Math 152: Applicable Mathematics and Computing

Similar documents
Math 152: Applicable Mathematics and Computing

Contents. MA 327/ECO 327 Introduction to Game Theory Fall 2017 Notes. 1 Wednesday, August Friday, August Monday, August 28 6

A Brief Introduction to Game Theory

A Brief Introduction to Game Theory

Tutorial 1. (ii) There are finite many possible positions. (iii) The players take turns to make moves.

Topics in Computer Mathematics. two or more players Uncertainty (regarding the other player(s) resources and strategies)

CS510 \ Lecture Ariel Stolerman

Advanced Microeconomics: Game Theory

CS 1571 Introduction to AI Lecture 12. Adversarial search. CS 1571 Intro to AI. Announcements

2. The Extensive Form of a Game

Background. Game Theory and Nim. The Game of Nim. Game is Finite 1/27/2011

Math 152: Applicable Mathematics and Computing

SF2972 Game Theory Written Exam March 17, 2011

Japanese. Sail North. Search Search Search Search

final examination on May 31 Topics from the latter part of the course (covered in homework assignments 4-7) include:

Sequential games. We may play the dating game as a sequential game. In this case, one player, say Connie, makes a choice before the other.

Name: Final Exam May 7, 2014

ARTIFICIAL INTELLIGENCE (CS 370D)

Combined Games. Block, Alexander Huang, Boao. icamp Summer Research Program University of California, Irvine Irvine, CA

Lecture Notes on Game Theory (QTM)

Announcements. Homework 1 solutions posted. Test in 2 weeks (27 th ) -Covers up to and including HW2 (informed search)

CSC304: Algorithmic Game Theory and Mechanism Design Fall 2016

Mohammad Hossein Manshaei 1394

Lecture 33: How can computation Win games against you? Chess: Mechanical Turk

Realizing Strategies for winning games. Senior Project Presented by Tiffany Johnson Math 498 Fall 1999

CS 2710 Foundations of AI. Lecture 9. Adversarial search. CS 2710 Foundations of AI. Game search

Game Theory Lecturer: Ji Liu Thanks for Jerry Zhu's slides

Math 147 Lecture Notes: Lecture 21

Grade 7/8 Math Circles. February 14 th /15 th. Game Theory. If they both confess, they will both serve 5 hours of detention.

Game theory attempts to mathematically. capture behavior in strategic situations, or. games, in which an individual s success in

Math 611: Game Theory Notes Chetan Prakash 2012

Module 3. Problem Solving using Search- (Two agent) Version 2 CSE IIT, Kharagpur

GAME THEORY. Part II. Two-Person Zero-Sum Games. Thomas S. Ferguson

37 Game Theory. Bebe b1 b2 b3. a Abe a a A Two-Person Zero-Sum Game

CSC/MTH 231 Discrete Structures II Spring, Homework 5

Self-interested agents What is Game Theory? Example Matrix Games. Game Theory Intro. Lecture 3. Game Theory Intro Lecture 3, Slide 1

Introduction: What is Game Theory?

Distributed Optimization and Games

Math 464: Linear Optimization and Game

Computing Nash Equilibrium; Maxmin

Plan. Related courses. A Take-Away Game. Mathematical Games , (21-801) - Mathematical Games Look for it in Spring 11

Mixed Strategies; Maxmin

Game Tree Search. CSC384: Introduction to Artificial Intelligence. Generalizing Search Problem. General Games. What makes something a game?

Game Theory Intro. Lecture 3. Game Theory Intro Lecture 3, Slide 1

game tree complete all possible moves

What is... Game Theory? By Megan Fava

Game-playing AIs: Games and Adversarial Search I AIMA

Chapter 15: Game Theory: The Mathematics of Competition Lesson Plan

Overview GAME THEORY. Basic notions

CSC384: Introduction to Artificial Intelligence. Game Tree Search

Introduction to Game Theory a Discovery Approach. Jennifer Firkins Nordstrom

Game Tree Search. Generalizing Search Problems. Two-person Zero-Sum Games. Generalizing Search Problems. CSC384: Intro to Artificial Intelligence

ECO 220 Game Theory. Objectives. Agenda. Simultaneous Move Games. Be able to structure a game in normal form Be able to identify a Nash equilibrium

Section Notes 6. Game Theory. Applied Math 121. Week of March 22, understand the difference between pure and mixed strategies.

Aspects of Game Theory & John Nash

CMU Lecture 22: Game Theory I. Teachers: Gianni A. Di Caro

Determine the Expected value for each die: Red, Blue and Green. Based on your calculations from Question 1, do you think the game is fair?

Homework 8 (for lectures on 10/14,10/16)

1. Simultaneous games All players move at same time. Represent with a game table. We ll stick to 2 players, generally A and B or Row and Col.

Exercises for Introduction to Game Theory SOLUTIONS

Modeling Strategic Environments 1 Extensive form games

the gamedesigninitiative at cornell university Lecture 6 Uncertainty & Risk

U strictly dominates D for player A, and L strictly dominates R for player B. This leaves (U, L) as a Strict Dominant Strategy Equilibrium.

Introduction to Game Theory

Games in Extensive Form

Math Games Ideas. For School or Home Education. by Teresa Evans. Copyright 2005 Teresa Evans. All rights reserved.

Week 1: Probability models and counting

Obliged Sums of Games

MIT 15.S50 LECTURE 5. Friday, January 27 th, 2012

Introduction To Game Theory: Two-Person Games of Perfect Information and Winning Strategies. Wes Weimer, University of Virginia

Multiagent Systems: Intro to Game Theory. CS 486/686: Introduction to Artificial Intelligence

Optimization of Multipurpose Reservoir Operation Using Game Theory

Multiagent Systems: Intro to Game Theory. CS 486/686: Introduction to Artificial Intelligence

Probability MAT230. Fall Discrete Mathematics. MAT230 (Discrete Math) Probability Fall / 37

ECON 282 Final Practice Problems

Ex 1: A coin is flipped. Heads, you win $1. Tails, you lose $1. What is the expected value of this game?

Advanced Automata Theory 4 Games

EXPLORING TIC-TAC-TOE VARIANTS

V. Adamchik Data Structures. Game Trees. Lecture 1. Apr. 05, Plan: 1. Introduction. 2. Game of NIM. 3. Minimax

Game, Set, and Match Carl W. Lee September 2016

Distributed Optimization and Games

Funny Money. The Big Idea. Supplies. Key Prep: What s the Math? Valuing units of money Counting by 5s and 10s. Grades K-2

Multiagent Systems: Intro to Game Theory. CS 486/686: Introduction to Artificial Intelligence

Adversarial Search and Game- Playing C H A P T E R 6 C M P T : S P R I N G H A S S A N K H O S R A V I

Senior Math Circles February 10, 2010 Game Theory II

Solution Concepts 4 Nash equilibrium in mixed strategies

Summary Overview of Topics in Econ 30200b: Decision theory: strong and weak domination by randomized strategies, domination theorem, expected utility

Section Summary. Finite Probability Probabilities of Complements and Unions of Events Probabilistic Reasoning

Game Playing AI. Dr. Baldassano Yu s Elite Education

CMPUT 396 Tic-Tac-Toe Game

DYNAMIC GAMES. Lecture 6

GCSE MATHEMATICS Intermediate Tier, topic sheet. PROBABILITY

Solving Problems by Searching: Adversarial Search

GAME THEORY. Part II. Two-Person Zero-Sum Games. Class notes for Math 167, Fall 2000 Thomas S. Ferguson

Game Playing State of the Art

Unit 9: Probability Assignments

Dominant Strategies (From Last Time)

Topic 1: defining games and strategies. SF2972: Game theory. Not allowed: Extensive form game: formal definition

Numan Sheikh FC College Lahore

Programming Project 1: Pacman (Due )

Transcription:

Math 152: Applicable Mathematics and Computing April 16, 2017 April 16, 2017 1 / 17

Announcements Please bring a blue book for the midterm on Friday. Some students will be taking the exam in Center 201, will announce which students before Wednesday s class. Exam covers Part I (chapters 1-4). Use the homeworks and lecture notes as a guide. April 16, 2017 2 / 17

Two-Player Zero Sum Games As before, we will be concerned with two player games. In particular we will study zero sum games: these are games where what one player wins is exactly what the other player loses. For example two-player poker: your winnings are exactly my losses. April 16, 2017 3 / 17

Two-Person Zero Sum Games Def. A two-person zero sum game is a game with two players, which we will call Player I and Player II, where one player wins what the other player loses. Eg. If Player I wins 5 dollars, this means that Player II loses 5 dollars. The prize-money is called the payoff. April 16, 2017 4 / 17

Two-Player Zero Sum Games Zero sum games are nice mathematically, because we can represent the outcome of the game as a single number x. x represents the winnings of Player I. For example, if x is 100 dollars, player I has taken 100 dollars from player II. But if x is 100 dollars, player II has taken 100 dollars from player I. April 16, 2017 5 / 17

Strategic Form Def. The strategic form of a two-person zero sum game is given by the triplet (X, Y, A), where 1 X is a nonempty set, called the strategies of Player I 2 Y is a nonempty set, called the strategies of Player II 3 A is a function mapping X Y to R (ie. for each x X and y Y, A(x, y) is a real number). This represents the payoff, given the strategies of the players. This is a mathematical way to represent a two-person zero sum game. Board example. Write Rock-Paper-Scissors in strategic form, where the winner wins 1, and both players receive 0 in the case of a draw. April 16, 2017 6 / 17

Strategic Form of a Game We imagine the game being played in the following way: simultaneously, player I chooses her strategy x from X and player II chooses his strategy y from Y. Both players do not know what the other player chooses. At the same moment, both players announce what strategy they picked. The players then consult A(x, y) to see who wins, and the winner pays the loser (remember that a positive number means Player II pays Player I, if A(x, y) is negative, then Player I pays Player II). April 16, 2017 7 / 17

Pure and Mixed Strategies Def. The elements of the player s strategy sets X and Y are called pure strategies. These strategies involves no randomness. Def. A mixed strategy is a random combination of pure strategies. For example, a player s strategy might consist of choosing pure strategy x 1 with probability 1/4 and another pure strategy x 2 with probability 3/4. April 16, 2017 8 / 17

Strategic Form Example II Game (Even/Odd) At the same time, both players will say either one or two. These two numbers will be added together, if the sum is odd then player I wins, otherwise player II wins. The winner receives x dollars, where x is the sum of the two numbers chosen. In this case each player only has two strategies: X = {1, 2} and Y = {1, 2}. The outcomes are: ( y = 1 y = 2 ) x = 1 2 +3 x = 2 +3 4 (this is called the payoff matrix, which is a nice way to represent A(x, y)). April 16, 2017 9 / 17

Strategic Form Example: Even/Odd Player I has an advantage in this game. For example, here is one approach where, on average, Player I will not lose money: With probability 3/5, player I picks one, and with probability 2/5 she picks two. (This is a mixed strategy). If player II calls one : then with probability 3/5 player II loses 2 dollars; with probability 2/5 player II wins 3. On average: 2(3/5) + 3(2/5) = 0 If player II calls two : then with probability 3/5 player II wins 3 dollars; with probability 2/5 player II loses 4. On average: 3(3/5) 4(2/5) = 1/5 So on average player I can only win money, not lose money. In fact player I can do even better than this. April 16, 2017 10 / 17

Minimax Theorem Theorem (Minimax) For every two-person zero sum game where the sets X and Y are finite, (1) there is a number V, called the value of the game, (2) there is a mixed strategy for Player I such that I s average gain is at least V no matter what II does, and (3) there is a mixed strategy for Player II such that II s average loss is at most V no matter what I does. Def. A game is fair if V = 0, otherwise it is unfair. Goal. We want a way to find the value of a game, given the payoff matrix, and the corresponding mixed strategy. April 16, 2017 11 / 17

Even/Odd: Optimal Play We return to the Even/Odd example. Let s try to find a way for player I to always win a positive amount, on average, no matter what player II does. We just need to decide on what probability p to choose 1. To simplify things, let us try to find a p so that player I s average winnings is the same no matter what player II does. Such a strategy is called an equalizing strategy. It does not exist for every game. April 16, 2017 12 / 17

Even/Odd: Optimal Play Let p be the probability that player I chooses one. If player II selects one, player I wins on average 2p + 3(1 p) = 5p + 3 If player II selects two, player I wins on average 3p 4(1 p) = 7p 4 In an equalizing strategy, these are equal, so: 5p + 3 = 7p 4 Solving these yields p = 7/12. In this case, 5p + 3 = 7p 4 = 1/12 That is, player I wins on average 1/12 dollars no matter what player II does. April 16, 2017 13 / 17

Even/Odd: Optimal Play We have seen that in the even/odd game player I has a way to ensure she wins 1/12 dollars on average. Similarly, player II has a strategy that ensures he loses no more than 1/12 on average (to see this, repeat the computation from the previous slide form the perspective of player II). This is the value of this game. Because of player I s advantage, this is an unfair game. April 16, 2017 14 / 17

Strategic Form: Not so restrictive It seems that games in strategic form are very restrictive. Both players appear to only take a single turn. Actually many real, complicated games fit this form. For example, chess, tic-tac-toe, go, etc. For example, tic-tac-toe. The strategies for player I, consist of a list of all possible moves that player II can make, and what player I does in response. If both players choose such a strategy before the game starts, the outcome is determined without playing the game. April 16, 2017 15 / 17

Example 2: Equal Game (Equal) Each player picks either one or two. If both players say the same number, the player I wins x dollars, where x is the number both players chose. If both players say different numbers, the player II wins whatever player II said. April 16, 2017 16 / 17

Definition Review Zero Sum game: Player I s gain is equivalent to Player II s loss. Pure strategy: An explicit description of what the player should do in all eventualities. Mixed strategy: A random combination of pure strategies. Equalizing strategy: A strategy where the player s average gain is the same no matter what the opposing player does. Value of a game: A number V such that Player I has a strategy that wins at least V on average, and Player II has a strategy such that Player II loses no more than V on average. Unfair game: A game with value V 0. April 16, 2017 17 / 17