Exact Probability of Battles with Random Components

[manny405]

Presently, the outcome of a battle is win lose or draw. However, outcomes of battle's with random components, for example a mosquito versus a pig(3,1)+otter(1,2) team, can have multiple outcomes. The only way to evaluate these outcomes presently is bootstrapping (i.e. running the same battle many times and collecting statistics). Instead, all possible combinations of outcomes, or battle paths, could be evaluated at a single time to return the exact probabilities of the battle's outcome.

Implementation for this has already started in effects.py. This needs to be connected and built out in battle.py. I am imagining this as being controlled by an argument to the battle class which will turn the feature on or off.

[alexdriedger]

+1 for making it a parameter. All of the use cases I'm using just needs to run the battle once, but I can definitely see how it's helpful for doing more exhaustive searches