 | a game so there are a variety of 74 | 一个游戏，所以有各种各样的 75 | 76 | 16 77 | 00:00:43,020 --> 00:00:45,579 78 | different roads that the UN could choose 79 | 联合国可以选择的不同道路 80 | 81 | 17 82 | 00:00:45,780 --> 00:00:48,189 83 | to set up its checkpoint on every hour 84 | 每小时设置其检查点 85 | 86 | 18 87 | 00:00:48,390 --> 00:00:54,549 88 | and for each one of those roads the the 89 | 对于其中的每条道路， 90 | 91 | 19 92 | 00:00:54,750 --> 00:00:58,588 93 | potential attacker could decide to 94 | 潜在的攻击者可以决定 95 | 96 | 20 97 | 00:00:58,789 --> 00:01:01,928 98 | attack that road and if it's the case 99 | 攻击那条路，如果是这样 100 | 101 | 21 102 | 00:01:02,128 --> 00:01:05,379 103 | that the defender defended that road and 104 | 捍卫者捍卫了这条路， 105 | 106 | 22 107 | 00:01:05,579 --> 00:01:07,329 108 | the attacker attacks it then the 109 | 攻击者攻击它，然后 110 | 111 | 23 112 | 00:01:07,530 --> 00:01:09,579 113 | attacker gets a large negative payoff 114 | 攻击者获得了巨大的负收益 115 | 116 | 24 117 | 00:01:09,780 --> 00:01:11,379 118 | because they're captured and their 119 | 因为他们被捕获了 120 | 121 | 25 122 | 00:01:11,579 --> 00:01:14,230 123 | attack is not successful if on the other 124 | 如果对方攻击不成功 125 | 126 | 26 127 | 00:01:14,430 --> 00:01:16,918 128 | hand the attacker attacks any other road 129 | 攻击者攻击其他道路 130 | 131 | 27 132 | 00:01:17,118 --> 00:01:20,409 133 | then the attack is successful and then 134 | 然后攻击成功，然后 135 | 136 | 28 137 | 00:01:20,609 --> 00:01:22,599 138 | the utility of the attacker depends on 139 | 攻击者的效用取决于 140 | 141 | 29 142 | 00:01:22,799 --> 00:01:24,849 143 | the value of the target that was 144 | 目标的价值是 145 | 146 | 30 147 | 00:01:25,049 --> 00:01:28,750 148 | attacked and wasn't defended now it's 149 | 被攻击，现在没有被捍卫 150 | 151 | 31 152 | 00:01:28,950 --> 00:01:31,869 153 | pretty clear that if the UN were to 154 | 很清楚，如果联合国要 155 | 156 | 32 157 | 00:01:32,069 --> 00:01:34,418 158 | commit to any deterministic strategy if 159 | 致力于任何确定性策略，如果 160 | 161 | 33 162 | 00:01:34,618 --> 00:01:36,399 163 | they were to choose their action in any 164 | 他们将在任何情况下选择自己的行动 165 | 166 | 34 167 | 00:01:36,599 --> 00:01:38,980 168 | deterministic way things are going to go 169 | 确定性的方式 170 | 171 | 35 172 | 00:01:39,180 --> 00:01:41,649 173 | pretty badly for them because the 174 | 对他们来说非常糟糕，因为 175 | 176 | 36 177 | 00:01:41,849 --> 00:01:43,808 178 | defender asserting the attacker would be 179 | 声称攻击者是 180 | 181 | 37 182 | 00:01:44,009 --> 00:01:45,308 183 | able to look at what they're doing then 184 | 然后可以看看他们在做什么 185 | 186 | 38 187 | 00:01:45,509 --> 00:01:47,259 188 | they'd be able to watch for a while see 189 | 他们可以看一会儿 190 | 191 | 39 192 | 00:01:47,459 --> 00:01:49,090 193 | what strategy the UN is following and 194 | 联合国正在采取什么策略？ 195 | 196 | 40 197 | 00:01:49,290 --> 00:01:50,890 198 | then attack something different and 199 | 然后攻击不同的东西 200 | 201 | 41 202 | 00:01:51,090 --> 00:01:53,558 203 | attacks would always be successful so it 204 | 攻击将永远是成功的，所以它 205 | 206 | 42 207 | 00:01:53,759 --> 00:01:55,238 208 | must be that this is not really how 209 | 必须是这不是真的 210 | 211 | 43 212 | 00:01:55,438 --> 00:01:58,119 213 | checkpoints get set up and indeed what 214 | 建立检查点，实际上是什么 215 | 216 | 44 217 | 00:01:58,319 --> 00:02:00,399 218 | really happens is that the checkpoints 219 | 真正发生的是检查站 220 | 221 | 45 222 | 00:02:00,599 --> 00:02:02,619 223 | are set up in a randomized way so that 224 | 以随机方式设置，以便 225 | 226 | 46 227 | 00:02:02,819 --> 00:02:04,778 228 | even if the attacker is watched for a 229 | 即使攻击者受到监视 230 | 231 | 47 232 | 00:02:04,978 --> 00:02:07,179 233 | while and figure out what the randomized 234 | 一会儿，弄清楚什么是随机的 235 | 236 | 48 237 | 00:02:07,379 --> 00:02:10,240 238 | strategy is they're not able to know on 239 | 策略是他们不知道 240 | 241 | 49 242 | 00:02:10,439 --> 00:02:12,368 243 | a given hour where the checkpoints are 244 | 给定小时内检查点所在的位置 245 | 246 | 50 247 | 00:02:12,568 --> 00:02:13,320 248 | going to be 249 | 将 250 | 251 | 51 252 | 00:02:13,520 --> 00:02:16,080 253 | and this means that their value of an 254 | 这意味着他们的价值 255 | 256 | 52 257 | 00:02:16,280 --> 00:02:18,990 258 | attack is limited so the Nash 259 | 攻击是有限的，所以纳什 260 | 261 | 53 262 | 00:02:19,189 --> 00:02:21,120 263 | equilibrium of a game like this is going 264 | 像这样的游戏平衡 265 | 266 | 54 267 | 00:02:21,319 --> 00:02:23,790 268 | to involve the defender defending in a 269 | 让辩护人参与 270 | 271 | 55 272 | 00:02:23,990 --> 00:02:26,040 273 | randomized way like this and this kind 274 | 像这样的随机方式 275 | 276 | 56 277 | 00:02:26,240 --> 00:02:28,110 278 | of a randomized strategy is called a 279 | 的随机策略称为 280 | 281 | 57 282 | 00:02:28,310 --> 00:02:30,510 283 | mixed strategy and that's going to be 284 | 混合策略，这将是 285 | 286 | 58 287 | 00:02:30,710 --> 00:02:35,710 288 | the topic of the next sequence of videos 289 | 下一段视频的主题 290 | 291 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 2/GTO-2-04a - Hardness of Computing a Nash Equilibrium (Basic).srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,060 --> 00:00:02,468 3 | so let's speak a little bit about how 4 | 所以让我们谈谈如何 5 | 6 | 2 7 | 00:00:02,669 --> 00:00:05,259 8 | hard is it to compute a Nash equilibrium 9 | 很难计算纳什均衡 10 | 11 | 3 12 | 00:00:05,459 --> 00:00:09,220 13 | in a normal phone game this is an 14 | 在正常的电话游戏中，这是 15 | 16 | 4 17 | 00:00:09,419 --> 00:00:11,769 18 | involved topic and I'll just give you a 19 | 涉及的话题，我给你一个 20 | 21 | 5 22 | 00:00:11,968 --> 00:00:14,500 23 | taste for it let me draw your attention 24 | 尝起来让我引起你的注意 25 | 26 | 6 27 | 00:00:14,699 --> 00:00:17,050 28 | to two specific algorithms for computing 29 | 两种特定的计算算法 30 | 31 | 7 32 | 00:00:17,250 --> 00:00:20,249 33 | a sample natural equilibrium and again 34 | 样本自然平衡 35 | 36 | 8 37 | 00:00:20,449 --> 00:00:23,230 38 | these are two out of a long line of 39 | 这是一长串的两个 40 | 41 | 9 42 | 00:00:23,429 --> 00:00:25,120 43 | algorithms that have been investigated 44 | 已经研究过的算法 45 | 46 | 10 47 | 00:00:25,320 --> 00:00:28,449 48 | and these are a sort of two extreme the 49 | 这是两个极端 50 | 51 | 11 52 | 00:00:28,649 --> 00:00:30,579 53 | one of them starts with a mathematical 54 | 其中之一以数学开始 55 | 56 | 12 57 | 00:00:30,778 --> 00:00:32,198 58 | formulation of the problem called a 59 | 问题的表达称为 60 | 61 | 13 62 | 00:00:32,399 --> 00:00:36,489 63 | linear complementarity problem and you 64 | 线性互补问题和你 65 | 66 | 14 67 | 00:00:36,689 --> 00:00:40,000 68 | want to set it up as a mathematical 69 | 想要将其设置为数学 70 | 71 | 15 72 | 00:00:40,200 --> 00:00:42,459 73 | optimization problem you can apply 74 | 您可以应用的优化问题 75 | 76 | 16 77 | 00:00:42,659 --> 00:00:45,250 78 | various algorithms to that and the most 79 | 最重要的各种算法 80 | 81 | 17 82 | 00:00:45,450 --> 00:00:50,739 83 | famous one for two players games is due 84 | 著名的两人游戏之一 85 | 86 | 18 87 | 00:00:50,939 --> 00:00:54,549 88 | to Lemke and houses and this is an 89 | 到伦克和房屋，这是一个 90 | 91 | 19 92 | 00:00:54,750 --> 00:01:01,239 93 | algorithm that really displays a deep 94 | 真正展现出深刻的算法 95 | 96 | 20 97 | 00:01:01,439 --> 00:01:02,829 98 | understanding of the mathematical 99 | 对数学的理解 100 | 101 | 21 102 | 00:01:03,030 --> 00:01:07,209 103 | structure of what a game is and the 104 | 游戏的结构以及 105 | 106 | 22 107 | 00:01:07,409 --> 00:01:10,390 108 | nature of Nash equilibria on perhaps the 109 | 纳什均衡的性质也许 110 | 111 | 23 112 | 00:01:10,590 --> 00:01:14,349 113 | other extreme is what is called the 114 | 另一个极端是所谓的 115 | 116 | 24 117 | 00:01:14,549 --> 00:01:18,759 118 | support enumeration method a recent 119 | 支持枚举方法最近 120 | 121 | 25 122 | 00:01:18,959 --> 00:01:21,969 123 | procedure that doesn't have as deep and 124 | 没有那么深和 125 | 126 | 26 127 | 00:01:22,170 --> 00:01:23,439 128 | insight into the structure of the 129 | 洞察结构 130 | 131 | 27 132 | 00:01:23,640 --> 00:01:25,929 133 | problem it says simply the following it 134 | 问题，它说只需跟随它 135 | 136 | 28 137 | 00:01:26,129 --> 00:01:30,959 138 | says if you fix the support of the 139 | 说如果你修复的支持 140 | 141 | 29 142 | 00:01:31,159 --> 00:01:34,119 143 | strategies of the player and the support 144 | 玩家的策略和支持 145 | 146 | 30 147 | 00:01:34,319 --> 00:01:35,579 148 | of the strategy players are those 149 | 战略参与者中的那些是 150 | 151 | 31 152 | 00:01:35,780 --> 00:01:37,659 153 | actions are played with nonzero 154 | 动作以非零播放 155 | 156 | 32 157 | 00:01:37,859 --> 00:01:41,049 158 | probability if you fix that support then 159 | 如果您修复该支持的可能性 160 | 161 | 33 162 | 00:01:41,250 --> 00:01:42,819 163 | the problem becomes very easy you can 164 | 问题变得非常容易，您可以 165 | 166 | 34 167 | 00:01:43,019 --> 00:01:44,980 168 | set it up as a linear program and solve 169 | 将其设置为线性程序并求解 170 | 171 | 35 172 | 00:01:45,180 --> 00:01:48,939 173 | it efficiently and that will be the end 174 | 它有效，这将是结局 175 | 176 | 36 177 | 00:01:49,140 --> 00:01:51,579 178 | of it if it weren't for the case that 179 | 如果不是这样的话 180 | 181 | 37 182 | 00:01:51,780 --> 00:01:54,399 183 | there that indeed there are an 184 | 那里确实有一个 185 | 186 | 38 187 | 00:01:54,599 --> 00:01:56,679 188 | exponential number of supports to 189 | 支持的指数数量 190 | 191 | 39 192 | 00:01:56,879 --> 00:02:00,759 193 | explore and so the trick in this 194 | 探索，所以这个技巧 195 | 196 | 40 197 | 00:02:00,959 --> 00:02:04,329 198 | procedure is to explore them cleverly 199 | 程序是巧妙地探索它们 200 | 201 | 41 202 | 00:02:04,530 --> 00:02:07,209 203 | using Clarrie eristic s-- and that's 204 | 使用Clarrie eristic s-- 205 | 206 | 42 207 | 00:02:07,409 --> 00:02:09,130 208 | called the support enumeration there's a 209 | 称为支持枚举 210 | 211 | 43 212 | 00:02:09,330 --> 00:02:11,360 213 | clever heuristic work for 214 | 巧妙的启发式工作 215 | 216 | 44 217 | 00:02:11,560 --> 00:02:13,969 218 | how to enumerate those supports and 219 | 如何列举这些支持，以及 220 | 221 | 45 222 | 00:02:14,169 --> 00:02:16,670 223 | check them one by one although the 224 | 尽管他们一一检查 225 | 226 | 46 227 | 00:02:16,870 --> 00:02:21,410 228 | latter procedure is not as smart or as 229 | 后面的过程不那么聪明 230 | 231 | 47 232 | 00:02:21,610 --> 00:02:25,730 233 | insightful as the Lemke housen I turns 234 | 当我转身成为Lemke屋子时很有见地 235 | 236 | 48 237 | 00:02:25,930 --> 00:02:28,730 238 | out that in practice it tends to run 239 | 在实践中它倾向于运行 240 | 241 | 49 242 | 00:02:28,930 --> 00:02:33,680 243 | very fast so we've seen the algorithm 244 | 非常快，所以我们已经看过算法 245 | 246 | 50 247 | 00:02:33,879 --> 00:02:35,300 248 | people have tried very hard to find 249 | 人们努力寻找 250 | 251 | 51 252 | 00:02:35,500 --> 00:02:38,060 253 | algorithms computing a sample Nash 254 | 计算样本Nash的算法 255 | 256 | 52 257 | 00:02:38,259 --> 00:02:39,950 258 | equilibrium 259 | 平衡 260 | 261 | 53 262 | 00:02:40,150 --> 00:02:42,289 263 | and it does seem hard the question is 264 | 问题似乎确实很难 265 | 266 | 54 267 | 00:02:42,489 --> 00:02:45,590 268 | can we somehow capture that formally 269 | 我们可以以某种方式正式捕获它吗 270 | 271 | 55 272 | 00:02:45,789 --> 00:02:53,050 273 | within the complexity hierarchy and and 274 | 在复杂性层次结构中，以及 275 | 276 | 56 277 | 00:02:53,250 --> 00:02:55,130 278 | and for that we need to introduce some 279 | 为此，我们需要介绍一些 280 | 281 | 57 282 | 00:02:55,330 --> 00:02:59,330 283 | you know new new concepts the essential 284 | 你知道新概念必不可少 285 | 286 | 58 287 | 00:02:59,530 --> 00:03:02,539 288 | concept is that of the new class of 289 | 概念是新类别的 290 | 291 | 59 292 | 00:03:02,739 --> 00:03:06,649 293 | problems called P P ad for polynomial 294 | 多项式的PP ad问题 295 | 296 | 60 297 | 00:03:06,848 --> 00:03:09,050 298 | parity arguments directed graph 299 | 奇偶参数有向图 300 | 301 | 61 302 | 00:03:09,250 --> 00:03:12,050 303 | introduced by christos papadimitriou in 304 | 由克里斯托斯·帕帕迪米特里乌（Christos papadimitriou）在 305 | 306 | 62 307 | 00:03:12,250 --> 00:03:15,770 308 | 1994 we won't go into detail but just so 309 | 1994年，我们不再赘述，只是如此 310 | 311 | 63 312 | 00:03:15,969 --> 00:03:18,700 313 | you know the chronology P P ad is a 314 | 您知道时间顺序PP广告是 315 | 316 | 64 317 | 00:03:18,900 --> 00:03:21,620 318 | specialization of a class called TF NP 319 | TF NP类的专业化 320 | 321 | 65 322 | 00:03:21,819 --> 00:03:23,750 323 | which is in turn with a specialization 324 | 反过来又有专长 325 | 326 | 66 327 | 00:03:23,949 --> 00:03:27,349 328 | of a problem called f NP a going detail 329 | f NP问题的细节 330 | 331 | 67 332 | 00:03:27,549 --> 00:03:31,460 333 | here is is beyond the scope of what we 334 | 这超出了我们的范围 335 | 336 | 68 337 | 00:03:31,659 --> 00:03:34,810 338 | want to speak about but it does help us 339 | 想谈一谈，但这确实对我们有帮助 340 | 341 | 69 342 | 00:03:35,009 --> 00:03:40,430 343 | now position the complexity of finding a 344 | 现在定位找到一个 345 | 346 | 70 347 | 00:03:40,629 --> 00:03:41,629 348 | sample Nash equilibrium in the 349 | 样本中的纳什均衡 350 | 351 | 71 352 | 00:03:41,829 --> 00:03:44,170 353 | complexity hierarchy and again we have 354 | 复杂性层次结构 355 | 356 | 72 357 | 00:03:44,370 --> 00:03:50,200 358 | the class of polynomial time problems of 359 | 多项式时间问题的类别 360 | 361 | 73 362 | 00:03:50,400 --> 00:03:52,368 363 | problem that can be verified in 364 | 可以验证的问题 365 | 366 | 74 367 | 00:03:52,568 --> 00:03:54,830 368 | polynomial time with these being the 369 | 多项式时间，这些是 370 | 371 | 75 372 | 00:03:55,030 --> 00:03:58,939 373 | hardest among them and given that P PAE 374 | 其中最难的是P PAE 375 | 376 | 76 377 | 00:03:59,139 --> 00:04:03,409 378 | turns out to reside somewhere within 379 | 原来是住在里面的某个地方 380 | 381 | 77 382 | 00:04:03,609 --> 00:04:05,409 383 | this class now again we don't know 384 | 再次上这堂课，我们不知道 385 | 386 | 78 387 | 00:04:05,609 --> 00:04:08,840 388 | whether this entire class did not 389 | 整个班级都没有 390 | 391 | 79 392 | 00:04:09,039 --> 00:04:11,110 393 | collapse and all become one of the same 394 | 崩溃，都成为同一个 395 | 396 | 80 397 | 00:04:11,310 --> 00:04:13,460 398 | it's why do you believe that it does not 399 | 这就是为什么您认为它不会 400 | 401 | 81 402 | 00:04:13,659 --> 00:04:17,930 403 | but proof doesn't exist however we do 404 | 但是我们不存在证明 405 | 406 | 82 407 | 00:04:18,129 --> 00:04:20,658 408 | know that P P 80 lies someplace in 409 | 知道PP 80位于 410 | 411 | 83 412 | 00:04:20,858 --> 00:04:23,730 413 | between P and 414 | 在P和 415 | 416 | 84 417 | 00:04:23,930 --> 00:04:27,189 418 | now what does that have to do with the 419 | 现在这与 420 | 421 | 85 422 | 00:04:27,389 --> 00:04:29,639 423 | problem computing a Nash equilibrium 424 | 计算纳什均衡的问题 425 | 426 | 86 427 | 00:04:29,839 --> 00:04:32,439 428 | well that's where the the following 429 | 嗯，这是以下内容 430 | 431 | 87 432 | 00:04:32,639 --> 00:04:36,429 433 | theorems come in originally it was shown 434 | 定理最初来自它表明 435 | 436 | 88 437 | 00:04:36,629 --> 00:04:39,218 438 | that the problem of computing a Nash 439 | 计算纳什的问题 440 | 441 | 89 442 | 00:04:39,418 --> 00:04:41,350 443 | equilibrium is complete for this class 444 | 该课程的均衡是完全的 445 | 446 | 90 447 | 00:04:41,550 --> 00:04:42,129 448 | pped 449 | pped 450 | 451 | 91 452 | 00:04:42,329 --> 00:04:45,040 453 | that is it's the hardest among all 454 | 那是最困难的 455 | 456 | 92 457 | 00:04:45,240 --> 00:04:47,679 458 | problem that class initially proved for 459 | 班级最初证明的问题 460 | 461 | 93 462 | 00:04:47,879 --> 00:04:51,459 463 | four players then for all four games 464 | 然后四场比赛的四名球员 465 | 466 | 94 467 | 00:04:51,658 --> 00:04:53,980 468 | with three or more players and then 469 | 三个或更多玩家，然后 470 | 471 | 95 472 | 00:04:54,180 --> 00:04:58,329 473 | finally in oh six for all class of games 474 | 最终在所有游戏中都达到了六分 475 | 476 | 96 477 | 00:04:58,528 --> 00:05:01,989 478 | and so we widely believed that the 479 | 因此，我们广泛认为 480 | 481 | 97 482 | 00:05:02,189 --> 00:05:05,468 483 | problem is not polynomial cannot prove 484 | 问题不是多项式无法证明 485 | 486 | 98 487 | 00:05:05,668 --> 00:05:07,899 488 | it but we do know where it resides and 489 | 它，但我们确实知道它的位置， 490 | 491 | 99 492 | 00:05:08,098 --> 00:05:09,910 493 | within the complexity hierarchy that we 494 | 在我们的复杂性层次结构中 495 | 496 | 100 497 | 00:05:10,110 --> 00:05:15,110 498 | are familiar with 499 | 熟悉 500 | 501 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 3/GTO-3-01 - Other solution concepts - A High-Level Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,060 --> 00:00:02,228 3 | in this video we're going to look at 4 | 在这个视频中，我们将要看 5 | 6 | 2 7 | 00:00:02,428 --> 00:00:04,929 8 | some additional solution concepts other 9 | 其他一些解决方案概念 10 | 11 | 3 12 | 00:00:05,129 --> 00:00:07,149 13 | than the Nash equilibrium so these are 14 | 比纳什均衡，所以这些是 15 | 16 | 4 17 | 00:00:07,349 --> 00:00:10,120 18 | different ways of talking about which 19 | 不同的谈论方式 20 | 21 | 5 22 | 00:00:10,320 --> 00:00:12,189 23 | outcomes of a game make sense from a 24 | 游戏的结果从 25 | 26 | 6 27 | 00:00:12,388 --> 00:00:16,539 28 | game theoretic perspective first of all 29 | 博弈论的观点 30 | 31 | 7 32 | 00:00:16,739 --> 00:00:18,400 33 | I want to talk about a solution concept 34 | 我想谈一谈解决方案的概念 35 | 36 | 8 37 | 00:00:18,600 --> 00:00:20,739 38 | called iterated removal of dominated 39 | 称为反复删除主导 40 | 41 | 9 42 | 00:00:20,939 --> 00:00:24,249 43 | strategies and I want to illustrate this 44 | 策略，我想说明一下 45 | 46 | 10 47 | 00:00:24,449 --> 00:00:27,190 48 | by the example of grace shown in this 49 | 以这个例子中的恩典为例 50 | 51 | 11 52 | 00:00:27,390 --> 00:00:30,280 53 | picture here who decided to jump out of 54 | 决定跳出图片的人 55 | 56 | 12 57 | 00:00:30,480 --> 00:00:33,239 58 | a plane to celebrate her 91st birthday 59 | 一架庆祝她91岁生日的飞机 60 | 61 | 13 62 | 00:00:33,439 --> 00:00:36,189 63 | so I want to think about a game between 64 | 所以我想考虑一下 65 | 66 | 14 67 | 00:00:36,390 --> 00:00:39,159 68 | Grace and the guy that she chose to 69 | 格蕾丝和她选择的那个人 70 | 71 | 15 72 | 00:00:39,359 --> 00:00:41,858 73 | strap herself to who you can also see in 74 | 把自己绑在你也可以看到的人身上 75 | 76 | 16 77 | 00:00:42,058 --> 00:00:45,669 78 | the picture and in particular I want to 79 | 图片，尤其是我想 80 | 81 | 17 82 | 00:00:45,869 --> 00:00:47,829 83 | think about his decision of whether to 84 | 考虑他的决定是否 85 | 86 | 18 87 | 00:00:48,030 --> 00:00:51,338 88 | pack the parachute safely or not and her 89 | 降落伞是否安全打包，她 90 | 91 | 19 92 | 00:00:51,539 --> 00:00:54,909 93 | decision about whether to jump out of 94 | 关于是否跳出的决定 95 | 96 | 20 97 | 00:00:55,109 --> 00:00:58,628 98 | the plane or not now in principle she 99 | 原则上她现在还是不在飞机上 100 | 101 | 21 102 | 00:00:58,829 --> 00:01:01,268 103 | might worry that he would choose not to 104 | 可能会担心他会选择不 105 | 106 | 22 107 | 00:01:01,469 --> 00:01:03,489 108 | pack the parachute safely and she would 109 | 安全地打包降落伞，她会 110 | 111 | 23 112 | 00:01:03,689 --> 00:01:05,560 113 | choose to jump out of the plane and if 114 | 选择跳出飞机，如果 115 | 116 | 24 117 | 00:01:05,760 --> 00:01:07,659 118 | that were to happen then she would never 119 | 那将要发生，那么她将永远不会 120 | 121 | 25 122 | 00:01:07,859 --> 00:01:10,750 123 | get to celebrate her 92nd birthday but 124 | 庆祝她的92岁生日， 125 | 126 | 26 127 | 00:01:10,950 --> 00:01:13,980 128 | you can see in fact she did choose and 129 | 你可以看到事实上她确实选择了 130 | 131 | 27 132 | 00:01:14,180 --> 00:01:16,750 133 | indeed she landed safely and her choice 134 | 的确，她安全着陆并且选择 135 | 136 | 28 137 | 00:01:16,950 --> 00:01:17,558 138 | was a good one 139 | 是一个好人 140 | 141 | 29 142 | 00:01:17,759 --> 00:01:19,539 143 | so how was she able to reason that this 144 | 所以她怎么能推理出 145 | 146 | 30 147 | 00:01:19,739 --> 00:01:22,000 148 | was sensible well if she looked at the 149 | 如果她看着 150 | 151 | 31 152 | 00:01:22,200 --> 00:01:23,859 153 | payoffs of the game she would see that 154 | 游戏的收益，她会看到 155 | 156 | 32 157 | 00:01:24,060 --> 00:01:27,269 158 | that this guy let's call him Bruce 159 | 这个家伙叫他布鲁斯 160 | 161 | 33 162 | 00:01:27,469 --> 00:01:30,250 163 | Bruce's action of not packing the 164 | 布鲁斯不包装行李箱的行为 165 | 166 | 34 167 | 00:01:30,450 --> 00:01:32,950 168 | parachute safely was very bad not only 169 | 安全降落伞不仅非常糟糕 170 | 171 | 35 172 | 00:01:33,150 --> 00:01:35,649 173 | for grace but also for himself in fact 174 | 为了恩典，也为了自己 175 | 176 | 36 177 | 00:01:35,849 --> 00:01:39,549 178 | it was a dominated strategy and knowing 179 | 这是一个主导的策略，并且知道 180 | 181 | 37 182 | 00:01:39,750 --> 00:01:42,789 183 | that he's rational grace reasoned that 184 | 他的理性宽限理由 185 | 186 | 38 187 | 00:01:42,989 --> 00:01:44,948 188 | he would never play a dominated strategy 189 | 他永远不会扮演主导策略 190 | 191 | 39 192 | 00:01:45,149 --> 00:01:47,709 193 | and so she was able to change the game 194 | 这样她就能改变比赛 195 | 196 | 40 197 | 00:01:47,909 --> 00:01:49,959 198 | by removing this dominated strategy and 199 | 通过删除这一主导策略， 200 | 201 | 41 202 | 00:01:50,159 --> 00:01:53,230 203 | instead to reason that she only had to 204 | 而是说她只需要 205 | 206 | 42 207 | 00:01:53,430 --> 00:01:55,329 208 | care about the remainder of the game in 209 | 在乎游戏的其余部分 210 | 211 | 43 212 | 00:01:55,530 --> 00:01:57,189 213 | which his dominated strategies didn't 214 | 他的主导策略没有 215 | 216 | 44 217 | 00:01:57,390 --> 00:01:59,980 218 | exist this is the idea of iterated 219 | 存在，这是迭代的想法 220 | 221 | 45 222 | 00:02:00,180 --> 00:02:02,140 223 | removal of dominated strategies which 224 | 删除主导策略 225 | 226 | 46 227 | 00:02:02,340 --> 00:02:05,308 228 | you'll hear about more formally later 229 | 您稍后会听到更多正式消息 230 | 231 | 47 232 | 00:02:05,509 --> 00:02:07,808 233 | secondly I'd like to revisit our 234 | 其次，我想回顾一下我们的 235 | 236 | 48 237 | 00:02:08,008 --> 00:02:11,289 238 | question of soccer goal kicking and I'd 239 | 踢足球的问题，我会 240 | 241 | 49 242 | 00:02:11,489 --> 00:02:13,390 243 | like to ask is it really the 244 | 想问的是真的吗 245 | 246 | 50 247 | 00:02:13,590 --> 00:02:17,350 248 | is that when a player prepares to take a 249 | 是当玩家准备接受 250 | 251 | 51 252 | 00:02:17,550 --> 00:02:19,990 253 | penalty kick he's really solving for the 254 | 罚球，他真的解决了 255 | 256 | 52 257 | 00:02:20,189 --> 00:02:21,880 258 | Nash equilibrium now we did see 259 | 纳什均衡现在我们确实看到了 260 | 261 | 53 262 | 00:02:22,080 --> 00:02:23,950 263 | experimental evidence that shows that 264 | 实验证据表明 265 | 266 | 54 267 | 00:02:24,150 --> 00:02:25,929 268 | the Nash equilibrium is a pretty good 269 | 纳什均衡是相当不错的 270 | 271 | 55 272 | 00:02:26,128 --> 00:02:28,270 273 | description of what actually happens in 274 | 实际发生的情况的描述 275 | 276 | 56 277 | 00:02:28,469 --> 00:02:30,399 278 | these situations but is it the case that 279 | 这些情况，但事实是 280 | 281 | 57 282 | 00:02:30,598 --> 00:02:32,439 283 | the players are really thinking about 284 | 玩家们真的在考虑 285 | 286 | 58 287 | 00:02:32,639 --> 00:02:34,630 288 | the idea of Nash equilibrium that 289 | 纳什均衡的思想 290 | 291 | 59 292 | 00:02:34,830 --> 00:02:36,849 293 | doesn't seem right it seems like the 294 | 似乎不正确，好像 295 | 296 | 60 297 | 00:02:37,049 --> 00:02:38,890 298 | players are thinking about how best to 299 | 玩家正在考虑如何最好地 300 | 301 | 61 302 | 00:02:39,090 --> 00:02:41,890 303 | kick the ball into the goal in order to 304 | 将球踢入球门以 305 | 306 | 62 307 | 00:02:42,090 --> 00:02:43,899 308 | hurt the other guy as much as possible 309 | 尽可能伤害另一个人 310 | 311 | 63 312 | 00:02:44,098 --> 00:02:47,770 313 | or in order to do as well for themselves 314 | 或者为了自己做得好 315 | 316 | 64 317 | 00:02:47,969 --> 00:02:50,950 318 | as possible it turns out that this isn't 319 | 事实证明这不是 320 | 321 | 65 322 | 00:02:51,150 --> 00:02:53,500 323 | an accident in the case of zero-sum 324 | 零和的情况下发生事故 325 | 326 | 66 327 | 00:02:53,699 --> 00:02:56,770 328 | games these three ideas doing as well 329 | 游戏这三个想法也做得很好 330 | 331 | 67 332 | 00:02:56,969 --> 00:02:59,170 333 | for yourself as possible hurting the 334 | 为自己尽可能的伤害 335 | 336 | 68 337 | 00:02:59,370 --> 00:03:00,879 338 | other player as much as possible and 339 | 尽可能多的其他玩家 340 | 341 | 69 342 | 00:03:01,079 --> 00:03:04,060 343 | being in Nash equilibrium all turn out 344 | 处于纳什均衡状态 345 | 346 | 70 347 | 00:03:04,259 --> 00:03:08,860 348 | to coincide finally I want to revisit 349 | 恰好最后我想重温 350 | 351 | 71 352 | 00:03:09,060 --> 00:03:12,368 353 | the battle of the sexes and ask is it 354 | 性别之战，问是 355 | 356 | 72 357 | 00:03:12,568 --> 00:03:14,710 358 | really the case that as we saw before 359 | 确实如我们之前所见 360 | 361 | 73 362 | 00:03:14,909 --> 00:03:16,480 363 | with the Nash equilibria of this game 364 | 这场比赛的纳什均衡 365 | 366 | 74 367 | 00:03:16,680 --> 00:03:19,509 368 | were doomed either to an unfair outcome 369 | 注定要么得到不公平的结果 370 | 371 | 75 372 | 00:03:19,709 --> 00:03:21,219 373 | where one member of the couple always 374 | 一对夫妻中总是有一个 375 | 376 | 76 377 | 00:03:21,419 --> 00:03:23,860 378 | gets their preferred activity or a miss 379 | 得到他们喜欢的活动或错过 380 | 381 | 77 382 | 00:03:24,060 --> 00:03:27,219 383 | coordination where sometimes the the two 384 | 有时两者在哪里协调 385 | 386 | 78 387 | 00:03:27,419 --> 00:03:28,599 388 | members of the couple end up doing 389 | 这对夫妇的成员最终做 390 | 391 | 79 392 | 00:03:28,799 --> 00:03:31,330 393 | different activities it doesn't seem 394 | 似乎没有不同的活动 395 | 396 | 80 397 | 00:03:31,530 --> 00:03:33,219 398 | like this is a good model of how people 399 | 像这样是人们如何做的好榜样 400 | 401 | 81 402 | 00:03:33,419 --> 00:03:35,530 403 | really do solve disputes like this 404 | 确实确实解决了这样的纠纷 405 | 406 | 82 407 | 00:03:35,729 --> 00:03:37,750 408 | between themselves so I want to think 409 | 彼此之间，所以我想 410 | 411 | 83 412 | 00:03:37,949 --> 00:03:39,490 413 | about a new solution concept called 414 | 关于一个称为“ 415 | 416 | 84 417 | 00:03:39,689 --> 00:03:42,789 418 | correlated equilibrium in which we don't 419 | 相关均衡，我们不这样做 420 | 421 | 85 422 | 00:03:42,989 --> 00:03:44,530 423 | have this problem and were able to 424 | 有这个问题，并能够 425 | 426 | 86 427 | 00:03:44,729 --> 00:03:46,599 428 | achieve fairness without miss 429 | 公平无误 430 | 431 | 87 432 | 00:03:46,799 --> 00:03:51,799 433 | coordination 434 | 协调 435 | 436 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 4/GTO-4-01 - Perfect Information Extensive Form - Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,500 --> 00:00:03,159 3 | sometimes time plays an important role 4 | 有时时间很重要 5 | 6 | 2 7 | 00:00:03,359 --> 00:00:08,350 8 | in in in strategic situations things 9 | 在战略形势下 10 | 11 | 3 12 | 00:00:08,550 --> 00:00:12,939 13 | take step one after the other and not 14 | 一步一步走，而不是 15 | 16 | 4 17 | 00:00:13,138 --> 00:00:15,609 18 | only do they do but the actors know that 19 | 他们只是这样做，但是演员们知道 20 | 21 | 5 22 | 00:00:15,808 --> 00:00:17,560 23 | they will and that influences how they 24 | 他们会并且那会影响他们 25 | 26 | 6 27 | 00:00:17,760 --> 00:00:23,109 28 | behave so the year was 1519 and amnon 29 | 举止如此，这一年是1519和阿农 30 | 31 | 7 32 | 00:00:23,309 --> 00:00:26,560 33 | Cortez Spaniard was leading a flotilla 34 | 西班牙科尔特斯（Cortez Spaniard）带领一支舰队 35 | 36 | 8 37 | 00:00:26,760 --> 00:00:30,999 38 | of eleven boats and about 600 men about 39 | 共有11艘船，约600名男子 40 | 41 | 9 42 | 00:00:31,199 --> 00:00:34,989 43 | to invade a continent later to be known 44 | 后来入侵一个大陆 45 | 46 | 10 47 | 00:00:35,189 --> 00:00:37,779 48 | as America they were vastly outnumbered 49 | 作为美国，它们的数量远远超过了 50 | 51 | 11 52 | 00:00:37,979 --> 00:00:41,858 53 | and were well aware of the the heavy art 54 | 并且非常了解繁重的艺术 55 | 56 | 12 57 | 00:00:42,058 --> 00:00:46,239 58 | they were facing and as is famously 59 | 他们面对着，众所周知 60 | 61 | 13 62 | 00:00:46,439 --> 00:00:52,959 63 | known as they landed Cortes ordered that 64 | 当他们降落时，科尔特斯下令 65 | 66 | 14 67 | 00:00:53,159 --> 00:00:55,209 68 | all the both be burnt this not 69 | 都被烧毁了这不是 70 | 71 | 15 72 | 00:00:55,409 --> 00:00:56,890 73 | controversy whether this was done in 74 | 争议是否在 75 | 76 | 16 77 | 00:00:57,090 --> 00:00:58,838 78 | complete coordination and agreement with 79 | 与之完全协调并达成协议 80 | 81 | 17 82 | 00:00:59,039 --> 00:01:01,149 83 | his men or it was sprung on them but 84 | 他的手下，或者突然冒出来，但是 85 | 86 | 18 87 | 00:01:01,350 --> 00:01:03,158 88 | either way it's clear what the logic 89 | 无论哪种方式，很清楚逻辑是什么 90 | 91 | 19 92 | 00:01:03,359 --> 00:01:08,500 93 | behind it was the quartet Cortes knew 94 | 后面是四重奏Cortes知道 95 | 96 | 20 97 | 00:01:08,700 --> 00:01:11,500 98 | that as they faced these daunting odds 99 | 当他们面对这些艰巨的可能性时 100 | 101 | 21 102 | 00:01:11,700 --> 00:01:15,569 103 | the men would be tempted to turn back 104 | 男人会很想回头 105 | 106 | 22 107 | 00:01:15,769 --> 00:01:18,939 108 | board the boat and and flee and by 109 | 登船逃离 110 | 111 | 23 112 | 00:01:19,140 --> 00:01:22,778 113 | removing eruption that increased their 114 | 消除增加他们的喷发 115 | 116 | 24 117 | 00:01:22,978 --> 00:01:27,719 118 | fighting resolve going forward and so 119 | 战斗决心前进等等 120 | 121 | 25 122 | 00:01:27,920 --> 00:01:31,390 123 | again there's not only the fact that 124 | 不仅如此， 125 | 126 | 26 127 | 00:01:31,590 --> 00:01:34,899 128 | time passes between actions but that one 129 | 时间在行动之间流逝，但是那一个 130 | 131 | 27 132 | 00:01:35,099 --> 00:01:38,140 133 | reasons about that fact and impacts how 134 | 有关该事实的原因并影响 135 | 136 | 28 137 | 00:01:38,340 --> 00:01:41,859 138 | the strategic situation unfolds we see 139 | 我们看到的战略形势正在发展 140 | 141 | 29 142 | 00:01:42,060 --> 00:01:42,969 143 | it not only when there are multiple 144 | 不仅有多个 145 | 146 | 30 147 | 00:01:43,170 --> 00:01:46,929 148 | agents involved but even in a single 149 | 代理商参与，但即使是一个 150 | 151 | 31 152 | 00:01:47,129 --> 00:01:49,480 153 | decision maker here it's not only Cortes 154 | 这里的决策者不仅是科尔特斯 155 | 156 | 32 157 | 00:01:49,680 --> 00:01:53,698 158 | and his men the two actors or the two 159 | 和他的手下两个演员或两个 160 | 161 | 33 162 | 00:01:53,899 --> 00:01:58,090 163 | sets of actors who are somehow actions 164 | 演员以某种方式行动 165 | 166 | 34 167 | 00:01:58,290 --> 00:02:01,869 168 | are intertwined but even with a single 169 | 交织在一起，即使只有一个 170 | 171 | 35 172 | 00:02:02,069 --> 00:02:04,808 173 | actor the fact that time unfolds can 174 | 演员事实证明时间可以展开 175 | 176 | 36 177 | 00:02:05,009 --> 00:02:09,279 178 | impact the situation here's another 179 | 影响情况，这是另一个 180 | 181 | 37 182 | 00:02:09,479 --> 00:02:11,800 183 | famous historical tale in this case of 184 | 在这种情况下的著名历史故事 185 | 186 | 38 187 | 00:02:12,000 --> 00:02:13,660 188 | Ulysses and the sirens 189 | 尤利西斯和警报器 190 | 191 | 39 192 | 00:02:13,860 --> 00:02:18,730 193 | Ulysses is captaining his boat and about 194 | 尤利西斯（Ulysses）正在指挥他的船， 195 | 196 | 40 197 | 00:02:18,930 --> 00:02:21,790 198 | to pass through the straits of the 199 | 穿越海峡 200 | 201 | 41 202 | 00:02:21,990 --> 00:02:27,610 203 | sirens and as is commonly known the 204 | 警报器，众所周知 205 | 206 | 42 207 | 00:02:27,810 --> 00:02:31,270 208 | sirens song are so seductive that they 209 | 警笛声很诱人 210 | 211 | 43 212 | 00:02:31,469 --> 00:02:35,020 213 | would cause any person in victory 214 | 会导致任何人胜利 215 | 216 | 44 217 | 00:02:35,219 --> 00:02:38,469 218 | Ulysses to do things there are in his 219 | 尤利西斯做他的事 220 | 221 | 45 222 | 00:02:38,669 --> 00:02:42,250 223 | own not not in his own best interests he 224 | 自己不符合自己的最大利益 225 | 226 | 46 227 | 00:02:42,449 --> 00:02:43,719 228 | would jump into the sea he would crash 229 | 会跳入海中，他会坠毁 230 | 231 | 47 232 | 00:02:43,919 --> 00:02:46,360 233 | the boat against the rocks and so what 234 | 船靠在岩石上等等 235 | 236 | 48 237 | 00:02:46,560 --> 00:02:49,900 238 | he does according to the tale he order 239 | 他按照他订购的故事来做 240 | 241 | 49 242 | 00:02:50,099 --> 00:02:52,300 243 | is always a man to first put wax in the 244 | 总是一个男人先在上面放蜡 245 | 246 | 50 247 | 00:02:52,500 --> 00:02:53,950 248 | ears so they will not be seduced by the 249 | 耳朵，所以他们不会被诱惑 250 | 251 | 51 252 | 00:02:54,150 --> 00:02:58,180 253 | song he himself who in fact wants to 254 | 自己想唱的他自己 255 | 256 | 52 257 | 00:02:58,379 --> 00:03:00,280 258 | hear the song knowing that he would not 259 | 听这首歌知道他不会 260 | 261 | 53 262 | 00:03:00,479 --> 00:03:03,300 263 | be able to withstand the seduction 264 | 能够承受诱惑 265 | 266 | 54 267 | 00:03:03,500 --> 00:03:06,039 268 | orders he's meant to time to the mast 269 | 他本来想去桅杆的命令 270 | 271 | 55 272 | 00:03:06,239 --> 00:03:10,060 273 | and should he possibly get free from the 274 | 他是否应该摆脱困境 275 | 276 | 56 277 | 00:03:10,259 --> 00:03:14,500 278 | from the tithe to restrain him with his 279 | 从什一税来约束他 280 | 281 | 57 282 | 00:03:14,699 --> 00:03:18,700 283 | swords and and and and so so it comes to 284 | 剑与和与之所以如此 285 | 286 | 58 287 | 00:03:18,900 --> 00:03:23,560 288 | pass they sailed through the Strait when 289 | 当他们驶过海峡时 290 | 291 | 59 292 | 00:03:23,759 --> 00:03:25,330 293 | they when he hears the siren song he 294 | 他们当他听到警笛声时 295 | 296 | 60 297 | 00:03:25,530 --> 00:03:27,689 298 | goes temporarily insane and tries to 299 | 暂时疯了，试图 300 | 301 | 61 302 | 00:03:27,889 --> 00:03:32,259 303 | escape from the bond that fails and all 304 | 逃离失败的纽带 305 | 306 | 62 307 | 00:03:32,459 --> 00:03:36,189 308 | is good so here again is single action 309 | 好，所以这里又是一个动作 310 | 311 | 63 312 | 00:03:36,389 --> 00:03:38,500 313 | in this case Ulysses reasoning about the 314 | 在这种情况下，尤利西斯（Ulysses）关于 315 | 316 | 64 317 | 00:03:38,699 --> 00:03:40,960 318 | future thinking about what will the 319 | 未来思考什么 320 | 321 | 65 322 | 00:03:41,159 --> 00:03:43,270 323 | situation be at taking action now to 324 | 情况正在采取行动 325 | 326 | 66 327 | 00:03:43,469 --> 00:03:46,090 328 | impact what the strategic situation will 329 | 影响战略形势 330 | 331 | 67 332 | 00:03:46,289 --> 00:03:50,140 333 | look like in the future to model such 334 | 将来像这样建模 335 | 336 | 68 337 | 00:03:50,340 --> 00:03:53,080 338 | situation we turn to games in extensive 339 | 情况我们转向广泛的游戏 340 | 341 | 69 342 | 00:03:53,280 --> 00:03:56,319 343 | form as the term is called sometimes 344 | 有时称为形式 345 | 346 | 70 347 | 00:03:56,519 --> 00:04:01,519 348 | simply known as game truth 349 | 简称为游戏真相 350 | 351 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 4/GTO-4-07 - Imperfect Information Extensive Form - Poker.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,030 --> 00:00:02,799 3 | hello again folks so this is Matt and 4 | 再次向大家问好，这是马特， 5 | 6 | 2 7 | 00:00:03,000 --> 00:00:05,439 8 | we're talking about imperfect 9 | 我们在谈论不完美 10 | 11 | 3 12 | 00:00:05,639 --> 00:00:07,509 13 | information in the extensive form now so 14 | 现在以广泛的形式提供信息 15 | 16 | 4 17 | 00:00:07,710 --> 00:00:09,930 18 | we're going to be talking now about 19 | 我们现在要谈论 20 | 21 | 5 22 | 00:00:10,130 --> 00:00:12,519 23 | games where we have some sequential 24 | 我们有一些顺序的游戏 25 | 26 | 6 27 | 00:00:12,718 --> 00:00:14,769 28 | moves and there can be some uncertainty 29 | 移动，可能会有一些不确定性 30 | 31 | 7 32 | 00:00:14,968 --> 00:00:19,780 33 | in players minds about both the possible 34 | 在玩家心目中，两种可能性 35 | 36 | 8 37 | 00:00:19,980 --> 00:00:22,659 38 | payoffs of others and the strategies 39 | 他人的收益和策略 40 | 41 | 9 42 | 00:00:22,859 --> 00:00:23,980 43 | that others might be following 44 | 其他人可能会关注 45 | 46 | 10 47 | 00:00:24,179 --> 00:00:27,129 48 | so we'll start those just by you know to 49 | 所以我们将以您知道的方式开始 50 | 51 | 11 52 | 00:00:27,329 --> 00:00:29,109 53 | give you some ideas about this let's 54 | 给你一些想法，让我们 55 | 56 | 12 57 | 00:00:29,309 --> 00:00:31,510 58 | talk a little bit about poker which is a 59 | 谈一点关于扑克的知识 60 | 61 | 13 62 | 00:00:31,710 --> 00:00:33,459 63 | game has been becoming incredibly 64 | 游戏变得越来越不可思议 65 | 66 | 14 67 | 00:00:33,659 --> 00:00:36,429 68 | popular and recently both for people 69 | 受欢迎，最近都受到人们的欢迎 70 | 71 | 15 72 | 00:00:36,630 --> 00:00:37,809 73 | playing and on television and other 74 | 玩和看电视等 75 | 76 | 16 77 | 00:00:38,009 --> 00:00:39,669 78 | kinds of things and it's one of the 79 | 各种各样的东西，这是其中之一 80 | 81 | 17 82 | 00:00:39,869 --> 00:00:48,659 83 | oldest games which has very extensive 84 | 最古老的游戏 85 | 86 | 18 87 | 00:00:48,859 --> 00:00:52,149 88 | experience and for a lot of people and 89 | 经验，对很多人来说， 90 | 91 | 19 92 | 00:00:52,350 --> 00:00:54,549 93 | the the you know one of the critical 94 | 你知道的关键之一 95 | 96 | 20 97 | 00:00:54,750 --> 00:00:56,049 98 | aspects of poker is that there's 99 | 扑克方面是 100 | 101 | 21 102 | 00:00:56,250 --> 00:00:59,259 103 | actually sequential play in betting 104 | 实际上是连续下注 105 | 106 | 22 107 | 00:00:59,460 --> 00:01:02,799 108 | calling folding so one player gets you 109 | 召唤折叠，让一个玩家抓住你 110 | 111 | 23 112 | 00:01:03,000 --> 00:01:05,799 113 | know two to make a decision in terms of 114 | 认识两个人就 115 | 116 | 24 117 | 00:01:06,000 --> 00:01:08,560 118 | how much they're going to make a bet at 119 | 他们要下多少赌注 120 | 121 | 25 122 | 00:01:08,760 --> 00:01:10,000 123 | a certain point in time other players 124 | 在某个时间点其他玩家 125 | 126 | 26 127 | 00:01:10,200 --> 00:01:11,319 128 | have to react to that 129 | 必须对此做出反应 130 | 131 | 27 132 | 00:01:11,519 --> 00:01:14,980 133 | so the sequential play you see some 134 | 所以顺序播放，您会看到一些 135 | 136 | 28 137 | 00:01:15,180 --> 00:01:17,140 138 | cards in in many of these games but not 139 | 这些游戏中有很多是纸牌，但不是 140 | 141 | 29 142 | 00:01:17,340 --> 00:01:18,730 143 | all so you might see some of the cards 144 | 所有这些，所以您可能会看到一些卡片 145 | 146 | 30 147 | 00:01:18,930 --> 00:01:19,929 148 | that the other players are holding but 149 | 其他玩家正在举行，但 150 | 151 | 31 152 | 00:01:20,129 --> 00:01:21,399 153 | you don't know how strong their hand is 154 | 你不知道他们的手有多强 155 | 156 | 32 157 | 00:01:21,599 --> 00:01:24,879 158 | and you have to be inferring things 159 | 而你必须在推断事物 160 | 161 | 33 162 | 00:01:25,079 --> 00:01:27,369 163 | about their their possible cards both 164 | 关于他们可能的牌 165 | 166 | 34 167 | 00:01:27,569 --> 00:01:32,109 168 | from odds in the game and based on what 169 | 从游戏中的赔率以及基于什么 170 | 171 | 35 172 | 00:01:32,310 --> 00:01:34,950 173 | they're doing in terms of their strategy 174 | 他们正在根据自己的策略进行 175 | 176 | 36 177 | 00:01:35,150 --> 00:01:37,299 178 | so you see the bets and you react to 179 | 所以你看到了赌注，你对 180 | 181 | 37 182 | 00:01:37,500 --> 00:01:38,619 183 | them and you have to make inferences 184 | 他们，你必须做出推断 185 | 186 | 38 187 | 00:01:38,819 --> 00:01:42,698 188 | based on that so that involves having 189 | 基于此，这涉及到 190 | 191 | 39 192 | 00:01:42,899 --> 00:01:44,439 193 | beliefs about the motivations the 194 | 关于动机的信念 195 | 196 | 40 197 | 00:01:44,640 --> 00:01:46,719 198 | rationality of other players what their 199 | 其他玩家的合理性 200 | 201 | 41 202 | 00:01:46,920 --> 00:01:48,668 203 | payoffs are what their potential payoffs 204 | 收益是他们潜在的收益 205 | 206 | 42 207 | 00:01:48,868 --> 00:01:51,099 208 | are which in poker might come from from 209 | 扑克中的哪些可能来自 210 | 211 | 43 212 | 00:01:51,299 --> 00:01:54,278 213 | the cards so when we think about these 214 | 卡，所以当我们考虑这些 215 | 216 | 44 217 | 00:01:54,478 --> 00:01:55,840 218 | kinds of games you know there's many 219 | 您知道很多类型的游戏 220 | 221 | 45 222 | 00:01:56,040 --> 00:01:57,698 223 | possible hands that's gonna make poker a 224 | 可能使扑克成为可能 225 | 226 | 46 227 | 00:01:57,899 --> 00:02:01,028 228 | fairly complicated game to keep track of 229 | 相当复杂的游戏，以跟踪 230 | 231 | 47 232 | 00:02:01,228 --> 00:02:03,009 233 | there's many betting strategies which 234 | 有很多下注策略 235 | 236 | 48 237 | 00:02:03,209 --> 00:02:06,340 238 | means that the overall tree that we have 239 | 意味着我们拥有的整体树 240 | 241 | 49 242 | 00:02:06,540 --> 00:02:07,750 243 | to work with is gonna be quite 244 | 与之共事会相当 245 | 246 | 50 247 | 00:02:07,950 --> 00:02:10,029 248 | complicated so it's actually going to be 249 | 复杂，所以实际上 250 | 251 | 51 252 | 00:02:10,229 --> 00:02:11,800 253 | almost impossible to draw the tree in 254 | 几乎不可能画树 255 | 256 | 52 257 | 00:02:12,000 --> 00:02:12,990 258 | the sense of just 259 | 正义感 260 | 261 | 53 262 | 00:02:13,189 --> 00:02:16,050 263 | on the screen but there's nonetheless a 264 | 在屏幕上，但仍然有 265 | 266 | 54 267 | 00:02:16,250 --> 00:02:17,850 268 | lot that we can learn about analyzing 269 | 我们可以从分析中学到很多 270 | 271 | 55 272 | 00:02:18,050 --> 00:02:20,550 273 | such games and analyzing the types of 274 | 这类游戏并分析 275 | 276 | 56 277 | 00:02:20,750 --> 00:02:22,920 278 | strategies that they have how how do we 279 | 他们有如何做的策略 280 | 281 | 57 282 | 00:02:23,120 --> 00:02:24,360 283 | represent extensive form games with 284 | 代表广泛的形式游戏 285 | 286 | 58 287 | 00:02:24,560 --> 00:02:25,980 288 | incomplete information how might we but 289 | 不完整的信息我们怎么可能 290 | 291 | 59 292 | 00:02:26,180 --> 00:02:27,780 293 | reason about these things 294 | 这些事情的原因 295 | 296 | 60 297 | 00:02:27,979 --> 00:02:29,219 298 | and moreover there'll be simpler 299 | 而且会更简单 300 | 301 | 61 302 | 00:02:29,419 --> 00:02:30,990 303 | settings you know poker is actually a 304 | 您知道扑克的设置实际上是 305 | 306 | 62 307 | 00:02:31,189 --> 00:02:33,930 308 | fairly complicated game and there's 309 | 相当复杂的游戏， 310 | 311 | 63 312 | 00:02:34,129 --> 00:02:35,520 313 | other fairly complicated games but very 314 | 其他相当复杂的游戏，但非常 315 | 316 | 64 317 | 00:02:35,719 --> 00:02:37,650 318 | high stakes games so for instance you 319 | 高风险游戏，例如您 320 | 321 | 65 322 | 00:02:37,849 --> 00:02:39,120 323 | know we could have one country thinking 324 | 知道我们可以有一个国家的思考 325 | 326 | 66 327 | 00:02:39,319 --> 00:02:41,370 328 | about invading another one a potential 329 | 关于入侵另一个潜在的 330 | 331 | 67 332 | 00:02:41,569 --> 00:02:45,060 333 | war or a conflict there trying to decide 334 | 战争或冲突试图决定 335 | 336 | 68 337 | 00:02:45,259 --> 00:02:47,100 338 | what the other country is going to do in 339 | 另一个国家在做什么 340 | 341 | 69 342 | 00:02:47,300 --> 00:02:49,920 343 | response so if you invaded what would 344 | 回应，因此，如果您入侵了 345 | 346 | 70 347 | 00:02:50,120 --> 00:02:51,240 348 | they do that's a game of incomplete 349 | 他们这样做是不完整的游戏 350 | 351 | 71 352 | 00:02:51,439 --> 00:02:52,590 353 | information because you might not know 354 | 信息，因为您可能不知道 355 | 356 | 72 357 | 00:02:52,789 --> 00:02:56,070 358 | exactly how strong they are or how was 359 | 他们到底有多坚强 360 | 361 | 73 362 | 00:02:56,270 --> 00:02:58,020 363 | the the willingness of the population to 364 | 人口的意愿 365 | 366 | 74 367 | 00:02:58,219 --> 00:03:01,469 368 | fight what might happen politically how 369 | 对抗政治上可能发生的事情 370 | 371 | 75 372 | 00:03:01,669 --> 00:03:03,390 373 | how strong are they and if there was a 374 | 他们有多坚强，如果有 375 | 376 | 76 377 | 00:03:03,590 --> 00:03:06,300 378 | war so there's these are situations 379 | 战争，所以有这些情况 380 | 381 | 77 382 | 00:03:06,500 --> 00:03:09,090 383 | where one party might have to move first 384 | 一方可能必须先走的地方 385 | 386 | 78 387 | 00:03:09,289 --> 00:03:12,090 388 | anticipating reaction of the other the 389 | 期待对方的反应 390 | 391 | 79 392 | 00:03:12,289 --> 00:03:14,219 393 | second one has to anticipate what the 394 | 第二个必须预料到 395 | 396 | 80 397 | 00:03:14,419 --> 00:03:16,110 398 | the the fact they're being invaded means 399 | 他们受到入侵的事实意味着 400 | 401 | 81 402 | 00:03:16,310 --> 00:03:18,390 403 | about the strength of the other do you 404 | 关于对方的力量 405 | 406 | 82 407 | 00:03:18,590 --> 00:03:20,939 408 | surrender do you fight so those are 409 | 你投降了，所以那些 410 | 411 | 83 412 | 00:03:21,139 --> 00:03:22,259 413 | games that they're going to have similar 414 | 他们将拥有相似的游戏 415 | 416 | 84 417 | 00:03:22,459 --> 00:03:23,789 418 | kinds of features to these and it's 419 | 这些功能的特点是 420 | 421 | 85 422 | 00:03:23,989 --> 00:03:26,009 423 | gonna be very important to develop a set 424 | 开发一套非常重要 425 | 426 | 86 427 | 00:03:26,209 --> 00:03:27,780 428 | of a way of representing these things 429 | 代表这些东西的方式 430 | 431 | 87 432 | 00:03:27,979 --> 00:03:29,910 433 | and some thoughts about analyzing those 434 | 以及有关分析这些问题的一些想法 435 | 436 | 88 437 | 00:03:30,110 --> 00:03:32,390 438 | so that's where we're headed next and 439 | 这就是我们接下来要去的地方 440 | 441 | 89 442 | 00:03:32,590 --> 00:03:35,070 443 | we'll see more a lot more of this very 444 | 我们将在其中看到更多 445 | 446 | 90 447 | 00:03:35,270 --> 00:03:40,270 448 | shortly 449 | 不久 450 | 451 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 5/GTO-5-03 - Stochastic Games.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,060 --> 00:00:02,829 3 | let's now speak briefly about stochastic 4 | 现在让我们简要地谈论一下随机 5 | 6 | 2 7 | 00:00:03,029 --> 00:00:08,080 8 | games this is a topic that lends itself 9 | 游戏这是一个适合自己的话题 10 | 11 | 3 12 | 00:00:08,279 --> 00:00:11,859 13 | to a very long discussion and quite a 14 | 经过很长时间的讨论， 15 | 16 | 4 17 | 00:00:12,058 --> 00:00:16,240 18 | complicated one but we'll touch on the 19 | 复杂的一个，但我们将涉及 20 | 21 | 5 22 | 00:00:16,440 --> 00:00:19,499 23 | main points and position it to the 24 | 要点并定位到 25 | 26 | 6 27 | 00:00:19,699 --> 00:00:22,600 28 | landscape of topics we're discussing and 29 | 我们正在讨论的主题以及 30 | 31 | 7 32 | 00:00:22,800 --> 00:00:24,970 33 | so the strategy the striking point are 34 | 所以策略的重点是 35 | 36 | 8 37 | 00:00:25,170 --> 00:00:27,730 38 | repeated games as we know a repeated 39 | 我们知道重复的游戏 40 | 41 | 9 42 | 00:00:27,929 --> 00:00:31,390 43 | game is simply a game in normal form for 44 | 游戏只是一种普通形式的游戏 45 | 46 | 10 47 | 00:00:31,589 --> 00:00:33,640 48 | example that will repeat over and over 49 | 这个例子会一遍又一遍地重复 50 | 51 | 11 52 | 00:00:33,840 --> 00:00:37,599 53 | again for example we play prisoner's 54 | 再例如，我们扮演囚犯 55 | 56 | 12 57 | 00:00:37,799 --> 00:00:40,839 58 | dilemma once twice three times maybe 59 | 两次困境两次，也许 60 | 61 | 13 62 | 00:00:41,039 --> 00:00:42,518 63 | finite time maybe an infinite number of 64 | 有限的时间也许是无限的 65 | 66 | 14 67 | 00:00:42,719 --> 00:00:44,678 68 | time then we accumulate all the rewards 69 | 时间，然后我们积累所有的奖励 70 | 71 | 15 72 | 00:00:44,878 --> 00:00:47,369 73 | all the time to sum over all rewards 74 | 一直在总结所有奖励 75 | 76 | 16 77 | 00:00:47,570 --> 00:00:49,899 78 | stochastic gamely is a generalization of 79 | 随机博弈是对 80 | 81 | 17 82 | 00:00:50,100 --> 00:00:53,619 83 | it where we play games repeatedly but 84 | 它是我们反复玩游戏的地方，但是 85 | 86 | 18 87 | 00:00:53,820 --> 00:00:56,439 88 | not necessarily the same game so we play 89 | 不一定是同一游戏，所以我们玩 90 | 91 | 19 92 | 00:00:56,640 --> 00:00:59,768 93 | a game depending on how we played that 94 | 取决于我们如何玩的游戏 95 | 96 | 20 97 | 00:00:59,969 --> 00:01:01,809 98 | game let's say prisoner's dilemma we 99 | 游戏让我们说囚徒的困境 100 | 101 | 21 102 | 00:01:02,009 --> 00:01:04,629 103 | each got some payoff but depending on 104 | 每个人都有一些回报，但取决于 105 | 106 | 22 107 | 00:01:04,829 --> 00:01:06,869 108 | how we play the game we also probable 109 | 我们如何玩游戏，我们也可能 110 | 111 | 23 112 | 00:01:07,069 --> 00:01:08,769 113 | probabilistically transition to some 114 | 概率性地过渡到一些 115 | 116 | 24 117 | 00:01:08,969 --> 00:01:13,179 118 | other game play that in turn and the 119 | 依次进行的其他游戏和 120 | 121 | 25 122 | 00:01:13,379 --> 00:01:17,109 123 | process continues a graphical way to 124 | 过程继续以图形方式 125 | 126 | 26 127 | 00:01:17,310 --> 00:01:20,230 128 | look at it is here if if this here is a 129 | 看这是这里，如果这是一个 130 | 131 | 27 132 | 00:01:20,430 --> 00:01:23,259 133 | repeated game where you play the same 134 | 重复游戏，您玩同样的游戏 135 | 136 | 28 137 | 00:01:23,459 --> 00:01:26,079 138 | game over and over again here you play 139 | 在这里玩一遍又一遍的游戏 140 | 141 | 29 142 | 00:01:26,280 --> 00:01:30,730 143 | the game and then if you happen to play 144 | 游戏，然后如果您碰巧玩 145 | 146 | 30 147 | 00:01:30,930 --> 00:01:33,789 148 | this your transition to this game if you 149 | 这是您向该游戏的过渡，如果您 150 | 151 | 31 152 | 00:01:33,989 --> 00:01:35,349 153 | happen to just play this you maybe 154 | 碰巧只是玩这个，你也许 155 | 156 | 32 157 | 00:01:35,549 --> 00:01:37,628 158 | transition to the same game if you play 159 | 如果您玩的话，过渡到同一个游戏 160 | 161 | 33 162 | 00:01:37,828 --> 00:01:39,429 163 | this you change your transition here if 164 | 如果你改变这里的过渡 165 | 166 | 34 167 | 00:01:39,629 --> 00:01:40,689 168 | you play this maybe you'll play this 169 | 你玩这个也许你会玩这个 170 | 171 | 35 172 | 00:01:40,890 --> 00:01:44,198 173 | game again and so on from each game you 174 | 再玩一次，依此类推 175 | 176 | 36 177 | 00:01:44,399 --> 00:01:47,890 178 | transition probabilistically to toe it 179 | 概率过渡到脚趾 180 | 181 | 37 182 | 00:01:48,090 --> 00:01:51,959 183 | to other games this is a stochastic game 184 | 对其他游戏来说这是一个随机的游戏 185 | 186 | 38 187 | 00:01:52,159 --> 00:01:55,750 188 | formally speaking it's it's the 189 | 正式来说是 190 | 191 | 39 192 | 00:01:55,950 --> 00:01:57,759 193 | following tuple it's a lot of notation 194 | 以下元组有很多符号 195 | 196 | 40 197 | 00:01:57,959 --> 00:02:01,689 198 | but the concept is exactly so we have is 199 | 但是这个概念正是我们所拥有的 200 | 201 | 41 202 | 00:02:01,890 --> 00:02:06,488 203 | a finite set of states Q we have a set 204 | 有限的一组状态Q我们有一组 205 | 206 | 42 207 | 00:02:06,688 --> 00:02:11,340 208 | of players we have a set of actions 209 | 我们有一系列行动 210 | 211 | 43 212 | 00:02:11,539 --> 00:02:13,660 213 | where 214 | 哪里 215 | 216 | 44 217 | 00:02:13,860 --> 00:02:17,680 218 | actions are available to two specific 219 | 有两个特定的操作可用 220 | 221 | 45 222 | 00:02:17,879 --> 00:02:20,620 223 | players of a sub I is the action 224 | 子我是玩家的动作 225 | 226 | 46 227 | 00:02:20,819 --> 00:02:25,390 228 | available player I and then we have two 229 | 可用的球员我，然后我们有两个 230 | 231 | 47 232 | 00:02:25,590 --> 00:02:29,650 233 | two functions we have the transition 234 | 我们有两个功能过渡 235 | 236 | 48 237 | 00:02:29,849 --> 00:02:31,660 238 | probability function so depending on the 239 | 概率函数，因此取决于 240 | 241 | 49 242 | 00:02:31,860 --> 00:02:35,130 243 | state were in and on the actions we took 244 | 陈述我们采取的行动 245 | 246 | 50 247 | 00:02:35,330 --> 00:02:38,530 248 | we move to each of any of the other 249 | 我们移到其他任何一个 250 | 251 | 51 252 | 00:02:38,729 --> 00:02:41,110 253 | states or the very same state with a 254 | 状态或与一个状态完全相同的状态 255 | 256 | 52 257 | 00:02:41,310 --> 00:02:43,660 258 | certain probability as governed by this 259 | 受此约束的特定概率 260 | 261 | 53 262 | 00:02:43,860 --> 00:02:50,320 263 | probability distribution and and 264 | 概率分布和 265 | 266 | 54 267 | 00:02:50,520 --> 00:02:53,770 268 | similarly a reward is the reward 269 | 同样，奖励就是奖励 270 | 271 | 55 272 | 00:02:53,969 --> 00:02:57,580 273 | function which tells us if in a certain 274 | 告诉我们如果 275 | 276 | 56 277 | 00:02:57,780 --> 00:03:00,010 278 | State a certain action profile was taken 279 | 说明采取了某种行动 280 | 281 | 57 282 | 00:03:00,210 --> 00:03:03,000 283 | by the agents then this is a reward to 284 | 由代理商，那么这是对 285 | 286 | 58 287 | 00:03:03,199 --> 00:03:06,430 288 | to that particular agent to eat each of 289 | 给那个特工吃 290 | 291 | 59 292 | 00:03:06,629 --> 00:03:08,920 293 | the agents so our sub I is a reward to 294 | 代理商，所以我们的子我是对 295 | 296 | 60 297 | 00:03:09,120 --> 00:03:12,960 298 | two agent I that's the formal definition 299 | 我是正式的两个代理人 300 | 301 | 61 302 | 00:03:13,159 --> 00:03:15,580 303 | notice that it's sort of assumes 304 | 请注意，这是一种假设 305 | 306 | 62 307 | 00:03:15,780 --> 00:03:18,250 308 | implicitly that you have the same action 309 | 暗含你有相同的动作 310 | 311 | 63 312 | 00:03:18,449 --> 00:03:24,520 313 | spaces here but you could define it 314 | 这里有空格，但您可以定义它 315 | 316 | 64 317 | 00:03:24,719 --> 00:03:26,380 318 | otherwise it simply would involve more 319 | 否则，它将涉及更多 320 | 321 | 65 322 | 00:03:26,580 --> 00:03:27,990 323 | notations or nothing inherently 324 | 表示法或本来就不存在 325 | 326 | 66 327 | 00:03:28,189 --> 00:03:30,789 328 | important about the action spaces being 329 | 关于动作空间很重要 330 | 331 | 67 332 | 00:03:30,989 --> 00:03:32,500 333 | the same in the different games within 334 | 在不同的游戏中相同 335 | 336 | 68 337 | 00:03:32,699 --> 00:03:36,550 338 | this stochastic game so just a few final 339 | 这个随机的游戏，所以最后几场 340 | 341 | 69 342 | 00:03:36,750 --> 00:03:39,520 343 | comments on it first of all as we as we 344 | 首先像我们一样评论它 345 | 346 | 70 347 | 00:03:39,719 --> 00:03:42,100 348 | saw this obviously general generalizes 349 | 看到这显然可以概括 350 | 351 | 71 352 | 00:03:42,300 --> 00:03:44,710 353 | the notion of a repeated game but it 354 | 重复游戏的概念，但它 355 | 356 | 72 357 | 00:03:44,909 --> 00:03:47,469 358 | also generalizes the notion of an MDP or 359 | 也概括了MDP或 360 | 361 | 73 362 | 00:03:47,669 --> 00:03:51,939 363 | a Markov decision process if a if a 364 | 马尔可夫决策过程，如果 365 | 366 | 74 367 | 00:03:52,139 --> 00:03:54,009 368 | sarcastic game if a repeated game is a 369 | 如果重复的游戏是一个讽刺游戏 370 | 371 | 75 372 | 00:03:54,209 --> 00:03:57,090 373 | stochastic game with only one game a 374 | 一个只有一个游戏的随机游戏 375 | 376 | 76 377 | 00:03:57,289 --> 00:04:00,910 378 | Markov decision process on MVP is a game 379 | MVP的马尔可夫决策过程是一个游戏 380 | 381 | 77 382 | 00:04:01,110 --> 00:04:04,390 383 | with only one player and so you have 384 | 只有一个玩家，所以你有 385 | 386 | 78 387 | 00:04:04,590 --> 00:04:08,770 388 | States there where the agents take agent 389 | 代理商所在的国家 390 | 391 | 79 392 | 00:04:08,969 --> 00:04:11,620 393 | takes an action receives a remediate 394 | 采取行动得到补救 395 | 396 | 80 397 | 00:04:11,819 --> 00:04:13,920 398 | reward and probably move to some other 399 | 奖励并可能转移到其他人 400 | 401 | 81 402 | 00:04:14,120 --> 00:04:16,900 403 | state and the only difference is that he 404 | 状态，唯一的区别是他 405 | 406 | 82 407 | 00:04:17,100 --> 00:04:20,230 408 | is the only actor in the setting I 409 | 是背景中我唯一的演员 410 | 411 | 83 412 | 00:04:20,430 --> 00:04:22,240 413 | mentioned this because well empty peas 414 | 提到这个是因为豌豆很好 415 | 416 | 84 417 | 00:04:22,439 --> 00:04:24,680 418 | have been studied 419 | 已经研究过 420 | 421 | 85 422 | 00:04:24,879 --> 00:04:27,860 423 | naturally in a variety of disciplines 424 | 自然地在各种学科中 425 | 426 | 86 427 | 00:04:28,060 --> 00:04:30,139 428 | from optimization to computer science to 429 | 从优化到计算机科学，再到 430 | 431 | 87 432 | 00:04:30,339 --> 00:04:34,400 433 | pure math and beyond but also these two 434 | 纯粹的数学，以及超越这两个 435 | 436 | 88 437 | 00:04:34,600 --> 00:04:37,370 438 | perspectives of generalization repeated 439 | 泛化观点重复 440 | 441 | 89 442 | 00:04:37,569 --> 00:04:41,050 443 | games and of MVPs give you a sense for 444 | 游戏和MVP可以带给您 445 | 446 | 90 447 | 00:04:41,250 --> 00:04:44,030 448 | the theory and investigations into 449 | 理论与研究 450 | 451 | 91 452 | 00:04:44,230 --> 00:04:47,240 453 | stochastic games so from repeated games 454 | 重复游戏中的随机游戏 455 | 456 | 92 457 | 00:04:47,439 --> 00:04:52,129 458 | we inherit the definitions of different 459 | 我们继承了不同的定义 460 | 461 | 93 462 | 00:04:52,329 --> 00:04:53,810 463 | ways of the aggregating rewards over 464 | 汇总奖励的方式 465 | 466 | 94 467 | 00:04:54,009 --> 00:04:57,740 468 | time you can have limited average 469 | 有限的平均时间 470 | 471 | 95 472 | 00:04:57,939 --> 00:05:01,389 473 | rewards future discounted rewards 474 | 奖励未来的折扣奖励 475 | 476 | 96 477 | 00:05:01,589 --> 00:05:05,829 478 | whereas from the literature on 479 | 而根据文献 480 | 481 | 97 482 | 00:05:06,029 --> 00:05:11,180 483 | optimization and on MVPs we get notions 484 | 优化，在MVP上我们有了概念 485 | 486 | 98 487 | 00:05:11,379 --> 00:05:17,180 488 | such as stationarity and Markovian 489 | 例如平稳性和马尔可夫式 490 | 491 | 99 492 | 00:05:17,379 --> 00:05:20,329 493 | strategies these have to do with we also 494 | 这些与我们有关的策略 495 | 496 | 100 497 | 00:05:20,529 --> 00:05:22,370 498 | have notions of reach ability about the 499 | 有关于 500 | 501 | 101 502 | 00:05:22,569 --> 00:05:24,079 503 | structure of the underlying transition 504 | 基本过渡的结构 505 | 506 | 102 507 | 00:05:24,279 --> 00:05:26,569 508 | probability and so again these are 509 | 概率，这些又是 510 | 511 | 103 512 | 00:05:26,769 --> 00:05:28,970 513 | issues that are involved that we won't 514 | 我们不会涉及的问题 515 | 516 | 104 517 | 00:05:29,170 --> 00:05:33,620 518 | get to into more in this lecture but at 519 | 在本讲座中深入了解，但在 520 | 521 | 105 522 | 00:05:33,819 --> 00:05:38,819 523 | least we flagged their existence 524 | 至少我们标记了它们的存在 525 | 526 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 6/GTO-6-03 - Bayesian Games - Second Definition.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,030 --> 00:00:02,738 3 | so here is an alternative definition of 4 | 所以这是对 5 | 6 | 2 7 | 00:00:02,939 --> 00:00:05,408 8 | Asian games that is essentially 9 | 从本质上讲，亚洲运动会 10 | 11 | 3 12 | 00:00:05,609 --> 00:00:09,100 13 | identical mathematically but presented 14 | 数学上相同但呈现 15 | 16 | 4 17 | 00:00:09,300 --> 00:00:14,740 18 | differently it's based on types or more 19 | 根据类型或更多而不同 20 | 21 | 5 22 | 00:00:14,939 --> 00:00:19,060 23 | fully epistemic types and the type of 24 | 完全认识的类型和类型 25 | 26 | 6 27 | 00:00:19,260 --> 00:00:22,419 28 | the agent is supposed to capture 29 | 该代理应该捕获 30 | 31 | 7 32 | 00:00:22,618 --> 00:00:24,489 33 | everything that's private information to 34 | 所有的私人信息 35 | 36 | 8 37 | 00:00:24,689 --> 00:00:29,589 38 | the agents so if you look at you know 39 | 代理商，所以如果你看着你知道 40 | 41 | 9 42 | 00:00:29,789 --> 00:00:32,948 43 | the first definition of Asian game that 44 | 亚洲游戏的第一个定义是 45 | 46 | 10 47 | 00:00:33,149 --> 00:00:35,890 48 | we saw having to do with certainty about 49 | 我们看到与确定有关 50 | 51 | 11 52 | 00:00:36,090 --> 00:00:38,768 53 | types and the common pryor then the type 54 | 类型和普通的pryor然后类型 55 | 56 | 12 57 | 00:00:38,969 --> 00:00:42,579 58 | of the agent was her private signal that 59 | 她的私人信号是 60 | 61 | 13 62 | 00:00:42,780 --> 00:00:46,390 63 | is the information set in which the the 64 | 是其中的信息集 65 | 66 | 14 67 | 00:00:46,590 --> 00:00:52,079 68 | get chosen game lies as well as 69 | 被选择的游戏谎言以及 70 | 71 | 15 72 | 00:00:52,280 --> 00:00:54,959 73 | everything that emanates from it namely 74 | 由此产生的一切 75 | 76 | 16 77 | 00:00:55,159 --> 00:00:57,608 78 | her beliefs about what the possible 79 | 她对可能的信念 80 | 81 | 17 82 | 00:00:57,808 --> 00:00:59,529 83 | information of the other agents are and 84 | 其他代理商的信息是和 85 | 86 | 18 87 | 00:00:59,729 --> 00:01:00,759 88 | information of the other agent about 89 | 其他代理商的有关信息 90 | 91 | 19 92 | 00:01:00,960 --> 00:01:03,929 93 | yourself and so forth so all of that is 94 | 自己等等，所有这些都是 95 | 96 | 20 97 | 00:01:04,129 --> 00:01:07,238 98 | folded into the notion of a type so 99 | 折叠成一个类型的概念 100 | 101 | 21 102 | 00:01:07,438 --> 00:01:09,340 103 | that's mathematically very convenient 104 | 从数学上讲这很方便 105 | 106 | 22 107 | 00:01:09,540 --> 00:01:12,659 108 | packing all this information into a type 109 | 将所有这些信息打包成一个类型 110 | 111 | 23 112 | 00:01:12,859 --> 00:01:15,459 113 | formally speaking then the Bayesian game 114 | 正式来说贝叶斯博弈 115 | 116 | 24 117 | 00:01:15,659 --> 00:01:19,558 118 | is defined as follows it's defined as 119 | 定义如下 120 | 121 | 25 122 | 00:01:19,759 --> 00:01:24,369 123 | this tuple that is as follows we have a 124 | 这个元组如下 125 | 126 | 26 127 | 00:01:24,569 --> 00:01:27,789 128 | set of agents we have the actions 129 | 我们有行动的代理人集合 130 | 131 | 27 132 | 00:01:27,989 --> 00:01:29,709 133 | available to the agent so now we don't 134 | 提供给代理商，所以现在我们不 135 | 136 | 28 137 | 00:01:29,909 --> 00:01:32,230 138 | have sets of games we have very directly 139 | 有几组游戏我们非常直接 140 | 141 | 29 142 | 00:01:32,430 --> 00:01:34,168 143 | the actions available to the agent and 144 | 代理可用的操作和 145 | 146 | 30 147 | 00:01:34,368 --> 00:01:37,750 148 | now we have the type this abstract 149 | 现在我们有这个抽象的类型 150 | 151 | 31 152 | 00:01:37,950 --> 00:01:39,969 153 | mathematical object that care that 154 | 关心那个的数学对象 155 | 156 | 32 157 | 00:01:40,170 --> 00:01:42,189 158 | captures the private information of the 159 | 捕获的私人信息 160 | 161 | 33 162 | 00:01:42,390 --> 00:01:45,369 163 | agents so we have a type for each agent 164 | 代理商，所以我们为每个代理商都有一个类型 165 | 166 | 34 167 | 00:01:45,569 --> 00:01:49,840 168 | and we have a common prior as B as in 169 | 我们有一个共同的先验B 170 | 171 | 35 172 | 00:01:50,040 --> 00:01:53,799 173 | the first definition of games we have a 174 | 游戏的第一个定义是 175 | 176 | 36 177 | 00:01:54,000 --> 00:01:55,750 178 | common prior but now it's not over games 179 | 以前很常见，但现在还没有结束 180 | 181 | 37 182 | 00:01:55,950 --> 00:01:59,439 183 | it's over types so agent each agent had 184 | 这是类型，所以每个代理都有代理 185 | 186 | 38 187 | 00:01:59,640 --> 00:02:04,238 188 | the type and that prior is common the 189 | 类型和先验是常见的 190 | 191 | 39 192 | 00:02:04,438 --> 00:02:06,899 193 | type is chosen according to a 194 | 类型根据选择 195 | 196 | 40 197 | 00:02:07,099 --> 00:02:09,460 198 | probability distribution that's commonly 199 | 通常的概率分布 200 | 201 | 41 202 | 00:02:09,659 --> 00:02:12,610 203 | known by all the agents and each agent 204 | 所有代理商和每个代理商都知道 205 | 206 | 42 207 | 00:02:12,810 --> 00:02:13,240 208 | knows 209 | 知道 210 | 211 | 43 212 | 00:02:13,439 --> 00:02:15,850 213 | type therefore they also have a 214 | 因此，他们也有一个 215 | 216 | 44 217 | 00:02:16,050 --> 00:02:17,950 218 | posterior about the type of the other 219 | 关于其他类型的后验 220 | 221 | 45 222 | 00:02:18,150 --> 00:02:21,130 223 | agents and beliefs about what the other 224 | 代理人和其他人的信念 225 | 226 | 46 227 | 00:02:21,330 --> 00:02:22,750 228 | agents might believe it about their own 229 | 特工可能会相信自己 230 | 231 | 47 232 | 00:02:22,949 --> 00:02:26,140 233 | type and so on and so forth this is the 234 | 类型，依此类推，这就是 235 | 236 | 48 237 | 00:02:26,340 --> 00:02:28,450 238 | type of the agents and we have the 239 | 代理商的类型，我们有 240 | 241 | 49 242 | 00:02:28,650 --> 00:02:33,340 243 | utility function now depend not only are 244 | 实用功能现在不仅取决于 245 | 246 | 50 247 | 00:02:33,539 --> 00:02:36,100 248 | the actions taken by the agents but of 249 | 代理商采取的行动，但 250 | 251 | 51 252 | 00:02:36,300 --> 00:02:40,830 253 | their type that's the formal definition 254 | 他们的类型就是正式定义 255 | 256 | 52 257 | 00:02:41,030 --> 00:02:45,280 258 | again it's mathematically in something 259 | 再次从数学上讲 260 | 261 | 53 262 | 00:02:45,479 --> 00:02:47,980 263 | very simple but the intuition is 264 | 很简单，但直觉是 265 | 266 | 54 267 | 00:02:48,180 --> 00:02:50,469 268 | complicated because the notion of a type 269 | 复杂，因为类型的概念 270 | 271 | 55 272 | 00:02:50,669 --> 00:02:53,650 273 | packs into it a lot of things so let's 274 | 包装了很多东西，所以让我们 275 | 276 | 56 277 | 00:02:53,849 --> 00:02:56,389 278 | see it in action 279 | 看到它在行动 280 | 281 | 57 282 | 00:02:57,409 --> 00:03:02,230 283 | consider this game that we saw when we 284 | 考虑一下我们看到的这个游戏 285 | 286 | 58 287 | 00:03:02,430 --> 00:03:05,939 288 | discussed the first definition of 289 | 讨论了第一个定义 290 | 291 | 59 292 | 00:03:06,139 --> 00:03:10,630 293 | evasion games again we had four possible 294 | 再次进行回避游戏，我们有四种可能 295 | 296 | 60 297 | 00:03:10,830 --> 00:03:14,830 298 | games being played chosen by at random 299 | 随机选择的正在玩的游戏 300 | 301 | 61 302 | 00:03:15,030 --> 00:03:18,790 303 | by nature according to this prior and we 304 | 根据这个先天的本质，我们 305 | 306 | 62 307 | 00:03:18,990 --> 00:03:22,830 308 | all we had the private signals that the 309 | 我们所有的私人信号都表明 310 | 311 | 63 312 | 00:03:23,030 --> 00:03:25,480 313 | information sets that the agents found 314 | 代理商发现的信息集 315 | 316 | 64 317 | 00:03:25,680 --> 00:03:28,750 318 | themselves in here is the type 319 | 自己在这里是类型 320 | 321 | 65 322 | 00:03:28,949 --> 00:03:31,630 323 | perspective on this so what are the 324 | 对此的观点是什么 325 | 326 | 66 327 | 00:03:31,830 --> 00:03:33,790 328 | actions available to agents very simply 329 | 代理可以轻松执行的操作 330 | 331 | 67 332 | 00:03:33,990 --> 00:03:39,000 333 | the row agent has the up or down action 334 | 行代理具有向上或向下操作 335 | 336 | 68 337 | 00:03:39,199 --> 00:03:42,039 338 | actions and the column agents have the 339 | 操作和列代理具有 340 | 341 | 69 342 | 00:03:42,239 --> 00:03:45,670 343 | left-to-right actions that pay off 344 | 从左到右的行动有回报 345 | 346 | 70 347 | 00:03:45,870 --> 00:03:48,580 348 | however will depend on their type so 349 | 但是将取决于它们的类型，所以 350 | 351 | 71 352 | 00:03:48,780 --> 00:03:50,980 353 | let's let's look for example what 354 | 让我们来看例如 355 | 356 | 72 357 | 00:03:51,180 --> 00:03:53,860 358 | happens here so what is the payoff when 359 | 发生在这里，那么什么是回报 360 | 361 | 73 362 | 00:03:54,060 --> 00:03:57,689 363 | the agent the raw agent plays up and the 364 | 代理，原始代理出现，并且 365 | 366 | 74 367 | 00:03:57,889 --> 00:04:01,509 368 | column player plays lift well that 369 | 专栏运动员发挥良好， 370 | 371 | 75 372 | 00:04:01,709 --> 00:04:05,830 373 | depends if the type of the agents is 374 | 取决于代理商的类型是 375 | 376 | 76 377 | 00:04:06,030 --> 00:04:08,259 378 | this one or what is the type the type 379 | 这一种还是什么类型 380 | 381 | 77 382 | 00:04:08,459 --> 00:04:12,100 383 | corresponds now to this information that 384 | 现在对应于该信息， 385 | 386 | 78 387 | 00:04:12,300 --> 00:04:16,449 388 | they have and the type of the second 389 | 他们有和第二种 390 | 391 | 79 392 | 00:04:16,649 --> 00:04:22,660 393 | player is this well what the path then 394 | 玩家好在那走什么路 395 | 396 | 80 397 | 00:04:22,860 --> 00:04:25,329 398 | well the path corresponds now to this 399 | 路径现在与此对应 400 | 401 | 81 402 | 00:04:25,529 --> 00:04:26,840 403 | cell right there 404 | 那里的细胞 405 | 406 | 82 407 | 00:04:27,040 --> 00:04:29,509 408 | is what happens when they play up and 409 | 当他们玩耍时会发生什么 410 | 411 | 83 412 | 00:04:29,709 --> 00:04:35,180 413 | left and so you get 2 and 0 when the 414 | 左，所以当得到2和0时 415 | 416 | 84 417 | 00:04:35,379 --> 00:04:39,500 418 | types are as they are let's take some 419 | 类型是一样，让我们​​来一些 420 | 421 | 85 422 | 00:04:39,699 --> 00:04:42,620 423 | other random example here is clear the 424 | 其他随机示例在这里很明显 425 | 426 | 86 427 | 00:04:42,819 --> 00:04:45,230 428 | slide let's take some rather rather 429 | 幻灯片让我们采取一些宁可 430 | 431 | 87 432 | 00:04:45,430 --> 00:04:50,600 433 | example here clearing the slide and let 434 | 这里的示例清除幻灯片，然后让 435 | 436 | 88 437 | 00:04:50,800 --> 00:04:58,210 438 | us look for example at down and left 439 | 我们看一下向下和向左 440 | 441 | 89 442 | 00:04:58,410 --> 00:05:04,250 443 | when the types are these well what is 444 | 当这些类型很好的时候 445 | 446 | 90 447 | 00:05:04,449 --> 00:05:06,560 448 | the what are the what what are the what 449 | 什么是什么什么是什么 450 | 451 | 91 452 | 00:05:06,759 --> 00:05:10,220 453 | are the types so the type is sick is 454 | 是类型，所以类型是生病的是 455 | 456 | 92 457 | 00:05:10,420 --> 00:05:15,439 458 | this one right here so the this is the 459 | 这个就在这里，这是 460 | 461 | 93 462 | 00:05:15,639 --> 00:05:18,470 463 | information available to the first agent 464 | 第一代理人可获得的信息 465 | 466 | 94 467 | 00:05:18,670 --> 00:05:22,460 468 | the second agent has this information 469 | 第二个代理商有此信息 470 | 471 | 95 472 | 00:05:22,660 --> 00:05:25,850 473 | available to her which means that this 474 | 向她提供这意味着 475 | 476 | 96 477 | 00:05:26,050 --> 00:05:31,600 478 | right here is the game being played and 479 | 在这里是正在玩的游戏， 480 | 481 | 97 482 | 00:05:31,800 --> 00:05:35,660 483 | what is DL DL means that we're playing 484 | DL DL是什么意思我们在玩 485 | 486 | 98 487 | 00:05:35,860 --> 00:05:38,270 488 | down and lifts there will be this one 489 | 向下并举起会有这个 490 | 491 | 99 492 | 00:05:38,470 --> 00:05:40,579 493 | and therefore the payoff will be 494 | 因此，回报将是 495 | 496 | 100 497 | 00:05:40,779 --> 00:05:46,520 498 | correspondingly 0 and 0 so you can look 499 | 分别对应0和0，所以您可以看一下 500 | 501 | 101 502 | 00:05:46,720 --> 00:05:49,310 503 | at other examples and figure out what 504 | 在其他例子中，找出什么 505 | 506 | 102 507 | 00:05:49,509 --> 00:05:55,490 508 | the type based formulation means by just 509 | 基于类型的表述意味着 510 | 511 | 103 512 | 00:05:55,689 --> 00:05:58,310 513 | taking some random row here and figuring 514 | 在这里采取一些随机行并计算 515 | 516 | 104 517 | 00:05:58,509 --> 00:06:04,930 518 | out why the row is the way it is the 519 | 找出为什么行是这样的 520 | 521 | 105 522 | 00:06:05,129 --> 00:06:09,230 523 | last thing to say here before we move on 524 | 在我们继续之前要在这里说的最后一件事 525 | 526 | 106 527 | 00:06:09,430 --> 00:06:12,290 528 | to analysis is that in this particular 529 | 要分析的是 530 | 531 | 107 532 | 00:06:12,490 --> 00:06:17,750 533 | example by fixing the type you ended up 534 | 通过修复最终类型的示例 535 | 536 | 108 537 | 00:06:17,949 --> 00:06:22,310 538 | with a very specific game and this is a 539 | 一个非常具体的游戏，这是一个 540 | 541 | 109 542 | 00:06:22,509 --> 00:06:26,750 543 | complicated topic where in fact if you 544 | 复杂的话题实际上，如果您 545 | 546 | 110 547 | 00:06:26,949 --> 00:06:29,689 548 | wanted to map it uncertainty over games 549 | 想将不确定性映射到游戏上 550 | 551 | 111 552 | 00:06:29,889 --> 00:06:31,819 553 | you may not have unique game and you 554 | 您可能没有独特的游戏，而您 555 | 556 | 112 557 | 00:06:32,019 --> 00:06:33,860 558 | need to look at the set of games and the 559 | 需要看一组游戏和 560 | 561 | 113 562 | 00:06:34,060 --> 00:06:35,220 563 | expectation there 564 | 那里的期望 565 | 566 | 114 567 | 00:06:35,420 --> 00:06:38,930 568 | but I'm just flanging this as a topic 569 | 但我只是把它当作主题 570 | 571 | 115 572 | 00:06:39,129 --> 00:06:41,340 573 | what we discussed will give you a good 574 | 我们讨论的内容将为您带来好处 575 | 576 | 116 577 | 00:06:41,540 --> 00:06:43,590 578 | handle on the two formulations of 579 | 处理以下两个公式 580 | 581 | 117 582 | 00:06:43,790 --> 00:06:45,689 583 | Bayesian games the explicit from a 584 | 贝叶斯游戏从 585 | 586 | 118 587 | 00:06:45,889 --> 00:06:49,410 588 | listing of games and a common prior over 589 | 游戏清单和共同的优先权 590 | 591 | 119 592 | 00:06:49,610 --> 00:06:51,930 593 | them and a partition structure for the 594 | 他们和一个分区结构 595 | 596 | 120 597 | 00:06:52,129 --> 00:06:57,129 598 | agents or the type based formulation 599 | 剂或基于类型的配方 600 | 601 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory I-Week 7/GTO-7-01 - Coalitional Game Theory - Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:01,310 --> 00:00:04,569 3 | coalitional game theory is an approach 4 | 联盟博弈论是一种方法 5 | 6 | 2 7 | 00:00:04,769 --> 00:00:06,729 8 | to modeling strategic situations that 9 | 对战略形势进行建模 10 | 11 | 3 12 | 00:00:06,929 --> 00:00:09,220 13 | stands in contrast was usually called 14 | 相反，通常称为 15 | 16 | 4 17 | 00:00:09,419 --> 00:00:13,809 18 | non cooperative game theory and in fact 19 | 非合作博弈论，实际上 20 | 21 | 5 22 | 00:00:14,009 --> 00:00:15,580 23 | coalitional game theory is often called 24 | 联盟博弈理论通常被称为 25 | 26 | 6 27 | 00:00:15,779 --> 00:00:18,460 28 | cooperative game theory the names are 29 | 合作博弈论的名字是 30 | 31 | 7 32 | 00:00:18,660 --> 00:00:20,230 33 | luminous reading and I'll get back to 34 | 夜光阅读，我会回到 35 | 36 | 8 37 | 00:00:20,429 --> 00:00:22,989 38 | that in a moment but let's first speak 39 | 片刻之后，让我们先说 40 | 41 | 9 42 | 00:00:23,189 --> 00:00:26,919 43 | about what situations does coalitional 44 | 关于什么情况联盟 45 | 46 | 10 47 | 00:00:27,118 --> 00:00:31,149 48 | game theory try to model you may 49 | 博弈论尝试为您建模 50 | 51 | 11 52 | 00:00:31,349 --> 00:00:33,209 53 | recognize these two fine-looking 54 | 认出这两个好看的 55 | 56 | 12 57 | 00:00:33,409 --> 00:00:37,320 58 | gentlemen the person on the left is 59 | 先生们，左边的人是 60 | 61 | 13 62 | 00:00:37,520 --> 00:00:40,839 63 | David Cameron Prime Minister of Britain 64 | 英国首相戴维·卡梅伦 65 | 66 | 14 67 | 00:00:41,039 --> 00:00:46,329 68 | and to his to his right as we look at it 69 | 看着他在他右边 70 | 71 | 15 72 | 00:00:46,530 --> 00:00:50,169 73 | he's left as they sit is Nicholas Clegg 74 | 当他们坐下时，他离开的是尼古拉斯·克莱格 75 | 76 | 16 77 | 00:00:50,369 --> 00:00:54,899 78 | his cultural partner now here are two 79 | 他的文化伙伴现在有两个 80 | 81 | 17 82 | 00:00:55,100 --> 00:00:58,689 83 | political rivals who nonetheless come 84 | 仍然来的政治对手 85 | 86 | 18 87 | 00:00:58,890 --> 00:01:00,640 88 | together and presumably there's a reason 89 | 在一起，大概是有原因的 90 | 91 | 19 92 | 00:01:00,840 --> 00:01:01,628 93 | there's something where they can 94 | 他们可以在某些地方 95 | 96 | 20 97 | 00:01:01,829 --> 00:01:03,819 98 | accomplish together that they cannot 99 | 一起完成，他们不能 100 | 101 | 21 102 | 00:01:04,019 --> 00:01:06,459 103 | accomplish alone in particular in this 104 | 尤其是独自完成 105 | 106 | 22 107 | 00:01:06,659 --> 00:01:09,009 108 | case command a majority in the in the 109 | case命令中的大多数 110 | 111 | 23 112 | 00:01:09,209 --> 00:01:11,560 113 | parliament and so that's a classical 114 | 议会，所以这是经典 115 | 116 | 24 117 | 00:01:11,760 --> 00:01:15,730 118 | example where a coalition formed in fact 119 | 实际上组成联盟的例子 120 | 121 | 25 122 | 00:01:15,930 --> 00:01:16,659 123 | we usually when you think about 124 | 我们通常在您想到的时候 125 | 126 | 26 127 | 00:01:16,859 --> 00:01:19,269 128 | coalition we think about political 129 | 我们考虑政治的联盟 130 | 131 | 27 132 | 00:01:19,469 --> 00:01:24,640 133 | parties but coalition's form not only in 134 | 政党，但联盟的形式不仅在 135 | 136 | 28 137 | 00:01:24,840 --> 00:01:27,640 138 | politics they certainly form in business 139 | 他们在业务中形成的政治 140 | 141 | 29 142 | 00:01:27,840 --> 00:01:31,959 143 | so the wind coalition is a the coming 144 | 因此，风电联盟即将到来 145 | 146 | 30 147 | 00:01:32,159 --> 00:01:35,799 148 | together of a number of firms in the 149 | 多家公司 150 | 151 | 31 152 | 00:01:36,000 --> 00:01:37,988 153 | United States to promote the joint 154 | 美国联合推广 155 | 156 | 32 157 | 00:01:38,188 --> 00:01:44,009 158 | agenda of wind energy turbines and such 159 | 风力发电机等的议程 160 | 161 | 33 162 | 00:01:44,209 --> 00:01:47,429 163 | again these are companies who are 164 | 这些都是公司 165 | 166 | 34 167 | 00:01:47,629 --> 00:01:50,409 168 | competitors and nonetheless they feel 169 | 竞争对手，尽管如此 170 | 171 | 35 172 | 00:01:50,609 --> 00:01:52,329 173 | that together there's things that can 174 | 在一起的东西可以 175 | 176 | 36 177 | 00:01:52,530 --> 00:01:54,969 178 | exist do that they can't do alone 179 | 存在，他们不能一个人做 180 | 181 | 37 182 | 00:01:55,170 --> 00:01:58,448 183 | for example Lobby government established 184 | 例如建立游说政府 185 | 186 | 38 187 | 00:01:58,649 --> 00:02:01,018 188 | standards and things of that nature 189 | 这种性质的标准和事物 190 | 191 | 39 192 | 00:02:01,218 --> 00:02:04,509 193 | now coalition's aren't always among 194 | 现在联盟并不总是 195 | 196 | 40 197 | 00:02:04,709 --> 00:02:07,058 198 | organizations or parties of ferns or 199 | 蕨类植物的组织或团体或 200 | 201 | 41 202 | 00:02:07,259 --> 00:02:09,580 203 | heavyweight things we as individuals 204 | 作为个人的重量级事物 205 | 206 | 42 207 | 00:02:09,780 --> 00:02:11,410 208 | routinely come together to accomplish 209 | 经常聚在一起完成 210 | 211 | 43 212 | 00:02:11,610 --> 00:02:12,010 213 | things 214 | 东西 215 | 216 | 44 217 | 00:02:12,210 --> 00:02:13,780 218 | together whether it's to enter in 219 | 一起进入 220 | 221 | 45 222 | 00:02:13,979 --> 00:02:18,030 223 | marriage or for example build a house 224 | 结婚或盖房子 225 | 226 | 46 227 | 00:02:18,229 --> 00:02:21,310 228 | when you have a carpenter Ln efficient 229 | 当你有一个木匠Ln高效 230 | 231 | 47 232 | 00:02:21,509 --> 00:02:24,580 233 | and a painter that come together they 234 | 和一个画家在一起 235 | 236 | 48 237 | 00:02:24,780 --> 00:02:26,800 238 | together can accomplish something that 239 | 一起可以完成一些 240 | 241 | 49 242 | 00:02:27,000 --> 00:02:30,280 243 | they can't on their own now the mere 244 | 他们现在不能只靠自己 245 | 246 | 50 247 | 00:02:30,479 --> 00:02:32,410 248 | fact that people come together doesn't 249 | 人们聚在一起的事实 250 | 251 | 51 252 | 00:02:32,610 --> 00:02:34,270 253 | mean that their interests are aligned or 254 | 表示他们的利益一致或 255 | 256 | 52 257 | 00:02:34,469 --> 00:02:37,510 258 | that they bring the same amount of value 259 | 他们带来相同的价值 260 | 261 | 53 262 | 00:02:37,710 --> 00:02:40,000 263 | to the coalition they've formed it could 264 | 他们组成的联盟可以 265 | 266 | 54 267 | 00:02:40,199 --> 00:02:44,200 268 | be for example that the framer of a 269 | 例如， 270 | 271 | 55 272 | 00:02:44,400 --> 00:02:47,800 273 | construction crew is irreplaceable 274 | 施工人员是不可替代的 275 | 276 | 56 277 | 00:02:48,000 --> 00:02:49,990 278 | but electricians are easy to find and 279 | 但是电工很容易找到 280 | 281 | 57 282 | 00:02:50,189 --> 00:02:51,790 283 | one could easily replace and presumably 284 | 一个可以轻松替换，大概 285 | 286 | 58 287 | 00:02:51,990 --> 00:02:53,890 288 | when they get paid for the house they 289 | 当他们得到房子的报酬时 290 | 291 | 59 292 | 00:02:54,090 --> 00:02:55,570 293 | build that should be reflected in how 294 | 建立应该反映在如何 295 | 296 | 60 297 | 00:02:55,770 --> 00:02:58,600 298 | they divide the the payment and so 299 | 他们分摊付款，所以 300 | 301 | 61 302 | 00:02:58,800 --> 00:03:00,750 303 | there's a competition of both 304 | 两者都有竞争 305 | 306 | 62 307 | 00:03:00,949 --> 00:03:05,380 308 | cooperation and competition here and and 309 | 合作与竞争在这里和 310 | 311 | 63 312 | 00:03:05,580 --> 00:03:08,010 313 | so for that reason calling these 314 | 所以出于这个原因 315 | 316 | 64 317 | 00:03:08,210 --> 00:03:12,600 318 | cooperative games as the common term is 319 | 合作游戏是通用术语 320 | 321 | 65 322 | 00:03:12,800 --> 00:03:17,140 323 | it is a little misleading just as much 324 | 同样有点误导 325 | 326 | 66 327 | 00:03:17,340 --> 00:03:18,939 328 | as non cooperative game theories 329 | 作为非合作博弈论 330 | 331 | 67 332 | 00:03:19,139 --> 00:03:23,860 333 | misleading as well because for example 334 | 也会引起误解，例如 335 | 336 | 68 337 | 00:03:24,060 --> 00:03:25,900 338 | if you look at the normal form game the 339 | 如果你看普通的形式游戏 340 | 341 | 69 342 | 00:03:26,099 --> 00:03:29,830 343 | canonical representation of a non 344 | 非的规范表示 345 | 346 | 70 347 | 00:03:30,030 --> 00:03:33,280 348 | cooperative game one can easily describe 349 | 合作游戏可以轻松描述 350 | 351 | 71 352 | 00:03:33,479 --> 00:03:35,770 353 | a completely harmonious situation of 354 | 完全和谐的情况 355 | 356 | 72 357 | 00:03:35,969 --> 00:03:37,660 358 | so-called team gains or common payoff 359 | 所谓的团队收益或共同收益 360 | 361 | 73 362 | 00:03:37,860 --> 00:03:39,790 363 | games where the interests of the ages 364 | 历代利益的游戏 365 | 366 | 74 367 | 00:03:39,990 --> 00:03:43,150 368 | are completely aligned so both non 369 | 完全对齐 370 | 371 | 75 372 | 00:03:43,349 --> 00:03:45,430 373 | cooperative game theory and coalitional 374 | 合作博弈论与联盟 375 | 376 | 76 377 | 00:03:45,629 --> 00:03:48,630 378 | or cooperative game theory model both 379 | 或合作博弈模型 380 | 381 | 77 382 | 00:03:48,830 --> 00:03:51,670 383 | competition and coordination the 384 | 竞争与协调 385 | 386 | 78 387 | 00:03:51,870 --> 00:03:53,439 388 | essential difference is the basic 389 | 基本区别是基本 390 | 391 | 79 392 | 00:03:53,639 --> 00:03:55,780 393 | modeling unit in coalitional game theory 394 | 联盟博弈模型的建模单元 395 | 396 | 80 397 | 00:03:55,979 --> 00:03:58,300 398 | the basic modeling unit is the group the 399 | 基本建模单位是 400 | 401 | 81 402 | 00:03:58,500 --> 00:04:01,030 403 | team and what they kind of can 404 | 团队，他们能做什么 405 | 406 | 82 407 | 00:04:01,229 --> 00:04:03,040 408 | accomplish and the analysis is based on 409 | 完成，分析基于 410 | 411 | 83 412 | 00:04:03,240 --> 00:04:08,240 413 | this basic modeling 414 | 这个基本的模型 415 | 416 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 1/GTO2-1-01 - Social Choice - Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,060 --> 00:00:01,958 3 | welcome to our advanced course on game 4 | 欢迎来到我们的游戏高级课程 5 | 6 | 2 7 | 00:00:02,158 --> 00:00:04,509 8 | theory this course covers the topics of 9 | 理论上，本课程涵盖以下主题 10 | 11 | 3 12 | 00:00:04,710 --> 00:00:07,000 13 | social choice mechanism design and 14 | 社会选择机制设计与 15 | 16 | 4 17 | 00:00:07,200 --> 00:00:09,939 18 | auctions and essentially it asks the 19 | 拍卖，本质上它要求 20 | 21 | 5 22 | 00:00:10,138 --> 00:00:12,700 23 | question if I know that agents are going 24 | 问我是否知道代理商要去 25 | 26 | 6 27 | 00:00:12,900 --> 00:00:15,130 28 | to behave game theoretically how can I 29 | 从理论上表现游戏我该如何 30 | 31 | 7 32 | 00:00:15,330 --> 00:00:17,829 33 | design interactions for them that lead 34 | 为他们设计互动 35 | 36 | 8 37 | 00:00:18,028 --> 00:00:21,339 38 | to good social outcomes the first topic 39 | 获得良好的社会成果的第一个主题 40 | 41 | 9 42 | 00:00:21,539 --> 00:00:22,629 43 | we're going to think about is called 44 | 我们要考虑的称为 45 | 46 | 10 47 | 00:00:22,829 --> 00:00:25,600 48 | social choice and here leaving aside the 49 | 社会选择，这里抛开 50 | 51 | 11 52 | 00:00:25,800 --> 00:00:28,239 53 | question of strategic behavior we asked 54 | 我们问的战略行为问题 55 | 56 | 12 57 | 00:00:28,439 --> 00:00:29,769 58 | about how to aggregate different 59 | 关于如何汇总不同 60 | 61 | 13 62 | 00:00:29,969 --> 00:00:31,510 63 | people's preferences together in a 64 | 人们的喜好在一起 65 | 66 | 14 67 | 00:00:31,710 --> 00:00:35,409 68 | sensible way I'm Kevin Layton Brown I'm 69 | 我是明智的方式，我是Kevin Layton Brown，我是 70 | 71 | 15 72 | 00:00:35,609 --> 00:00:36,839 73 | from the University of British Columbia 74 | 来自不列颠哥伦比亚大学 75 | 76 | 16 77 | 00:00:37,039 --> 00:00:39,038 78 | my co instructors who you'll meet in 79 | 我将与您见面的共同指导者 80 | 81 | 17 82 | 00:00:39,238 --> 00:00:41,378 83 | subsequent videos are Matt Jackson and 84 | 随后的视频是Matt Jackson和 85 | 86 | 18 87 | 00:00:41,579 --> 00:00:44,159 88 | Yoav shoham from Stanford University 89 | 斯坦福大学的Yoav Shoham 90 | 91 | 19 92 | 00:00:44,359 --> 00:00:49,000 93 | let's get started so social choice is 94 | 让我们开始吧，所以社会选择是 95 | 96 | 20 97 | 00:00:49,200 --> 00:00:51,809 98 | essentially the problem of voting and 99 | 本质上是投票和 100 | 101 | 21 102 | 00:00:52,009 --> 00:00:54,128 103 | the first thing that you might think is 104 | 您可能会想到的第一件事是 105 | 106 | 22 107 | 00:00:54,329 --> 00:00:55,989 108 | that voting seems like an incredibly 109 | 投票似乎令人难以置信 110 | 111 | 23 112 | 00:00:56,189 --> 00:00:58,570 113 | simple thing to study how hard can it be 114 | 研究它有多难的简单事情 115 | 116 | 24 117 | 00:00:58,770 --> 00:01:01,239 118 | to study something that after all 119 | 毕竟学习一些东西 120 | 121 | 25 122 | 00:01:01,439 --> 00:01:03,969 123 | involves just filling in one of the 124 | 涉及只是填写其中之一 125 | 126 | 26 127 | 00:01:04,170 --> 00:01:07,569 128 | boxes out of some small set of different 129 | 一小盒不同的盒子 130 | 131 | 27 132 | 00:01:07,769 --> 00:01:11,709 133 | candidates well I want to argue that 134 | 候选人很好我想说 135 | 136 | 28 137 | 00:01:11,909 --> 00:01:14,079 138 | that voting actually is tricky enough 139 | 投票实际上很棘手 140 | 141 | 29 142 | 00:01:14,280 --> 00:01:16,209 143 | that we want to have a mathematical 144 | 我们想要数学 145 | 146 | 30 147 | 00:01:16,409 --> 00:01:20,469 148 | theory of it here's an example in the 149 | 理论，这是一个例子 150 | 151 | 31 152 | 00:01:20,670 --> 00:01:23,319 153 | year 2000 the US presidential election 154 | 2000年美国总统大选 155 | 156 | 32 157 | 00:01:23,519 --> 00:01:27,609 158 | came down to who won Florida and it was 159 | 归结于谁赢得了佛罗里达，那是 160 | 161 | 33 162 | 00:01:27,810 --> 00:01:30,459 163 | incredibly close after complicated 164 | 复杂之后难以置信的关闭 165 | 166 | 34 167 | 00:01:30,659 --> 00:01:32,019 168 | recounts and an appeal to the Supreme 169 | 重述并向最高​​法院上诉 170 | 171 | 35 172 | 00:01:32,219 --> 00:01:34,299 173 | Court George Bush turned out to be the 174 | 乔治·布什法院原来是 175 | 176 | 36 177 | 00:01:34,500 --> 00:01:36,840 178 | winner by 537 votes 179 | 537票冠军 180 | 181 | 37 182 | 00:01:37,040 --> 00:01:40,659 183 | however Ralph Nader a third party 184 | 但是拉尔夫·纳德（Ralph Nader）是第三方 185 | 186 | 38 187 | 00:01:40,859 --> 00:01:42,250 188 | candidate with a pretty limited amount 189 | 数量有限的候选人 190 | 191 | 39 192 | 00:01:42,450 --> 00:01:45,808 193 | of support got 97 thousand votes in 194 | 的支持者中有9.7万票 195 | 196 | 40 197 | 00:01:46,009 --> 00:01:48,429 198 | surveys that were taken after the 199 | 在 200 | 201 | 41 202 | 00:01:48,629 --> 00:01:51,519 203 | election it turned out that about twice 204 | 选举结果表明，大约两次 205 | 206 | 42 207 | 00:01:51,719 --> 00:01:54,668 208 | as many Nader voters would have chosen 209 | 正如许多纳德选民所选择的 210 | 211 | 43 212 | 00:01:54,868 --> 00:01:57,399 213 | Al Gore instead of George Bush if Nader 214 | 如果纳德（Alder），则用戈尔（Al Gore）代替乔治·布什（George Bush） 215 | 216 | 44 217 | 00:01:57,599 --> 00:01:59,219 218 | hadn't been a candidate in the election 219 | 还没参加选举 220 | 221 | 45 222 | 00:01:59,420 --> 00:02:01,390 223 | now leaving the site whatever 224 | 现在无论离开网站 225 | 226 | 46 227 | 00:02:01,590 --> 00:02:03,099 228 | preferences we might have about who 229 | 我们可能对谁有偏好 230 | 231 | 47 232 | 00:02:03,299 --> 00:02:04,299 233 | should be President of the United States 234 | 应该是美国总统 235 | 236 | 48 237 | 00:02:04,500 --> 00:02:07,569 238 | this seems like a bad outcome because it 239 | 这似乎是一个不好的结果，因为它 240 | 241 | 49 242 | 00:02:07,769 --> 00:02:11,020 243 | seems like more people overall preferred 244 | 似乎总体上有更多人偏爱 245 | 246 | 50 247 | 00:02:11,219 --> 00:02:11,868 248 | Gore 249 | 戈尔 250 | 251 | 51 252 | 00:02:12,068 --> 00:02:14,599 253 | bush in Florida and nevertheless Bush 254 | 布什在佛罗里达州，但布什 255 | 256 | 52 257 | 00:02:14,799 --> 00:02:16,700 258 | was the winner so it seems like we 259 | 是赢家，所以好像我们 260 | 261 | 53 262 | 00:02:16,900 --> 00:02:18,049 263 | didn't do a very good job in this 264 | 在这方面做得不好 265 | 266 | 54 267 | 00:02:18,248 --> 00:02:19,640 268 | election about taking everybody's 269 | 选择每个人的选举 270 | 271 | 55 272 | 00:02:19,840 --> 00:02:21,409 273 | preferences into account and selecting 274 | 偏好设置并选择 275 | 276 | 56 277 | 00:02:21,609 --> 00:02:23,200 278 | the right winner 279 | 正确的赢家 280 | 281 | 57 282 | 00:02:23,400 --> 00:02:26,450 283 | indeed recently there have been popular 284 | 确实最近有流行 285 | 286 | 58 287 | 00:02:26,650 --> 00:02:28,969 288 | movements around the world demanding new 289 | 要求新的世界运动 290 | 291 | 59 292 | 00:02:29,169 --> 00:02:32,269 293 | voting systems here I'm showing protests 294 | 我在这里投票制度示威 295 | 296 | 60 297 | 00:02:32,468 --> 00:02:35,840 298 | in London and in Ottawa by British and 299 | 在伦敦和渥太华由英国人和 300 | 301 | 61 302 | 00:02:36,039 --> 00:02:37,399 303 | Canadian citizens who think that they 304 | 认为自己的加拿大公民 305 | 306 | 62 307 | 00:02:37,598 --> 00:02:39,319 308 | should have new voting systems that do a 309 | 应该有新的投票系统 310 | 311 | 63 312 | 00:02:39,519 --> 00:02:41,360 313 | better job of taking their whole 314 | 做好全部工作 315 | 316 | 64 317 | 00:02:41,560 --> 00:02:45,110 318 | preferences into account in this week of 319 | 本周的偏好设置 320 | 321 | 65 322 | 00:02:45,310 --> 00:02:47,090 323 | the course we're going to ask whether 324 | 当然，我们要问是否 325 | 326 | 66 327 | 00:02:47,289 --> 00:02:49,039 328 | it's really true that there are other 329 | 确实有其他 330 | 331 | 67 332 | 00:02:49,239 --> 00:02:51,800 333 | voting systems that are better and what 334 | 更好的投票系统以及什么 335 | 336 | 68 337 | 00:02:52,000 --> 00:02:53,689 338 | better actually means so we're gonna 339 | 更好的是意味着我们要 340 | 341 | 69 342 | 00:02:53,889 --> 00:02:55,340 343 | start by looking at different voting 344 | 首先看不同的投票 345 | 346 | 70 347 | 00:02:55,539 --> 00:02:58,789 348 | schemes seeing how they work we're gonna 349 | 计划，看看他们如何工作，我们要 350 | 351 | 71 352 | 00:02:58,989 --> 00:03:00,920 353 | see that each of these voting schemes is 354 | 看到每个投票方案是 355 | 356 | 72 357 | 00:03:01,120 --> 00:03:03,200 358 | sometimes able to give rise to kind of 359 | 有时能够引起 360 | 361 | 73 362 | 00:03:03,400 --> 00:03:05,840 363 | paradoxical outcomes where it seems like 364 | 看起来很矛盾的结果 365 | 366 | 74 367 | 00:03:06,039 --> 00:03:07,759 368 | something wrong can sometimes happen and 369 | 有时可能会发生错误， 370 | 371 | 75 372 | 00:03:07,959 --> 00:03:10,640 373 | we're gonna culminate in a famous proof 374 | 我们将以著名的证明达到高潮 375 | 376 | 76 377 | 00:03:10,840 --> 00:03:13,069 378 | that shows that in a really formal 379 | 这表明在一个真正正式的 380 | 381 | 77 382 | 00:03:13,269 --> 00:03:15,200 383 | mathematical sense it just isn't 384 | 从数学意义上讲不是 385 | 386 | 78 387 | 00:03:15,400 --> 00:03:16,879 388 | possible to avoid these kinds of 389 | 可能避免这种情况 390 | 391 | 79 392 | 00:03:17,079 --> 00:03:20,149 393 | failures so thanks for joining our 394 | 失败，所以感谢您加入我们 395 | 396 | 80 397 | 00:03:20,348 --> 00:03:22,670 398 | course and look forward to having you 399 | 当然，期待有你 400 | 401 | 81 402 | 00:03:22,870 --> 00:03:27,870 403 | for the videos to come 404 | 让视频来 405 | 406 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 1/GTO2-1-04 - Social Choice - Impossibility of Non-Paradoxical Social Welfare Functions.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,000 --> 00:00:02,738 3 | this video is going to tell you at a 4 | 该视频将在 5 | 6 | 2 7 | 00:00:02,939 --> 00:00:06,609 8 | high level probably the most important 9 | 高水平可能是最重要的 10 | 11 | 3 12 | 00:00:06,809 --> 00:00:08,020 13 | and surprising thing that we have to 14 | 和我们必须要做的令人惊讶的事情 15 | 16 | 4 17 | 00:00:08,220 --> 00:00:10,569 18 | tell you about social choice which is 19 | 告诉你关于社会选择的 20 | 21 | 5 22 | 00:00:10,769 --> 00:00:14,740 23 | that it's impossible to come up with any 24 | 不可能提出任何 25 | 26 | 6 27 | 00:00:14,939 --> 00:00:16,659 28 | social welfare function which is 29 | 社会福利功能是 30 | 31 | 7 32 | 00:00:16,859 --> 00:00:20,499 33 | sensible in a particular formal way so 34 | 以一种特殊的正式方式明智 35 | 36 | 8 37 | 00:00:20,699 --> 00:00:22,120 38 | first of all let's remember what a 39 | 首先，让我们记住 40 | 41 | 9 42 | 00:00:22,320 --> 00:00:24,609 43 | social welfare function is a social 44 | 社会福利功能是一种社会 45 | 46 | 10 47 | 00:00:24,809 --> 00:00:27,609 48 | welfare function is a voting rule where 49 | 福利功能是一种投票规则 50 | 51 | 11 52 | 00:00:27,809 --> 00:00:30,399 53 | every agent tells me a set of 54 | 每个经纪人都告诉我一套 55 | 56 | 12 57 | 00:00:30,599 --> 00:00:33,250 58 | preferences so gives me a total ordering 59 | 偏好，所以给了我一个整体的顺序 60 | 61 | 13 62 | 00:00:33,450 --> 00:00:35,829 63 | over all of the choices that we have to 64 | 在我们必须要做的所有选择中 65 | 66 | 14 67 | 00:00:36,030 --> 00:00:38,878 68 | make over all of the candidates and 69 | 弥补所有候选人 70 | 71 | 15 72 | 00:00:39,079 --> 00:00:40,869 73 | getting this whole set of different 74 | 得到这整套不同的 75 | 76 | 16 77 | 00:00:41,070 --> 00:00:42,459 78 | rankings from all of the different 79 | 来自所有不同的排名 80 | 81 | 17 82 | 00:00:42,659 --> 00:00:44,349 83 | agents the Social Welfare function gives 84 | 代理商的社会福利功能 85 | 86 | 18 87 | 00:00:44,549 --> 00:00:47,378 88 | me back a single ranking which is the 89 | 我回到一个单一的排名是 90 | 91 | 19 92 | 00:00:47,579 --> 00:00:51,428 93 | social choice so so think about voting 94 | 社会选择，所以考虑投票 95 | 96 | 20 97 | 00:00:51,628 --> 00:00:53,498 98 | over a set of candidates everyone has 99 | 每个人都有一套候选人 100 | 101 | 21 102 | 00:00:53,698 --> 00:00:54,998 103 | preferences over all the candidates that 104 | 对所有候选人的偏好 105 | 106 | 22 107 | 00:00:55,198 --> 00:00:56,858 108 | we're going to vote over and getting 109 | 我们将投票并获得 110 | 111 | 23 112 | 00:00:57,058 --> 00:00:59,559 113 | back a single ranking but that is the 114 | 返回单个排名，但这是 115 | 116 | 24 117 | 00:00:59,759 --> 00:01:00,969 118 | kind of aggregate ranking for all the 119 | 所有类别的总排名 120 | 121 | 25 122 | 00:01:01,170 --> 00:01:03,788 123 | agents so what would we want such a 124 | 代理商，那么我们想要这样一个 125 | 126 | 26 127 | 00:01:03,988 --> 00:01:05,469 128 | ranking to do what would it look like 129 | 排名做什么样 130 | 131 | 27 132 | 00:01:05,670 --> 00:01:08,799 133 | for such a ranking to do a good job well 134 | 这样的排名才能做好 135 | 136 | 28 137 | 00:01:09,000 --> 00:01:10,299 138 | here's one thing that seems pretty 139 | 这是一件看起来很漂亮的东西 140 | 141 | 29 142 | 00:01:10,500 --> 00:01:11,918 143 | obvious that such a ranking function 144 | 显然，这种排名功能 145 | 146 | 30 147 | 00:01:12,118 --> 00:01:15,159 148 | should do well say that a ranking for a 149 | 应该说一个 150 | 151 | 31 152 | 00:01:15,359 --> 00:01:17,619 153 | social welfare function W is purrito 154 | 社会福利函数W是purrito 155 | 156 | 32 157 | 00:01:17,819 --> 00:01:21,129 158 | efficient if any time it's the case that 159 | 有效率的话 160 | 161 | 33 162 | 00:01:21,329 --> 00:01:23,500 163 | all of the agents agree on the ordering 164 | 所有代理商都同意订购 165 | 166 | 34 167 | 00:01:23,700 --> 00:01:26,109 168 | of two outcomes if everyone thinks that 169 | 如果每个人都认为这两个结果 170 | 171 | 35 172 | 00:01:26,310 --> 00:01:28,450 173 | candidate a is better than candidate B 174 | 候选人a胜过候选人B 175 | 176 | 36 177 | 00:01:28,650 --> 00:01:31,590 178 | then the social welfare function also 179 | 然后社会福利功能 180 | 181 | 37 182 | 00:01:31,790 --> 00:01:33,869 183 | agrees about that part of the ordering 184 | 同意订购的那部分 185 | 186 | 38 187 | 00:01:34,069 --> 00:01:36,219 188 | so it might it might say all kinds of 189 | 所以它可能会说各种 190 | 191 | 39 192 | 00:01:36,420 --> 00:01:38,469 193 | other things about candidates C D and E 194 | 关于候选CD和E的其他内容 195 | 196 | 40 197 | 00:01:38,670 --> 00:01:40,808 198 | but if everybody thinks that a is better 199 | 但是如果每个人都认为a更好 200 | 201 | 41 202 | 00:01:41,009 --> 00:01:42,878 203 | than B then the social welfare function 204 | 比B那么社会福利功能 205 | 206 | 42 207 | 00:01:43,078 --> 00:01:44,799 208 | also is going to tell us that a is 209 | 也要告诉我们a是 210 | 211 | 43 212 | 00:01:45,000 --> 00:01:46,750 213 | better than B it's gonna put a somewhere 214 | 比B还好 215 | 216 | 44 217 | 00:01:46,950 --> 00:01:49,299 218 | higher in the ranking than B so and 219 | 排名高于B，因此 220 | 221 | 45 222 | 00:01:49,500 --> 00:01:51,219 223 | that's all we're requiring by Pareto 224 | 这就是帕累托所需要的 225 | 226 | 46 227 | 00:01:51,420 --> 00:01:53,918 228 | efficiency we're just saying only in the 229 | 效率，我们只是在说 230 | 231 | 47 232 | 00:01:54,118 --> 00:01:56,049 233 | case when everybody agrees absolutely 234 | 每个人都绝对同意的情况 235 | 236 | 48 237 | 00:01:56,250 --> 00:01:57,789 238 | everybody agrees the social welfare 239 | 每个人都同意社会福利 240 | 241 | 49 242 | 00:01:57,989 --> 00:02:00,518 243 | function has to rank in the appropriate 244 | 功能必须排在适当的位置 245 | 246 | 50 247 | 00:02:00,718 --> 00:02:05,319 248 | way a second thing we might want is what 249 | 我们可能想要的第二件事是 250 | 251 | 51 252 | 00:02:05,519 --> 00:02:06,698 253 | we call independence of irrelevant 254 | 我们称无关的独立性 255 | 256 | 52 257 | 00:02:06,899 --> 00:02:10,149 258 | alternatives so we'll say that a social 259 | 替代方案，所以我们说 260 | 261 | 53 262 | 00:02:10,348 --> 00:02:11,920 263 | welfare function is independent of 264 | 福利功能独立于 265 | 266 | 54 267 | 00:02:12,120 --> 00:02:13,280 268 | irrelevant alternatives 269 | 不相关的选择 270 | 271 | 55 272 | 00:02:13,479 --> 00:02:16,759 273 | if the way that I rank between a and B 274 | 如果我在A和B之间的排名方式 275 | 276 | 56 277 | 00:02:16,959 --> 00:02:19,189 278 | only depends on the way that all of the 279 | 只取决于所有 280 | 281 | 57 282 | 00:02:19,389 --> 00:02:22,490 283 | individual agents ranked a and B so if I 284 | 个体代理商排名a和B，所以如果我 285 | 286 | 58 287 | 00:02:22,689 --> 00:02:24,469 288 | want to decide in the social ordering 289 | 想决定社会秩序 290 | 291 | 59 292 | 00:02:24,669 --> 00:02:27,890 293 | how it is that that I compare a to B I 294 | 我将BI与BI进行比较的方式 295 | 296 | 60 297 | 00:02:28,090 --> 00:02:30,679 298 | should do that only by looking at how 299 | 应该只看如何 300 | 301 | 61 302 | 00:02:30,878 --> 00:02:33,950 303 | all of the individual agents relatively 304 | 所有个体代理相对 305 | 306 | 62 307 | 00:02:34,150 --> 00:02:36,770 308 | ranked a to B so I shouldn't care about 309 | 将A排名为B，所以我不在乎 310 | 311 | 63 312 | 00:02:36,969 --> 00:02:39,649 313 | how everyone felt about a and C or about 314 | 每个人对a和C或大概的感觉 315 | 316 | 64 317 | 00:02:39,848 --> 00:02:42,050 318 | B and D I should just look at the a B 319 | B和DI应该只看a B 320 | 321 | 65 322 | 00:02:42,250 --> 00:02:43,580 323 | rankings for everyone and that should 324 | 每个人的排名，应该 325 | 326 | 66 327 | 00:02:43,780 --> 00:02:45,830 328 | tell me how the social welfare function 329 | 告诉我社会福利的功能 330 | 331 | 67 332 | 00:02:46,030 --> 00:02:51,800 333 | ranks between a and B finally I'll say 334 | 最终在a和B之间排名 335 | 336 | 68 337 | 00:02:52,000 --> 00:02:53,480 338 | that a social welfare function has a 339 | 社会福利功能具有 340 | 341 | 69 342 | 00:02:53,680 --> 00:02:56,860 343 | dictator if there exists a single agent 344 | 独裁者是否存在单个代理 345 | 346 | 70 347 | 00:02:57,060 --> 00:02:59,899 348 | whose preferences we always just return 349 | 我们总是会返回谁的偏好 350 | 351 | 71 352 | 00:03:00,098 --> 00:03:02,689 353 | as the social ordering so in other words 354 | 换句话说就是社会秩序 355 | 356 | 72 357 | 00:03:02,889 --> 00:03:04,399 358 | if I've got n different agents and they 359 | 如果我有n个不同的代理商，他们 360 | 361 | 73 362 | 00:03:04,598 --> 00:03:06,560 363 | always give me some kind of order in and 364 | 总是给我某种命令 365 | 366 | 74 367 | 00:03:06,759 --> 00:03:09,349 368 | I just say you know what I only care 369 | 我只是说你知道我只在乎 370 | 371 | 75 372 | 00:03:09,549 --> 00:03:11,629 373 | about agent 3 everybody else it just 374 | 关于特工3其他所有人 375 | 376 | 76 377 | 00:03:11,829 --> 00:03:13,129 378 | doesn't matter what you tell me I'm 379 | 你告诉我我没关系 380 | 381 | 77 382 | 00:03:13,329 --> 00:03:14,509 383 | gonna throw it in the garbage I'm not 384 | 要把它扔进垃圾桶，我不是 385 | 386 | 78 387 | 00:03:14,709 --> 00:03:15,980 388 | gonna think about it at all I'm just 389 | 我会想一想 390 | 391 | 79 392 | 00:03:16,180 --> 00:03:18,020 393 | gonna give you back agent 3 and I've 394 | 会给你特工3，我已经 395 | 396 | 80 397 | 00:03:18,219 --> 00:03:20,360 398 | decided in advance who agent 3 is it's 399 | 事先确定谁是代理商3 400 | 401 | 81 402 | 00:03:20,560 --> 00:03:22,490 403 | always gonna be the same guy that's what 404 | 永远都是同一个人，这就是 405 | 406 | 82 407 | 00:03:22,689 --> 00:03:23,599 408 | it means for a social welfare function 409 | 对社会福利功能意味着 410 | 411 | 83 412 | 00:03:23,799 --> 00:03:26,300 413 | to have a dictator now it's kind of 414 | 现在有一个独裁者 415 | 416 | 84 417 | 00:03:26,500 --> 00:03:27,710 418 | terrible for a social welfare function 419 | 糟糕的社会福利功能 420 | 421 | 85 422 | 00:03:27,909 --> 00:03:29,840 423 | to have a dictator because it means that 424 | 有一个独裁者，因为这意味着 425 | 426 | 86 427 | 00:03:30,039 --> 00:03:31,399 428 | it's not aggregating in any meaningful 429 | 它没有聚集在任何有意义的地方 430 | 431 | 87 432 | 00:03:31,598 --> 00:03:34,069 433 | way it's just always reporting back what 434 | 只是总是报告什么 435 | 436 | 88 437 | 00:03:34,269 --> 00:03:37,759 438 | one person thanks notice that a social 439 | 一个人感谢通知， 440 | 441 | 89 442 | 00:03:37,959 --> 00:03:39,080 443 | welfare function that does have a 444 | 福利功能确实具有 445 | 446 | 90 447 | 00:03:39,280 --> 00:03:41,149 448 | dictator trivially satisfies 449 | 独裁者平凡地满足 450 | 451 | 91 452 | 00:03:41,348 --> 00:03:42,469 453 | independence of irrelevant alternatives 454 | 不相关选择的独立性 455 | 456 | 92 457 | 00:03:42,669 --> 00:03:46,039 458 | and Pareto efficiency because prudhoe 459 | 和帕累托效率，因为普拉德霍 460 | 461 | 93 462 | 00:03:46,239 --> 00:03:48,618 463 | efficiency says anytime everybody agrees 464 | 效率表示任何人都同意 465 | 466 | 94 467 | 00:03:48,818 --> 00:03:50,118 468 | the social welfare function has to do 469 | 社会福利职能必须要做 470 | 471 | 95 472 | 00:03:50,318 --> 00:03:52,368 473 | the right thing and when everybody 474 | 正确的事情，当每个人 475 | 476 | 96 477 | 00:03:52,568 --> 00:03:54,439 478 | agrees in particular the dictator also 479 | 特别同意独裁者 480 | 481 | 97 482 | 00:03:54,639 --> 00:03:56,780 483 | agreed so so we would get free - 484 | 同意，所以我们可以免费- 485 | 486 | 98 487 | 00:03:56,979 --> 00:03:59,868 488 | efficiency for free likewise we would 489 | 免费的效率我们同样会 490 | 491 | 99 492 | 00:04:00,068 --> 00:04:00,800 493 | get independence of irrelevant 494 | 获得无关的独立性 495 | 496 | 100 497 | 00:04:01,000 --> 00:04:03,560 498 | alternatives for free because we're only 499 | 免费的替代品，因为我们只是 500 | 501 | 101 502 | 00:04:03,759 --> 00:04:05,749 503 | gonna care about the dictators a B 504 | 要关心独裁者A B 505 | 506 | 102 507 | 00:04:05,949 --> 00:04:08,420 508 | rankings in order to decide the a B 509 | 排名以决定A B 510 | 511 | 103 512 | 00:04:08,620 --> 00:04:10,429 513 | ranking and and so we're not looking at 514 | 排名，所以我们不在看 515 | 516 | 104 517 | 00:04:10,628 --> 00:04:11,960 518 | anything else from any of the other 519 | 其他任何东西 520 | 521 | 105 522 | 00:04:12,159 --> 00:04:14,780 523 | agents so so we see we can satisfy our 524 | 代理商，所以我们看到我们可以满足我们的要求 525 | 526 | 106 527 | 00:04:14,979 --> 00:04:17,629 528 | two properties from a dictatorial social 529 | 独裁社会的两个属性 530 | 531 | 107 532 | 00:04:17,829 --> 00:04:20,180 533 | welfare function but dictatorial social 534 | 福利功能但独裁社会 535 | 536 | 108 537 | 00:04:20,379 --> 00:04:21,860 538 | welfare functions are pretty undesirable 539 | 福利功能非常不可取 540 | 541 | 109 542 | 00:04:22,060 --> 00:04:25,620 543 | to us well here's the punchline 544 | 对我们来说这是重点 545 | 546 | 110 547 | 00:04:25,819 --> 00:04:29,220 548 | Kenneth arrow in 1951 famously proved 549 | 1951年的肯尼斯·阿斯（Kenneth Arrow）著名 550 | 551 | 111 552 | 00:04:29,420 --> 00:04:31,889 553 | the following theorem any social welfare 554 | 以下定理任何社会福利 555 | 556 | 112 557 | 00:04:32,089 --> 00:04:34,829 558 | function over three or more outcomes so 559 | 作用于三个或更多结果，因此 560 | 561 | 113 562 | 00:04:35,029 --> 00:04:37,079 563 | candidates in elections that have more 564 | 选举中候选人更多 565 | 566 | 114 567 | 00:04:37,279 --> 00:04:40,020 568 | than two candidates that is pre to 569 | 比两名候选人 570 | 571 | 115 572 | 00:04:40,220 --> 00:04:42,090 573 | efficient and independent of irrelevant 574 | 高效且无关紧要 575 | 576 | 116 577 | 00:04:42,290 --> 00:04:45,780 578 | alternatives must be dictatorial so in 579 | 替代品必须是独裁的，因此在 580 | 581 | 117 582 | 00:04:45,980 --> 00:04:47,819 583 | other words these these two things that 584 | 换句话说，这两件事 585 | 586 | 118 587 | 00:04:48,019 --> 00:04:49,590 588 | that seem very natural that we would 589 | 看起来我们很自然 590 | 591 | 119 592 | 00:04:49,790 --> 00:04:52,079 593 | want free to efficiency and independence 594 | 想要自由高效和独立 595 | 596 | 120 597 | 00:04:52,279 --> 00:04:54,240 598 | of irrelevant alternatives are so 599 | 无关的替代品是这样 600 | 601 | 121 602 | 00:04:54,439 --> 00:04:55,740 603 | constraining on the set of social 604 | 约束社会 605 | 606 | 122 607 | 00:04:55,939 --> 00:04:57,569 608 | welfare functions that they imply that 609 | 他们暗示的福利功能 610 | 611 | 123 612 | 00:04:57,769 --> 00:04:58,949 613 | we would have to have a dictatorial 614 | 我们将不得不独裁 615 | 616 | 124 617 | 00:04:59,149 --> 00:05:02,100 618 | social welfare function since we can't 619 | 社会福利功能，因为我们不能 620 | 621 | 125 622 | 00:05:02,300 --> 00:05:04,170 623 | possibly live with a dictatorial social 624 | 可能与独裁者生活在一起 625 | 626 | 126 627 | 00:05:04,370 --> 00:05:06,150 628 | welfare function what that really means 629 | 福利功能的真正含义 630 | 631 | 127 632 | 00:05:06,350 --> 00:05:08,460 633 | is that we have to give up either pre 634 | 是我们必须放弃 635 | 636 | 128 637 | 00:05:08,660 --> 00:05:10,710 638 | doe efficiency or independence of 639 | 能源效率或独立性 640 | 641 | 129 642 | 00:05:10,910 --> 00:05:12,810 643 | irrelevant alternatives and that's 644 | 无关的选择，那就是 645 | 646 | 130 647 | 00:05:13,009 --> 00:05:14,100 648 | pretty bad news because these are both 649 | 坏消息是因为这两个都是 650 | 651 | 131 652 | 00:05:14,300 --> 00:05:19,300 653 | pretty natural things 654 | 很自然的东西 655 | 656 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 1/GTO2-1-06 - Impossible of Non-paradoxical Social Choice Functions.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,060 --> 00:00:01,929 3 | in this lecture we'll speak about social 4 | 在本讲座中，我们将谈论社交 5 | 6 | 2 7 | 00:00:02,129 --> 00:00:04,750 8 | choice functions the most famous 9 | 选择功能最著名 10 | 11 | 3 12 | 00:00:04,950 --> 00:00:08,229 13 | impossibility results pertain to social 14 | 不可能结果与社会有关 15 | 16 | 4 17 | 00:00:08,429 --> 00:00:10,630 18 | welfare function and in particular the 19 | 福利功能，尤其是 20 | 21 | 5 22 | 00:00:10,830 --> 00:00:14,140 23 | arrows impossibility theorem is widely 24 | 箭头不可能定理广泛 25 | 26 | 6 27 | 00:00:14,339 --> 00:00:17,220 28 | known and we have discussed that before 29 | 已知，我们之前已经讨论过 30 | 31 | 7 32 | 00:00:17,420 --> 00:00:22,690 33 | now it might be thought that the problem 34 | 现在可能会认为问题出在 35 | 36 | 8 37 | 00:00:22,890 --> 00:00:24,699 38 | lies in the fact that social welfare 39 | 在于社会福利 40 | 41 | 9 42 | 00:00:24,899 --> 00:00:27,879 43 | functions require you to specify as an 44 | 函数需要您指定为 45 | 46 | 10 47 | 00:00:28,079 --> 00:00:29,949 48 | output of the process and entire 49 | 过程和整个过程的输出 50 | 51 | 11 52 | 00:00:30,149 --> 00:00:32,500 53 | ordering and that might be highly 54 | 订购，这可能是高度 55 | 56 | 12 57 | 00:00:32,700 --> 00:00:34,299 58 | constraining but if you only need it to 59 | 约束，但是如果您只需要它 60 | 61 | 13 62 | 00:00:34,500 --> 00:00:37,570 63 | pick a winner as does a social choice 64 | 选择优胜者，社会选择也一样 65 | 66 | 14 67 | 00:00:37,770 --> 00:00:40,469 68 | function then you would escape these 69 | 功能，那么您将逃脱这些 70 | 71 | 15 72 | 00:00:40,670 --> 00:00:44,858 73 | paradoxes it turns out that the answer 74 | 矛盾的是，答案 75 | 76 | 16 77 | 00:00:45,058 --> 00:00:47,108 78 | is no but first we need to sort of 79 | 不是，但是首先我们需要 80 | 81 | 17 82 | 00:00:47,308 --> 00:00:49,178 83 | redefine our criteria a little bit 84 | 重新定义我们的标准 85 | 86 | 18 87 | 00:00:49,378 --> 00:00:52,928 88 | because the notions of Pareto efficiency 89 | 因为帕累托效率的概念 90 | 91 | 19 92 | 00:00:53,128 --> 00:00:56,259 93 | and independence of irrelevant 94 | 与无关无关 95 | 96 | 20 97 | 00:00:56,460 --> 00:00:58,959 98 | alternatives just aren't well-defined in 99 | 替代品的定义不明确 100 | 101 | 21 102 | 00:00:59,159 --> 00:01:02,248 103 | the context of a social choice function 104 | 社会选择功能的背景 105 | 106 | 22 107 | 00:01:02,448 --> 00:01:04,810 108 | but we will see that there are closely 109 | 但我们会发现 110 | 111 | 23 112 | 00:01:05,010 --> 00:01:08,619 113 | related notions that are well defined so 114 | 定义明确的相关概念 115 | 116 | 24 117 | 00:01:08,819 --> 00:01:10,988 118 | let's first define weak Pareto 119 | 首先定义弱帕累托 120 | 121 | 25 122 | 00:01:11,188 --> 00:01:11,649 123 | efficiency 124 | 效率 125 | 126 | 26 127 | 00:01:11,849 --> 00:01:14,230 128 | and we'll say that a software choice 129 | 我们会说软件选择 130 | 131 | 27 132 | 00:01:14,430 --> 00:01:18,209 133 | function C is weakly Pareto efficient if 134 | 函数C弱于帕累托效率 135 | 136 | 28 137 | 00:01:18,409 --> 00:01:22,799 138 | essentially it never elects a dominated 139 | 从根本上说，它从不选举一个被统治者 140 | 141 | 29 142 | 00:01:23,000 --> 00:01:31,649 143 | outcome or candidate so if there is a a 144 | 结果或候选人，如果有的话 145 | 146 | 30 147 | 00:01:31,849 --> 00:01:35,738 148 | candidate o - such that there is some 149 | 候选o-这样有一些 150 | 151 | 31 152 | 00:01:35,938 --> 00:01:38,918 153 | other candidate or one that is always 154 | 其他候选人或永远是 155 | 156 | 32 157 | 00:01:39,118 --> 00:01:42,128 158 | preferred to Oh - by every by every 159 | 喜欢哦-每个人一个 160 | 161 | 33 162 | 00:01:42,328 --> 00:01:45,250 163 | voter then the weaker candidate Oh - 164 | 选民再弱者候选人哦- 165 | 166 | 34 167 | 00:01:45,450 --> 00:01:47,439 168 | would never be elected seems like a 169 | 永远不会当选似乎 170 | 171 | 35 172 | 00:01:47,640 --> 00:01:49,448 173 | reasonable criterion and we'll call that 174 | 合理的标准，我们称之为 175 | 176 | 36 177 | 00:01:49,649 --> 00:01:55,859 178 | weak Pareto efficiency in place of 179 | 帕累托效率低下 180 | 181 | 37 182 | 00:01:56,060 --> 00:01:57,939 183 | independent of irrelevant alternative 184 | 独立于无关的选择 185 | 186 | 38 187 | 00:01:58,140 --> 00:01:59,859 188 | the alternative will have the notion of 189 | 替代方案将具有 190 | 191 | 39 192 | 00:02:00,060 --> 00:02:03,340 193 | monotonicity and informally as I written 194 | 单调性和我写的非正式 195 | 196 | 40 197 | 00:02:03,540 --> 00:02:06,849 198 | here at the at the bottom it says that 199 | 在底部的底部说 200 | 201 | 41 202 | 00:02:07,049 --> 00:02:12,160 203 | if you have a winner if we increase the 204 | 如果您有赢家，如果我们增加 205 | 206 | 42 207 | 00:02:12,360 --> 00:02:12,939 208 | support 209 | 支持 210 | 211 | 43 212 | 00:02:13,139 --> 00:02:16,540 213 | for that candidate they would still 214 | 对于那个候选人，他们仍然会 215 | 216 | 44 217 | 00:02:16,740 --> 00:02:19,830 218 | remain a winner so formally speaking 219 | 如此正式地保持赢家 220 | 221 | 45 222 | 00:02:20,030 --> 00:02:22,870 223 | will say that the subject response is 224 | 会说主题回应是 225 | 226 | 46 227 | 00:02:23,069 --> 00:02:27,700 228 | monotonic if we if we take any candidate 229 | 如果我们接受任何候选人，则单调 230 | 231 | 47 232 | 00:02:27,900 --> 00:02:32,439 233 | oh and an infestation for any preference 234 | 哦，还有什么偏爱 235 | 236 | 48 237 | 00:02:32,639 --> 00:02:33,219 238 | profile 239 | 轮廓 240 | 241 | 49 242 | 00:02:33,419 --> 00:02:37,210 243 | this one over here if under this 244 | 如果在这个下面 245 | 246 | 50 247 | 00:02:37,409 --> 00:02:41,830 248 | preference profile always selected then 249 | 然后总是选择首选项配置文件 250 | 251 | 51 252 | 00:02:42,030 --> 00:02:44,080 253 | if we look at any other preference 254 | 如果我们看看其他任何偏好 255 | 256 | 52 257 | 00:02:44,280 --> 00:02:48,040 258 | profile if it has the property that for 259 | 配置文件，如果它具有 260 | 261 | 53 262 | 00:02:48,240 --> 00:02:51,310 263 | every agent and every other outcome Oh 264 | 每一个特工和每一个其他的结果哦 265 | 266 | 54 267 | 00:02:51,509 --> 00:02:56,860 268 | prime if under the original preference 269 | 如果在原始偏好下为素数 270 | 271 | 55 272 | 00:02:57,060 --> 00:03:00,730 273 | Oh was preferred to O prime which is 274 | 哦比O素好 275 | 276 | 56 277 | 00:03:00,930 --> 00:03:05,320 278 | also under the new preference so one the 279 | 在新的偏好下，所以一个 280 | 281 | 57 282 | 00:03:05,520 --> 00:03:08,259 283 | that original winner Oh never lost 284 | 原来的赢家哦，永不丢失 285 | 286 | 58 287 | 00:03:08,459 --> 00:03:11,050 288 | support maybe only gained support then 289 | 支持可能只是获得支持 290 | 291 | 59 292 | 00:03:11,250 --> 00:03:12,789 293 | under those condition it better be the 294 | 在这种情况下，最好是 295 | 296 | 60 297 | 00:03:12,989 --> 00:03:15,039 298 | case that on the new preference order in 299 | 在新的优先顺序上的情况 300 | 301 | 61 302 | 00:03:15,239 --> 00:03:18,490 303 | which Oh only got more support it would 304 | 哦，只会得到更多的支持 305 | 306 | 62 307 | 00:03:18,689 --> 00:03:21,150 308 | still be the so for choice the winner 309 | 仍然是获胜者的选择 310 | 311 | 63 312 | 00:03:21,349 --> 00:03:26,060 313 | again a reasonable property to require 314 | 再次需要合理的财产 315 | 316 | 64 317 | 00:03:27,799 --> 00:03:30,689 318 | the last notion of dictatorship are 319 | 独裁的最后一个概念是 320 | 321 | 65 322 | 00:03:30,889 --> 00:03:34,060 323 | similar to what you see in social 324 | 类似于您在社交中看到的 325 | 326 | 66 327 | 00:03:34,259 --> 00:03:36,099 328 | welfare simply says that C is 329 | 福利只是说C是 330 | 331 | 67 332 | 00:03:36,299 --> 00:03:39,520 333 | dictatorial if there's some agents whose 334 | 是否有一些代理人 335 | 336 | 68 337 | 00:03:39,719 --> 00:03:43,439 338 | top choice is always the social choice 339 | 最高的选择永远是社会的选择 340 | 341 | 69 342 | 00:03:43,639 --> 00:03:45,759 343 | this is the bad news 344 | 这是个坏消息 345 | 346 | 70 347 | 00:03:45,959 --> 00:03:49,390 348 | the molar status weight theorem tells us 349 | 摩尔状态权重定理告诉我们 350 | 351 | 71 352 | 00:03:49,590 --> 00:03:53,590 353 | that we can't have all three so if a 354 | 我们不能同时拥有这三个，所以如果一个 355 | 356 | 72 357 | 00:03:53,789 --> 00:03:54,939 358 | social choice function is Pareto 359 | 社会选择功能是帕累托 360 | 361 | 73 362 | 00:03:55,139 --> 00:03:57,430 363 | efficient and monotonic it must be 364 | 高效且单调的 365 | 366 | 74 367 | 00:03:57,629 --> 00:04:02,590 368 | dictatorial and so after all social 369 | 独裁的，所以毕竟是社会的 370 | 371 | 75 372 | 00:04:02,789 --> 00:04:05,230 373 | choice functions aren't more benign the 374 | 选择功能不是更好 375 | 376 | 76 377 | 00:04:05,430 --> 00:04:09,599 378 | social welfare function and we won't go 379 | 社会福利功能，我们不会去 380 | 381 | 77 382 | 00:04:09,799 --> 00:04:12,849 383 | through the proof but the intuition is 384 | 通过证明，但直觉是 385 | 386 | 78 387 | 00:04:13,049 --> 00:04:18,718 388 | that in order to determine the relative 389 | 为了确定相对 390 | 391 | 79 392 | 00:04:18,918 --> 00:04:21,218 393 | ordering among candidates we need to 394 | 我们需要在候选人中排序 395 | 396 | 80 397 | 00:04:21,418 --> 00:04:25,310 398 | sort of probe it everywhere and 399 | 到处都有探测 400 | 401 | 81 402 | 00:04:26,699 --> 00:04:30,430 403 | and if we probe it enough we'll get the 404 | 如果我们对其进行足够的研究，我们将获得 405 | 406 | 82 407 | 00:04:30,629 --> 00:04:32,139 408 | entire social social welfare function 409 | 整个社会的社会福利职能 410 | 411 | 83 412 | 00:04:32,339 --> 00:04:38,139 413 | and since the since the function social 414 | 自从功能社会以来 415 | 416 | 84 417 | 00:04:38,339 --> 00:04:40,180 418 | choice function must be defined for all 419 | 必须为所有定义选择功能 420 | 421 | 85 422 | 00:04:40,379 --> 00:04:43,420 423 | inputs as we vary the inputs we can find 424 | 输入，因为我们改变了我们可以找到的输入 425 | 426 | 86 427 | 00:04:43,620 --> 00:04:46,449 428 | the total software for ordering so in 429 | 订购总软件 430 | 431 | 87 432 | 00:04:46,649 --> 00:04:48,910 433 | fact or we can use a social choice 434 | 事实，或者我们可以使用社会选择 435 | 436 | 88 437 | 00:04:49,110 --> 00:04:52,300 438 | function to recover the social welfare 439 | 恢复社会福利的功能 440 | 441 | 89 442 | 00:04:52,500 --> 00:04:54,879 443 | fact that the intuition behind the proof 444 | 事实上，证明背后的直觉 445 | 446 | 90 447 | 00:04:55,079 --> 00:05:01,569 448 | and why social drug functions are as may 449 | 以及为什么社会毒品功能如此 450 | 451 | 91 452 | 00:05:01,769 --> 00:05:05,770 453 | be counterintuitive or complicated and 454 | 违反直觉或复杂 455 | 456 | 92 457 | 00:05:05,970 --> 00:05:09,520 458 | social welfare functions now just two 459 | 社会福利职能现在只有两个 460 | 461 | 93 462 | 00:05:09,720 --> 00:05:13,990 463 | tests on tuition again let's consider an 464 | 再次进行学费测试让我们考虑 465 | 466 | 94 467 | 00:05:14,189 --> 00:05:16,449 468 | example and let's take for allottee 469 | 例子，让我们来分配 470 | 471 | 95 472 | 00:05:16,649 --> 00:05:20,280 473 | which perhaps on the face of it might 474 | 也许从表面上看 475 | 476 | 96 477 | 00:05:20,480 --> 00:05:23,050 478 | contradict the malar shadows weight 479 | 与黄斑阴影的重量矛盾 480 | 481 | 97 482 | 00:05:23,250 --> 00:05:26,860 483 | theorem clearly poor allottee is Pareto 484 | 定理显然是可怜的分配者是帕累托 485 | 486 | 98 487 | 00:05:27,060 --> 00:05:29,920 488 | efficient in other words if everybody 489 | 换句话说，如果每个人都有效 490 | 491 | 99 492 | 00:05:30,120 --> 00:05:32,829 493 | prefers some candidate to a another 494 | 偏爱某个候选人 495 | 496 | 100 497 | 00:05:33,029 --> 00:05:34,600 498 | candidate that other weaker candidate 499 | 那个其他较弱的候选人 500 | 501 | 101 502 | 00:05:34,800 --> 00:05:36,790 503 | will never be the choice and it's 504 | 永远不会是选择，这是 505 | 506 | 102 507 | 00:05:36,990 --> 00:05:40,840 508 | founded dictatorial obviously so the 509 | 成立专政显然是这样 510 | 511 | 103 512 | 00:05:41,040 --> 00:05:44,189 513 | theorem says it cannot be monotonic 514 | 定理说它不能单调 515 | 516 | 104 517 | 00:05:44,389 --> 00:05:45,879 518 | intuitive you might think it is 519 | 直观的你可能会认为是 520 | 521 | 105 522 | 00:05:46,079 --> 00:05:48,990 523 | monotonic but here's a counter example 524 | 单调的，但这是一个反例 525 | 526 | 106 527 | 00:05:49,189 --> 00:05:54,939 528 | so here are these seven agents and there 529 | 所以这是这七个特工 530 | 531 | 107 532 | 00:05:55,139 --> 00:06:01,150 533 | are three preference profiles the three 534 | 是三个偏好配置文件，三个 535 | 536 | 108 537 | 00:06:01,350 --> 00:06:05,490 538 | of the agent prefer A to B to C etc and 539 | 的代理商更喜欢A而不是B而不是C等，并且 540 | 541 | 109 542 | 00:06:05,689 --> 00:06:09,009 543 | clearly then a would be the winner under 544 | 显然，那么a将是下一个赢家 545 | 546 | 110 547 | 00:06:09,209 --> 00:06:12,220 548 | poor ality because three agents would 549 | 差劲，因为三个特工会 550 | 551 | 111 552 | 00:06:12,420 --> 00:06:17,350 553 | vote for a now what happens if we go and 554 | 现在投票，如果我们去 555 | 556 | 112 557 | 00:06:17,550 --> 00:06:26,370 558 | we modify this we modify this preference 559 | 我们修改此，我们修改此首选项 560 | 561 | 113 562 | 00:06:26,569 --> 00:06:31,329 563 | to this preference we simply make a 564 | 为此，我们只需 565 | 566 | 114 567 | 00:06:31,529 --> 00:06:35,740 568 | situation bitter C was preferred to a 569 | 情况C比C偏爱 570 | 571 | 115 572 | 00:06:35,939 --> 00:06:37,840 573 | originally and now a is preferred to C 574 | 最初，现在A优于C 575 | 576 | 116 577 | 00:06:38,040 --> 00:06:39,910 578 | and all other preferences 579 | 和所有其他偏好 580 | 581 | 117 582 | 00:06:40,110 --> 00:06:44,110 583 | to a remain the same so clear surely you 584 | 保持不变，所以请确保您清楚 585 | 586 | 118 587 | 00:06:44,310 --> 00:06:47,980 588 | would say a as is doing as well as it 589 | 会说做得很好 590 | 591 | 119 592 | 00:06:48,180 --> 00:06:50,319 593 | would have done originally and 594 | 本来可以做到的， 595 | 596 | 120 597 | 00:06:50,519 --> 00:06:51,759 598 | monotonicity would say that your 599 | 单调性表示您 600 | 601 | 121 602 | 00:06:51,959 --> 00:06:53,410 603 | children the winner but clearly in this 604 | 孩子赢家，但显然 605 | 606 | 122 607 | 00:06:53,610 --> 00:06:56,319 608 | case they're not four agents were voting 609 | 如果他们不是四个特工在投票 610 | 611 | 123 612 | 00:06:56,519 --> 00:06:59,290 613 | for B versus only three for a so B would 614 | 对于B来说，只有三个 615 | 616 | 124 617 | 00:06:59,490 --> 00:07:02,290 618 | be the winner now so plurality is not 619 | 现在成为赢家，所以不是多元化 620 | 621 | 125 622 | 00:07:02,490 --> 00:07:07,490 623 | monotonic 624 | 单调的 625 | 626 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 2/GTO2-2-01 - Mechanism Design - Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,000 --> 00:00:02,948 3 | hi folks so welcome back and now we'll 4 | 大家好，欢迎回来，现在我们将 5 | 6 | 2 7 | 00:00:03,149 --> 00:00:04,919 8 | be talking a bit about mechanism design 9 | 谈论机制设计 10 | 11 | 3 12 | 00:00:05,120 --> 00:00:09,790 13 | in general so the idea here is that in a 14 | 一般来说，这里的想法是 15 | 16 | 4 17 | 00:00:09,990 --> 00:00:11,410 18 | lot of what we've looked at so far in 19 | 到目前为止，我们已经看过很多 20 | 21 | 5 22 | 00:00:11,609 --> 00:00:13,300 23 | the course we've been looking at 24 | 我们一直在看的课程 25 | 26 | 6 27 | 00:00:13,500 --> 00:00:15,550 28 | specific institution specific games and 29 | 特定机构特定的游戏和 30 | 31 | 7 32 | 00:00:15,750 --> 00:00:18,839 33 | trying to understand how they affect 34 | 试图了解它们如何影响 35 | 36 | 8 37 | 00:00:19,039 --> 00:00:21,460 38 | behavior and in particular how the 39 | 行为，尤其是 40 | 41 | 9 42 | 00:00:21,660 --> 00:00:23,470 43 | structure of the game can be used to 44 | 游戏的结构可以用来 45 | 46 | 10 47 | 00:00:23,670 --> 00:00:26,079 48 | analyze and predict how people going to 49 | 分析和预测人们如何去 50 | 51 | 11 52 | 00:00:26,278 --> 00:00:28,089 53 | behave and now what we want to do is 54 | 表现，现在我们要做的是 55 | 56 | 12 57 | 00:00:28,289 --> 00:00:30,129 58 | flip things upside down so this will be 59 | 颠倒东西，这将是 60 | 61 | 13 62 | 00:00:30,329 --> 00:00:32,049 63 | game theory on its head and the idea is 64 | 博弈论从头到尾都是 65 | 66 | 14 67 | 00:00:32,250 --> 00:00:33,729 68 | instead of taking the game is given and 69 | 而不是采取游戏， 70 | 71 | 15 72 | 00:00:33,929 --> 00:00:36,128 73 | trying to analyze it we'll look at 74 | 试图分析它，我们将看 75 | 76 | 16 77 | 00:00:36,329 --> 00:00:37,779 78 | situations where we have some idea of 79 | 我们有一些想法的情况 80 | 81 | 17 82 | 00:00:37,979 --> 00:00:40,029 83 | what we'd like to see be the outcome and 84 | 我们希望看到的是结果， 85 | 86 | 18 87 | 00:00:40,229 --> 00:00:41,619 88 | then the question is can we design a 89 | 那么问题是我们可以设计一个 90 | 91 | 19 92 | 00:00:41,820 --> 00:00:44,320 93 | game that will yield that outcome so can 94 | 会产生这种结果的游戏 95 | 96 | 20 97 | 00:00:44,520 --> 00:00:47,229 98 | we design institutions and you know one 99 | 我们设计机构，您知道一个 100 | 101 | 21 102 | 00:00:47,429 --> 00:00:51,338 103 | big puzzle that we have in terms of the 104 | 就我们而言，这是一个大难题 105 | 106 | 22 107 | 00:00:51,539 --> 00:00:53,259 108 | way people behave is when you think 109 | 人们的行为方式就是当您认为 110 | 111 | 23 112 | 00:00:53,460 --> 00:00:55,838 113 | about the inefficiency that goes on in 114 | 关于持续的低效率 115 | 116 | 24 117 | 00:00:56,039 --> 00:00:57,849 118 | terms of bargaining there's often 119 | 讨价还价的条件经常是 120 | 121 | 25 122 | 00:00:58,049 --> 00:00:59,858 123 | bargaining breakdowns and people spend a 124 | 讨价还价的故障，人们花了 125 | 126 | 26 127 | 00:01:00,058 --> 00:01:01,178 128 | lot of time trying to figure out how do 129 | 很多时间试图弄清楚怎么做 130 | 131 | 27 132 | 00:01:01,378 --> 00:01:02,649 133 | we mediate bargaining what's an 134 | 我们调解讨价还价 135 | 136 | 28 137 | 00:01:02,850 --> 00:01:05,019 138 | appropriate technique to bring two sides 139 | 带两个方面的适当技巧 140 | 141 | 29 142 | 00:01:05,219 --> 00:01:06,369 143 | together so that they can reach a 144 | 在一起，以便他们可以达到 145 | 146 | 30 147 | 00:01:06,569 --> 00:01:08,379 148 | bargain and in terms of complete 149 | 讨价还价和完整 150 | 151 | 31 152 | 00:01:08,579 --> 00:01:10,450 153 | information game theory if people can 154 | 信息博弈论 155 | 156 | 32 157 | 00:01:10,650 --> 00:01:13,359 158 | really see what the payoffs are and you 159 | 真的看到了回报，你 160 | 161 | 33 162 | 00:01:13,560 --> 00:01:15,539 163 | looked at a particular bargaining game 164 | 看过一个讨价还价的游戏 165 | 166 | 34 167 | 00:01:15,739 --> 00:01:18,009 168 | generally you would have situations 169 | 通常情况下 170 | 171 | 35 172 | 00:01:18,209 --> 00:01:19,179 173 | where people would tend to reach 174 | 人们倾向于到达的地方 175 | 176 | 36 177 | 00:01:19,379 --> 00:01:20,619 178 | agreements they can look down the tree 179 | 他们可以看不起树的协议 180 | 181 | 37 182 | 00:01:20,819 --> 00:01:21,939 183 | they can figure out what's going to 184 | 他们可以弄清楚会发生什么 185 | 186 | 38 187 | 00:01:22,140 --> 00:01:24,609 188 | happen and and unravel things and reach 189 | 发生并解散事物并达到 190 | 191 | 39 192 | 00:01:24,810 --> 00:01:28,269 193 | an agreement and so here we can ask a 194 | 协议，所以在这里我们可以问一个 195 | 196 | 40 197 | 00:01:28,469 --> 00:01:30,128 198 | question of well let's suppose now that 199 | 好问题，让我们现在假设 200 | 201 | 41 202 | 00:01:30,328 --> 00:01:32,528 203 | we're in a Bayesian setting so we have a 204 | 我们处于贝叶斯环境，所以我们有一个 205 | 206 | 42 207 | 00:01:32,728 --> 00:01:34,719 208 | situation where we have workers thinking 209 | 我们有工人思考的情况 210 | 211 | 43 212 | 00:01:34,920 --> 00:01:36,579 213 | about trying to get higher wages and we 214 | 关于试图获得更高的工资，我们 215 | 216 | 44 217 | 00:01:36,780 --> 00:01:38,890 218 | have management and shareholders on the 219 | 拥有管理层和股东 220 | 221 | 45 222 | 00:01:39,090 --> 00:01:41,259 223 | other side trying to keep wages low and 224 | 另一方试图保持低工资， 225 | 226 | 46 227 | 00:01:41,459 --> 00:01:43,869 228 | each wants different things coming out 229 | 每个人都想要不同的东西出来 230 | 231 | 47 232 | 00:01:44,069 --> 00:01:47,769 233 | of a new agreement a labor agreement can 234 | 新协议中的劳动协议可以 235 | 236 | 48 237 | 00:01:47,969 --> 00:01:49,209 238 | we design a way in which they're going 239 | 我们设计他们前进的方式 240 | 241 | 49 242 | 00:01:49,409 --> 00:01:51,819 243 | to come to a table and make offers and 244 | 来一张桌子并提出要约 245 | 246 | 50 247 | 00:01:52,019 --> 00:01:53,469 248 | counteroffers in such a way that they'd 249 | 还价的方式是 250 | 251 | 51 252 | 00:01:53,670 --> 00:01:55,509 253 | actually reach an agreement and avoid a 254 | 实际上达成协议，避免 255 | 256 | 52 257 | 00:01:55,709 --> 00:01:57,969 258 | strike and avoid wasting resources and 259 | 罢工，避免浪费资源， 260 | 261 | 53 262 | 00:01:58,170 --> 00:02:01,000 263 | and time we see strikes all the time so 264 | 和时间，我们一直看到罢工，所以 265 | 266 | 54 267 | 00:02:01,200 --> 00:02:03,549 268 | the answer evidently seems to be no we 269 | 答案显然是不，我们 270 | 271 | 55 272 | 00:02:03,750 --> 00:02:06,250 273 | can't necessarily design such a 274 | 不一定设计这样的 275 | 276 | 56 277 | 00:02:06,450 --> 00:02:09,360 278 | procedure so the question becomes what 279 | 程序，所以问题变成什么 280 | 281 | 57 282 | 00:02:09,560 --> 00:02:11,740 283 | procedures can we design what are the 284 | 我们可以设计什么程序 285 | 286 | 58 287 | 00:02:11,939 --> 00:02:13,510 288 | best that we can do 289 | 尽我们所能 290 | 291 | 59 292 | 00:02:13,710 --> 00:02:16,120 293 | how do we begin to think about that how 294 | 我们如何开始思考 295 | 296 | 60 297 | 00:02:16,319 --> 00:02:18,160 298 | do we begin to model designing an 299 | 我们开始进行模型设计吗 300 | 301 | 61 302 | 00:02:18,360 --> 00:02:19,840 303 | institution and then seeing what the 304 | 机构，然后看看 305 | 306 | 62 307 | 00:02:20,039 --> 00:02:21,310 308 | outcomes are going to be and trying to 309 | 结果将是并试图 310 | 311 | 63 312 | 00:02:21,509 --> 00:02:23,050 313 | design the best institution in terms of 314 | 设计最佳机构 315 | 316 | 64 317 | 00:02:23,250 --> 00:02:25,390 318 | those outcomes so we'll be looking at 319 | 这些结果，所以我们将研究 320 | 321 | 65 322 | 00:02:25,590 --> 00:02:26,890 323 | things like auction design we'll talk 324 | 拍卖设计之类的东西 325 | 326 | 66 327 | 00:02:27,090 --> 00:02:28,860 328 | about why we might have breakdown and 329 | 关于为什么我们可能会崩溃， 330 | 331 | 67 332 | 00:02:29,060 --> 00:02:31,719 333 | inefficiency in bargaining settings when 334 | 谈判时效率低下 335 | 336 | 68 337 | 00:02:31,919 --> 00:02:33,520 338 | people have private information about 339 | 人们有关于的私人信息 340 | 341 | 69 342 | 00:02:33,719 --> 00:02:35,490 343 | what they'd like to see be the outcome 344 | 他们希望看到的结果 345 | 346 | 70 347 | 00:02:35,689 --> 00:02:38,620 348 | we're going to look at we've already had 349 | 我们要看看我们已经有 350 | 351 | 71 352 | 00:02:38,819 --> 00:02:41,170 353 | a look at voting rules but we'll look at 354 | 看看投票规则，我们来看看 355 | 356 | 72 357 | 00:02:41,370 --> 00:02:43,840 358 | a general set of decision mechanisms 359 | 一套通用的决策机制 360 | 361 | 73 362 | 00:02:44,039 --> 00:02:47,439 363 | that public enterprises can use to 364 | 公营企业可以用来 365 | 366 | 74 367 | 00:02:47,639 --> 00:02:49,480 368 | improve efficiency so we're going to 369 | 提高效率，所以我们要 370 | 371 | 75 372 | 00:02:49,680 --> 00:02:53,439 373 | look at in some specific detail at how 374 | 详细了解如何 375 | 376 | 76 377 | 00:02:53,639 --> 00:02:55,900 378 | to design things when we want to achieve 379 | 在我们想要实现时设计事物 380 | 381 | 77 382 | 00:02:56,099 --> 00:02:57,580 383 | efficiency we want to get the best 384 | 我们想要获得最好的效率 385 | 386 | 78 387 | 00:02:57,780 --> 00:02:59,590 388 | possible outcomes and we realize that 389 | 可能的结果，我们意识到 390 | 391 | 79 392 | 00:02:59,789 --> 00:03:01,180 393 | people have private information and 394 | 人们拥有私人信息， 395 | 396 | 80 397 | 00:03:01,379 --> 00:03:02,140 398 | they're going to try and take advantage 399 | 他们将尝试并利用 400 | 401 | 81 402 | 00:03:02,340 --> 00:03:05,020 403 | of that and have incentives to try and 404 | 并有动机去尝试 405 | 406 | 82 407 | 00:03:05,219 --> 00:03:06,939 408 | get the best that they can get out which 409 | 得到最好的，他们可以摆脱 410 | 411 | 83 412 | 00:03:07,139 --> 00:03:09,880 413 | doesn't necessarily always coincide with 414 | 不一定总是与 415 | 416 | 84 417 | 00:03:10,080 --> 00:03:12,070 418 | social efficiency so when is it that we 419 | 社会效率，那么我们什么时候 420 | 421 | 85 422 | 00:03:12,270 --> 00:03:13,480 423 | can get socially efficient outcomes 424 | 可以获得社会有效的结果 425 | 426 | 86 427 | 00:03:13,680 --> 00:03:15,520 428 | that's going to be the central question 429 | 这将是核心问题 430 | 431 | 87 432 | 00:03:15,719 --> 00:03:20,719 433 | in mechanism design so let's get started 434 | 在机制设计上，让我们开始吧 435 | 436 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 3/GTO2-3-03 - VCG Example.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,120 --> 00:00:03,248 3 | let us look at an example of VCG in 4 | 让我们来看一个VCG的例子 5 | 6 | 2 7 | 00:00:03,448 --> 00:00:05,589 8 | action of the victory park roads 9 | 胜利公园道路行动 10 | 11 | 3 12 | 00:00:05,790 --> 00:00:08,970 13 | mechanism in action and it'll be a 14 | 作用机制，它将是一个 15 | 16 | 4 17 | 00:00:09,169 --> 00:00:13,059 18 | rafting transportation problem so here 19 | 漂流问题，所以在这里 20 | 21 | 5 22 | 00:00:13,259 --> 00:00:17,949 23 | is a directed network of links each link 24 | 是链接的定向网络，每个链接 25 | 26 | 6 27 | 00:00:18,149 --> 00:00:20,829 28 | had the cost of the length of it if you 29 | 如果你有长度的代价 30 | 31 | 7 32 | 00:00:21,028 --> 00:00:24,669 33 | wish and the goal is to find a shortest 34 | 希望，目标是找到最短的 35 | 36 | 8 37 | 00:00:24,868 --> 00:00:29,318 38 | path from A to F and obviously looking 39 | 从A到F的路径，显然 40 | 41 | 9 42 | 00:00:29,518 --> 00:00:31,930 43 | at it I'll be quite obvious that the 44 | 在这一点上，我会很明显地看到 45 | 46 | 10 47 | 00:00:32,130 --> 00:00:36,038 48 | shortest path is this one and so this is 49 | 最短的路径就是这个，所以这是 50 | 51 | 11 52 | 00:00:36,238 --> 00:00:40,979 53 | the path that would be selected by a 54 | 由a选择的路径 55 | 56 | 12 57 | 00:00:41,179 --> 00:00:43,959 58 | traveler who wishes to get there from A 59 | 希望从A到达那里的旅行者 60 | 61 | 13 62 | 00:00:44,159 --> 00:00:50,399 63 | to F most quickly now according to a BCG 64 | 根据BCG，现在最快地到达F 65 | 66 | 14 67 | 00:00:50,600 --> 00:00:55,119 68 | how much would the owner of the link AF 69 | 链接AF的所有者要多少钱 70 | 71 | 15 72 | 00:00:55,320 --> 00:01:00,279 73 | pay well intuitively he's not part of 74 | 凭直觉支付好他不属于 75 | 76 | 16 77 | 00:01:00,479 --> 00:01:04,539 78 | the other path so we shouldn't have to 79 | 另一条路，所以我们不必 80 | 81 | 17 82 | 00:01:04,739 --> 00:01:07,509 83 | pay or receive anything let's see what 84 | 支付或接收任何东西，让我们看看 85 | 86 | 18 87 | 00:01:07,709 --> 00:01:13,629 88 | VCG says so the shorter path with a C's 89 | VCG说，所以C的较短路径 90 | 91 | 19 92 | 00:01:13,829 --> 00:01:16,679 93 | declaration of what their cost is is 94 | 声明其成本是 95 | 96 | 20 97 | 00:01:16,879 --> 00:01:20,799 98 | five right and so the cost to all the 99 | 五权，所以所有的成本 100 | 101 | 21 102 | 00:01:21,000 --> 00:01:24,879 103 | agents is that shortest path totaling 104 | 代理商是最短的路径总计 105 | 106 | 22 107 | 00:01:25,079 --> 00:01:31,409 108 | five had AC not being in the picture and 109 | 有五个人不在照片中 110 | 111 | 23 112 | 00:01:31,609 --> 00:01:34,058 113 | not declared anything therefore what if 114 | 没有声明任何东西，如果 115 | 116 | 24 117 | 00:01:34,259 --> 00:01:36,939 118 | what would the total cost to the agents 119 | 代理商的总费用是多少 120 | 121 | 25 122 | 00:01:37,140 --> 00:01:38,528 123 | have been well it would have been the 124 | 很好，应该是 125 | 126 | 26 127 | 00:01:38,728 --> 00:01:41,140 128 | very same path so the shortest path and 129 | 路径完全相同，所以最短路径和 130 | 131 | 27 132 | 00:01:41,340 --> 00:01:44,349 133 | so the cost would have been five also 134 | 所以成本也将是五 135 | 136 | 28 137 | 00:01:44,549 --> 00:01:50,198 138 | and so the amount that AC pays if you 139 | 因此，如果您 140 | 141 | 29 142 | 00:01:50,399 --> 00:01:52,448 143 | wish is the difference between the cost 144 | 希望是成本之间的差异 145 | 146 | 30 147 | 00:01:52,649 --> 00:01:56,709 148 | to the agents within the picture and the 149 | 给图片中的代理商和 150 | 151 | 31 152 | 00:01:56,909 --> 00:02:00,250 153 | cost without him in the picture which of 154 | 图片中没有他的代价 155 | 156 | 32 157 | 00:02:00,450 --> 00:02:03,250 158 | course is zero and this is true for all 159 | 当然是零，这对所有人都是正确的 160 | 161 | 33 162 | 00:02:03,450 --> 00:02:06,579 163 | the other ages other edges that don't 164 | 其他年龄的其他优势 165 | 166 | 34 167 | 00:02:06,780 --> 00:02:09,929 168 | participate in the shortest path all 169 | 全部参加最短路径 170 | 171 | 35 172 | 00:02:10,128 --> 00:02:13,818 173 | these edges 174 | 这些边缘 175 | 176 | 36 177 | 00:02:18,250 --> 00:02:22,170 178 | draw in blue these edges not 179 | 将这些边缘涂成蓝色 180 | 181 | 37 182 | 00:02:22,370 --> 00:02:26,570 183 | participating the shortest path will 184 | 参加最短路径将 185 | 186 | 38 187 | 00:02:26,770 --> 00:02:29,189 188 | will neither pain or seep anything 189 | 既不会痛苦也不会渗入任何东西 190 | 191 | 39 192 | 00:02:29,389 --> 00:02:32,210 193 | according to BCG so that's the easy part 194 | 根据BCG的说法，这很容易 195 | 196 | 40 197 | 00:02:32,409 --> 00:02:35,340 198 | what about edges that do participate in 199 | 那么参与的边缘呢 200 | 201 | 41 202 | 00:02:35,539 --> 00:02:38,700 203 | the shortest path so let's look for 204 | 最短的路径，让我们寻找 205 | 206 | 42 207 | 00:02:38,900 --> 00:02:47,100 208 | example at one of them at a B well with 209 | 例如在一个B井中 210 | 211 | 43 212 | 00:02:47,300 --> 00:02:52,410 213 | the shortest path as we know is it is 214 | 我们知道的最短路径是 215 | 216 | 44 217 | 00:02:52,610 --> 00:02:56,570 218 | this one and how much would all agents 219 | 这个和所有代理商多少钱 220 | 221 | 45 222 | 00:02:56,770 --> 00:03:01,050 223 | other than a B itself what is the cost 224 | 除了B本身，成本是多少 225 | 226 | 46 227 | 00:03:01,250 --> 00:03:03,900 228 | of the agents on the shortest path well 229 | 最短路径上的代理商 230 | 231 | 47 232 | 00:03:04,099 --> 00:03:10,200 233 | it's 1 plus 1 is 2 cost of two had a B 234 | 它的1加1是2的成本，两个有B 235 | 236 | 48 237 | 00:03:10,400 --> 00:03:14,640 238 | not being in the picture what would have 239 | 不在图片中会有什么 240 | 241 | 49 242 | 00:03:14,840 --> 00:03:18,330 243 | happened well the shortest path would 244 | 发生得最好的最短路径 245 | 246 | 50 247 | 00:03:18,530 --> 00:03:23,910 248 | have been um this one over here so this 249 | 一直在这一个这里，所以这个 250 | 251 | 51 252 | 00:03:24,110 --> 00:03:26,939 253 | would have been the shortest path of a 254 | 本来是一条最短的路 255 | 256 | 52 257 | 00:03:27,139 --> 00:03:36,689 258 | cost of 2 plus 3 plus 1 and so the cost 259 | 成本2加3加1等等 260 | 261 | 53 262 | 00:03:36,889 --> 00:03:39,810 263 | to the agent a B is the difference 264 | 与代理商A B的区别 265 | 266 | 54 267 | 00:03:40,009 --> 00:03:42,719 268 | between the cost without them in the 269 | 在没有它们的情况下 270 | 271 | 55 272 | 00:03:42,919 --> 00:03:46,230 273 | picture which would be minus 6 and the 274 | 图片将是负6和 275 | 276 | 56 277 | 00:03:46,430 --> 00:03:50,100 278 | cost with them in the picture which is 279 | 与他们在图片中的成本是 280 | 281 | 57 282 | 00:03:50,300 --> 00:03:52,170 283 | minus 2 that is a cost to all the other 284 | 减2是其他所有费用 285 | 286 | 58 287 | 00:03:52,370 --> 00:03:56,850 288 | agents and the difference is minus 4 so 289 | 代理和差是负4所以 290 | 291 | 59 292 | 00:03:57,050 --> 00:03:59,430 293 | the crust to a B is minus 4 in other 294 | B的外壳在其他地方为负4 295 | 296 | 60 297 | 00:03:59,629 --> 00:04:06,890 298 | words a B will get a payment of four 299 | 单词B将获得四的付款 300 | 301 | 61 302 | 00:04:07,090 --> 00:04:09,960 303 | making if you wish a profit of 1 because 304 | 如果您希望获利1， 305 | 306 | 62 307 | 00:04:10,159 --> 00:04:12,840 308 | their cost is 3 we'll get a payment of 4 309 | 他们的成本是3我们将支付4 310 | 311 | 63 312 | 00:04:13,039 --> 00:04:17,848 313 | and they'll have a benefit of 1 so this 314 | 他们将受益1 315 | 316 | 64 317 | 00:04:18,048 --> 00:04:23,810 318 | is for a B what about B II for example 319 | 对于一个B例如B II 320 | 321 | 65 322 | 00:04:24,009 --> 00:04:27,869 323 | well the same sort of analysis 324 | 同样的分析 325 | 326 | 66 327 | 00:04:28,069 --> 00:04:34,579 328 | ve the cost to all other agents without 329 | 将所有其他代理商的费用 330 | 331 | 67 332 | 00:04:34,779 --> 00:04:37,170 333 | being the picture is minus six it's 334 | 图片是负六 335 | 336 | 68 337 | 00:04:37,370 --> 00:04:40,020 338 | still this would have been the shorter 339 | 仍然会更短 340 | 341 | 69 342 | 00:04:40,220 --> 00:04:42,930 343 | route had been on in picture and cost 344 | 路线已经在图片和费用中 345 | 346 | 70 347 | 00:04:43,129 --> 00:04:49,199 348 | would be minus six now the cost to the 349 | 现在将是负六 350 | 351 | 71 352 | 00:04:49,399 --> 00:04:51,360 353 | agents with be in the picture to the 354 | 与代理商在图片中 355 | 356 | 72 357 | 00:04:51,560 --> 00:04:54,389 358 | agent rather than be e is for why is 359 | 代理而不是e是为什么 360 | 361 | 73 362 | 00:04:54,589 --> 00:04:59,819 363 | that it's one plus three and so the cost 364 | 那是一加三，所以成本 365 | 366 | 74 367 | 00:05:00,019 --> 00:05:03,449 368 | to this is the minus four over here so 369 | 这是这里的负四 370 | 371 | 75 372 | 00:05:03,649 --> 00:05:08,670 373 | the net of it is that the cost imposed 374 | 最终的结果是所施加的成本 375 | 376 | 76 377 | 00:05:08,870 --> 00:05:14,430 378 | on by VCG on the link be e is minus two 379 | 由链接上的VCG开启，e为负2 380 | 381 | 77 382 | 00:05:14,629 --> 00:05:17,610 383 | in other words they would get a payment 384 | 换句话说，他们会得到付款 385 | 386 | 78 387 | 00:05:17,810 --> 00:05:20,910 388 | of two making again a profit of one if 389 | 的二分之一再次获利，如果 390 | 391 | 79 392 | 00:05:21,110 --> 00:05:23,470 393 | you wish 394 | 你希望 395 | 396 | 80 397 | 00:05:23,529 --> 00:05:31,819 398 | what about EF same same of calculation 399 | EF与计算相同 400 | 401 | 81 402 | 00:05:32,019 --> 00:05:33,959 403 | although the outcome would be a little 404 | 虽然结果会有点 405 | 406 | 82 407 | 00:05:34,158 --> 00:05:38,040 408 | different because in that case they will 409 | 不同，因为在这种情况下，他们会 410 | 411 | 83 412 | 00:05:38,240 --> 00:05:41,189 413 | be getting a payment of three why is 414 | 得到三分钱，为什么 415 | 416 | 84 417 | 00:05:41,389 --> 00:05:44,160 418 | that well without them in the picture 419 | 没有他们在照片中很好 420 | 421 | 85 422 | 00:05:44,360 --> 00:05:46,528 423 | this would no longer be the shortest 424 | 这将不再是最短的 425 | 426 | 86 427 | 00:05:46,728 --> 00:05:49,319 428 | path the shortest path would be a cost 429 | 最短的路径将是成本 430 | 431 | 87 432 | 00:05:49,519 --> 00:05:52,649 433 | of 7 whether it's this one or this one 434 | 的7，无论是这个还是这个 435 | 436 | 88 437 | 00:05:52,848 --> 00:05:55,499 438 | in both cases the cost of the first path 439 | 在这两种情况下，第一条路径的成本 440 | 441 | 89 442 | 00:05:55,699 --> 00:05:58,100 443 | will be 7 so the total cost to the 444 | 将为7，因此总费用为 445 | 446 | 90 447 | 00:05:58,300 --> 00:06:00,838 448 | population without them in the picture 449 | 图片中没有他们的人口 450 | 451 | 91 452 | 00:06:01,038 --> 00:06:06,300 453 | would be 7 with them the cost is again 454 | 和他们在一起将是7 455 | 456 | 92 457 | 00:06:06,500 --> 00:06:10,199 458 | the 1 plus 3 it's a 4 and so they get 459 | 1加3是4，所以他们得到 460 | 461 | 93 462 | 00:06:10,399 --> 00:06:15,420 463 | the difference of 3 so the payment would 464 | 3的差额，所以付款 465 | 466 | 94 467 | 00:06:15,620 --> 00:06:19,259 468 | be 3 and therefore the net profit if you 469 | 为3，因此净利润为 470 | 471 | 95 472 | 00:06:19,459 --> 00:06:22,819 473 | wish it would be 3 minus 1 name namely 2 474 | 希望它是3减1的名字，即2 475 | 476 | 96 477 | 00:06:23,019 --> 00:06:29,100 478 | so you could ask why are why is it the 479 | 所以你可以问为什么是为什么 480 | 481 | 97 482 | 00:06:29,300 --> 00:06:31,019 483 | difference you know you look at all 484 | 差异，你知道你看看所有 485 | 486 | 98 487 | 00:06:31,218 --> 00:06:36,329 488 | these links somehow their net payoff is 489 | 这些链接的净收益是 490 | 491 | 99 492 | 00:06:36,528 --> 00:06:40,620 493 | different and that's fair 494 | 不同，这很公平 495 | 496 | 100 497 | 00:06:40,819 --> 00:06:42,960 498 | and what fairness is in the eye of the 499 | 眼中的公平是什么 500 | 501 | 101 502 | 00:06:43,160 --> 00:06:45,960 503 | beholder but the rationale for it is 504 | 情人，但其理由是 505 | 506 | 102 507 | 00:06:46,160 --> 00:06:48,060 508 | that they have different market power 509 | 他们有不同的市场力量 510 | 511 | 103 512 | 00:06:48,259 --> 00:06:53,610 513 | the amount of social value they bring is 514 | 他们带来的社会价值是 515 | 516 | 104 517 | 00:06:53,810 --> 00:06:56,160 518 | different because without them the 519 | 不同，因为没有他们 520 | 521 | 105 522 | 00:06:56,360 --> 00:07:00,180 523 | outcome will be different into the 524 | 结果会有所不同 525 | 526 | 106 527 | 00:07:00,379 --> 00:07:02,879 528 | population and that's why they have a 529 | 人口，这就是为什么他们有一个 530 | 531 | 107 532 | 00:07:03,079 --> 00:07:06,170 533 | different sort of payments and different 534 | 不同的付款方式和不同的 535 | 536 | 108 537 | 00:07:06,370 --> 00:07:09,629 538 | cost structures here that's an example 539 | 成本结构就是一个例子 540 | 541 | 109 542 | 00:07:09,829 --> 00:07:14,829 543 | of BCG in operation 544 | 运营中的BCG 545 | 546 | -------------------------------------------------------------------------------- /subtitle en+zh/Game Theory II-Week 4/GTO2-4-01 - Auctions - Taste.srt: -------------------------------------------------------------------------------- 1 | 1 2 | 00:00:00,089 --> 00:00:02,829 3 | we're gonna talk about auctions in this 4 | 我们将在此谈论拍卖 5 | 6 | 2 7 | 00:00:03,029 --> 00:00:06,639 8 | segment and when we think about auctions 9 | 细分以及当我们考虑拍卖时 10 | 11 | 3 12 | 00:00:06,839 --> 00:00:09,100 13 | we usually think about something fairly 14 | 我们通常会公平考虑 15 | 16 | 4 17 | 00:00:09,300 --> 00:00:13,439 18 | specific often something fairly simple 19 | 具体往往很简单 20 | 21 | 5 22 | 00:00:13,638 --> 00:00:16,810 23 | this is probably what many of us think 24 | 这可能是我们许多人的想法 25 | 26 | 6 27 | 00:00:17,010 --> 00:00:18,190 28 | of when we think about auctions 29 | 当我们考虑拍卖的时候 30 | 31 | 7 32 | 00:00:18,390 --> 00:00:22,390 33 | initially an art auction where the 34 | 最初是艺术品拍卖， 35 | 36 | 8 37 | 00:00:22,589 --> 00:00:26,890 38 | auctioneer displays the art and people 39 | 拍卖师展示艺术品和人物 40 | 41 | 9 42 | 00:00:27,089 --> 00:00:27,970 43 | from the audience 44 | 听众 45 | 46 | 10 47 | 00:00:28,170 --> 00:00:31,350 48 | call out larger and larger numbers until 49 | 喊出越来越大的数字，直到 50 | 51 | 11 52 | 00:00:31,550 --> 00:00:34,059 53 | nobody says anything and the auctioneer 54 | 没有人说什么，拍卖师 55 | 56 | 12 57 | 00:00:34,259 --> 00:00:36,839 58 | famously says going once going twice 59 | 著名的说一次去两次 60 | 61 | 13 62 | 00:00:37,039 --> 00:00:40,809 63 | sold and that's this that's how the sale 64 | 出售，这就是销售的方式 65 | 66 | 14 67 | 00:00:41,009 --> 00:00:46,919 68 | is decided on somewhat differently is 69 | 决定的有所不同是 70 | 71 | 15 72 | 00:00:47,119 --> 00:00:49,750 73 | this what is this well this is the 74 | 这是什么好这是 75 | 76 | 16 77 | 00:00:49,950 --> 00:00:52,178 78 | flower market in Holland just outside 79 | 荷兰的花卉市场就在外面 80 | 81 | 17 82 | 00:00:52,378 --> 00:00:56,518 83 | Amsterdam and here every day large 84 | 阿姆斯特丹和这里每天都很大 85 | 86 | 18 87 | 00:00:56,719 --> 00:00:59,459 88 | number of flowers are trucked in and 89 | 运了一些花， 90 | 91 | 19 92 | 00:00:59,659 --> 00:01:01,989 93 | Seoul's sold rather quickly because 94 | 首尔的销量很快，因为 95 | 96 | 20 97 | 00:01:02,189 --> 00:01:05,289 98 | flowers are perishable and they're sold 99 | 花易腐烂，被卖掉 100 | 101 | 21 102 | 00:01:05,489 --> 00:01:08,769 103 | somewhat differently there is a trading 104 | 有所不同，有交易 105 | 106 | 22 107 | 00:01:08,969 --> 00:01:11,379 108 | pit where the pallets of flowers are 109 | 鲜花托盘所在的坑 110 | 111 | 23 112 | 00:01:11,579 --> 00:01:15,730 113 | rolled in and they are auctioned off 114 | 卷入，它们被拍卖掉 115 | 116 | 24 117 | 00:01:15,930 --> 00:01:18,099 118 | auctioned off rather differently you see 119 | 拍卖会与众不同 120 | 121 | 25 122 | 00:01:18,299 --> 00:01:21,480 123 | these large clocks in the background and 124 | 这些大钟在后台 125 | 126 | 26 127 | 00:01:21,680 --> 00:01:26,289 128 | they have a special role and what that 129 | 他们扮演着特殊的角色 130 | 131 | 27 132 | 00:01:26,489 --> 00:01:29,168 133 | is something to be discussed here it's a 134 | 这里要讨论的是 135 | 136 | 28 137 | 00:01:29,368 --> 00:01:31,599 138 | very different kind of auction these are 139 | 这些是非常不同的拍卖 140 | 141 | 29 142 | 00:01:31,799 --> 00:01:34,500 143 | auctions for the rights to use our 144 | 拍卖使用我们的权利 145 | 146 | 30 147 | 00:01:34,700 --> 00:01:39,009 148 | electric magnetic spectrum so our 149 | 电磁频谱所以我们 150 | 151 | 31 152 | 00:01:39,209 --> 00:01:42,659 153 | airwaves are filled with all kind of 154 | 电波充满了各种 155 | 156 | 32 157 | 00:01:42,859 --> 00:01:46,299 158 | broadcasts from television to Cellular 159 | 从电视广播到手机 160 | 161 | 33 162 | 00:01:46,500 --> 00:01:49,058 163 | to what-have-you and those are worth a 164 | 对你有什么，那些都值得 165 | 166 | 34 167 | 00:01:49,259 --> 00:01:53,859 168 | lot to commercially and socially so 169 | 在商业上和社会上都是如此 170 | 171 | 35 172 | 00:01:54,060 --> 00:01:56,799 173 | governments all over the world have 174 | 全世界的政府都有 175 | 176 | 36 177 | 00:01:57,000 --> 00:01:59,429 178 | auctioned off the rights to use this 179 | 拍卖使用权 180 | 181 | 37 182 | 00:01:59,629 --> 00:02:04,289 183 | spectrum famously starting with the u.s. 184 | 频谱始于美国 185 | 186 | 38 187 | 00:02:04,489 --> 00:02:08,039 188 | FCC or Federal Communication Commission 189 | FCC或联邦通讯委员会 190 | 191 | 39 192 | 00:02:08,239 --> 00:02:12,610 193 | besides being valuable to society and 194 | 除了对社会有价值 195 | 196 | 40 197 | 00:02:12,810 --> 00:02:15,840 198 | fetching very large sums in the billions 199 | 取得数十亿美元的巨额款项 200 | 201 | 41 202 | 00:02:16,039 --> 00:02:18,490 203 | these altars very interesting because 204 | 这些祭坛非常有趣，因为 205 | 206 | 42 207 | 00:02:18,689 --> 00:02:21,130 208 | they're complicated what's being sold 209 | 他们很复杂正在出售什么 210 | 211 | 43 212 | 00:02:21,330 --> 00:02:23,050 213 | here is not something as simple as a 214 | 这不是一个简单的东西 215 | 216 | 44 217 | 00:02:23,250 --> 00:02:26,140 218 | single piece piece of art or even a 219 | 一件艺术品，甚至一件 220 | 221 | 45 222 | 00:02:26,340 --> 00:02:28,000 223 | palette of flowers 224 | 花的调色板 225 | 226 | 46 227 | 00:02:28,199 --> 00:02:32,860 228 | what sold here are regions and 229 | 这里卖的是地区和 230 | 231 | 47 232 | 00:02:33,060 --> 00:02:37,030 233 | frequencies and so different regions in 234 | 频率等不同区域 235 | 236 | 48 237 | 00:02:37,229 --> 00:02:39,850 238 | the country are different goods and in 239 | 这个国家是不同的商品 240 | 241 | 49 242 | 00:02:40,050 --> 00:02:41,590 243 | each region one can broadcast at 244 | 每个地区都可以在 245 | 246 | 50 247 | 00:02:41,789 --> 00:02:44,110 248 | different frequencies so that makes for 249 | 不同的频率，因此 250 | 251 | 51 252 | 00:02:44,310 --> 00:02:46,420 253 | a different kind of auction where both 254 | 另一种拍卖方式 255 | 256 | 52 257 | 00:02:46,620 --> 00:02:47,939 258 | the design of the auction and the 259 | 拍卖的设计和 260 | 261 | 53 262 | 00:02:48,139 --> 00:02:50,980 263 | trading in the auction are non-trivial 264 | 拍卖中的交易是不平凡的 265 | 266 | 54 267 | 00:02:51,180 --> 00:02:55,689 268 | and finally in all those cases until now 269 | 最后在所有这些情况下直到现在 270 | 271 | 55 272 | 00:02:55,889 --> 00:02:58,120 273 | we've had a single seller and main 274 | 我们有一个卖方和主要 275 | 276 | 56 277 | 00:02:58,319 --> 00:03:01,150 278 | buyers here's a famous example we have 279 | 买家，这是一个著名的例子 280 | 281 | 57 282 | 00:03:01,349 --> 00:03:04,630 283 | many cells and many buyers this is the 284 | 许多细胞和许多买家，这是 285 | 286 | 58 287 | 00:03:04,830 --> 00:03:08,530 288 | Nasdaq one of the many stock exchanges 289 | 纳斯达克众多证券交易所之一 290 | 291 | 59 292 | 00:03:08,729 --> 00:03:11,259 293 | in the world we have many buyers and is 294 | 在世界上，我们有很多买家 295 | 296 | 60 297 | 00:03:11,459 --> 00:03:14,140 298 | very sellers of goods and in fact these 299 | 非常商品的卖家，实际上这些 300 | 301 | 61 302 | 00:03:14,340 --> 00:03:16,780 303 | Goods coming multiple units if you wish 304 | 如果您愿意的话，货品有多个单位 305 | 306 | 62 307 | 00:03:16,979 --> 00:03:20,819 308 | to buy Google stock then you can buy 309 | 买谷歌股票然后你可以买 310 | 311 | 63 312 | 00:03:21,019 --> 00:03:25,539 313 | many such shares and there are many 314 | 很多这样的股份，还有很多 315 | 316 | 64 317 | 00:03:25,739 --> 00:03:27,550 318 | others like you want to buy it and many 319 | 其他像您想购买的东西 320 | 321 | 65 322 | 00:03:27,750 --> 00:03:31,780 323 | many who want to sell it so the way you 324 | 许多人想以您的方式出售它 325 | 326 | 66 327 | 00:03:31,979 --> 00:03:34,660 328 | design a two-sided exchange and the way 329 | 设计双向交流和方式 330 | 331 | 67 332 | 00:03:34,860 --> 00:03:37,090 333 | you trade in it and again different from 334 | 您进行交易，再次与 335 | 336 | 68 337 | 00:03:37,289 --> 00:03:39,400 338 | the simple single sided option we know 339 | 我们知道的简单的单面选项 340 | 341 | 69 342 | 00:03:39,599 --> 00:03:42,129 343 | these are all examples of auctions there 344 | 这些都是那里的拍卖的例子 345 | 346 | 70 347 | 00:03:42,329 --> 00:03:45,370 348 | are many other examples and the theory 349 | 还有很多其他例子和理论 350 | 351 | 71 352 | 00:03:45,569 --> 00:03:47,980 353 | about how to design and how to be able 354 | 关于如何设计和如何能够 355 | 356 | 72 357 | 00:03:48,180 --> 00:03:53,180 358 | option is very rich 359 | 选项非常丰富 360 | 361 | -------------------------------------------------------------------------------- /update.sh: -------------------------------------------------------------------------------- 1 | git add -A 2 | git commit -am "$(date "+%Y-%m-%d %H:%M:%S")" 3 | git push --------------------------------------------------------------------------------