See the figures below for the data from our in$-$class activity $24.$ where teams took turns taking

flags from a starting pool. Note that in Game $1,6$ out of $9$ matches were won by the first team to take

flags and in Game $2,3$ of $9$ matches were won by the starting team.

$($a$)[10$ points$]$ Using the rollback equilibrium as a predictive model, how many times would you ex$-$

pect the starting team to win when all agents are perfectly informed, rational, and have common

knowledge of rationality? Based on the class data, should we accept or reject this hypothesis?

$($b$)[10$ points$]$ Based on your observations in class and the results from other groups, which of the

assumptions above do you think could be modified to create a more accurate model of this game?

What modifications would you make, and what alternative hypotheses could you test?

$4$You can also find the data on Canvas in the Activities folder of the files tab

Page $7$