Tie breaking setting in Blotto Game

[shizejin]

Dear all, I am trying to translate your code into Python and Julia in QuantEcon. I have a question about tie breaking in Blotto Game. In your paper, I saw the following sentence:

Each player has a value for each hill that he receives if he assigns strictly more soldiers to the hill than his opponent (ties are broken uniformly at random.)

In your C and python code of Blotto Game, I saw that you are determining winners with uniform probabilities when constructing the game. I wonder if we should use the expected payoff (which is the value of winning times 1/2 for each player) to break the tie?

[ptigwe]

Hi @shizejin,

Thanks for the info. I've updated the code base. The implementation has now been extended to allow for all four tie resolutions:

• favour the first player;
• favour the second player;
• for each tie choose a player to favour and;
• breaking uniformly at random.

[shizejin]

Thanks! We will add tie option into our implementation in Python and Julia following yours.