Chapter 5: Generalizing the Extensive Form: Imperfect-Information Games

• Page number:
  • Section number:
  • Date:
  • Name:
  • Email:
  • Content:

• Page number: 48
  • Section number: 5.3
  • Date: 11/26/08
  • Name: James Wright
  • Email: jrwright@csDELETEthisTEXT.ubc.ca
  • Content: Theorem 5.3.3 should read "In extensive-form games of perfect information, the set of subgame-perfect equilibria is exactly the set of sequential equilibria."
• Page number: 43
  • Section number:5.2
  • Date:7/21/09
  • Name:Kevin Leyton-Brown
  • Content:I cut the following text. It's not wrong, but people were finding it confusing. On balance I think it hurt more than it helped, by focusing very strongly on differences between strategies that do not impact play.
    
    "The difference is substantive, and we illustrate it in the special case of perfect-information games. For example, consider the game of Figure 4.2. A strategy for player 1 that selects A with probability .5 and G with probability .3 is a behavioral strategy. In contrast, the mixed strategy (.6(A,G),.4(B,H)) is not a behavioral strategy for that player, since the choices made by him at the two nodes are not independent (in fact, they are perfectly correlated)."

-- KevinLeytonBrown - 21 Nov 2008