Test 6 Review

Our final exam for Decision Theory/Game Theory will be Thursday, May 17, at 10:15-12:15 in MND 1028 Note the room change.

The final will be somewhat comprehensive. It will cover these topics, techniques, and problems from the semester. You can review your earlier tests for the sort of problems on these topics to expect.

Expected value theorem

Strict dominance

Nash Equilibria

Best Responses

IESDS

Maximin rules

Leximin rules

Minimax regret rule

Maximax and Optimism-Pessimism Rule

Principle of Insufficient Reason

Finding and calculating MSNE for a two person/two move game.

Calculating payouts for PSNE, and calculating payouts for MSNE.

Be prepared to find strictly dominating pure strategies in a two person competitive game.

Be prepared to use Mixed Strategy Dominance and IESDS to solve a two person competitive game.

Be prepared to apply Bayes' Theorem to problems.

Be prepared to explain the problem of unknown priors and how Bayes Theorem allows us to solve it.

Be prepared to guess at a prior probability in a word problem where some evidence is offered, and then use the evidence to revise the application of Bayes' Theorem to "wash out" the unknown prior.

What is the solution to Ultimatum style games? Why? Why does player 2 have to accept such a poor split?Be prepared to solve game tree problems of the sort in Spaniel 2.2, 2.3 using backward induction.

What does game theory predict should be the solution to Centipede games? What do people actually do? Why are they different?

What is Bayesian subjectivism? What is its main hypothesis? It's conceptual basis? The approach for discovering probabilities? Can it address single or unique cases? What are the problems typically associated with it?

Newcomb's Problem

Causal vs. Evidential Decision Theory

Making Threats Credible, Commitment Problems.