A statement from Ponssard and Zamir "Zero-sum sequential games with incomplete information" at the bottom of the first page reads

AM's result is for zero-sum two-person games in which the two players move simultaneously. This includes of course games in which the moves are made sequentially: Player I moves first, player II is informed of player I's move and then makes his own move. (Such a game can be looked at as a simultaneous game after redefining the pure strategies of player II.)

I am curious about the last statement in parentheses. How does one "redefine" the strategies of the follower to make the game simultaneous move?

  • 2
    $\begingroup$ I think they simply mean that you can rewrite an extensive-form game with a strategy like "first play A then play L if player 2 played X and play R if player 2 played Y" in matrix form and call this strategy ALR. That is, simply redefine the game such that the simultaneous-move action is a simultaneously set contingent plan of actions. $\endgroup$ – Bayesian Jun 4 at 16:32

This is basically the idea that any extensive form game has a normal form representation.


  • 1
    $\begingroup$ Just a general well-meant comment: This community prefers more detailed descriptions of contents behind links that are likely temporary, but I guess it's alright in this case. $\endgroup$ – Bayesian Jun 4 at 18:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.