I have two combatants, A and B. Each has three stats (X,Y,Z). Each round of combat, they make three choices, and roll a random value. All of this determines the outcome of the battle, which can be a win for A, a win for B, or a tie. I know the attributes of A, and want to work out the attributes of B given that I want A to win N% of the time. Right now my method for doing this is to pick starting values for B's stats, then run 10,000 simulations in which the three choices by both combatants are (mostly) random each round and figuring out what the odds of A winning are, then changing the stats of B and trying again, narrowing in on the wanted odds of winning. Needless to say, that's going to take a long time, especially as all of those bits (the three stats, the three choices) are independent of one another. I have, however, run a large number of such 10,000 trial simulations with various stats at X, Y, and Z compared to a fixed competitor (randomly selecting the three choices each round for both and randomly rolling) to see how often wins and ties happen. Can I use this to somehow work out a way to calculate the three stats involved instead of running dozens of simulations to see what happens?
EDIT: I can provide more details about how the combat works if that would be helpful to finding a better solution.
EDIT: I can provide more details about how the combat works if that would be helpful to finding a better solution.