I'm creating a fighting game where two opponents face off against each other in the ring. While I've been able to figure out the odds of a player winning based on previous wins/losses, I have yet to find a formula which modifies those odds based on opponent. For example:

  • Player 1: W:5 L:5 - 1/1 odds
  • Player 2: W:5 L:0 - 1/5 odds

I want to calculate the odds that Player 1 will wins against player 2. Compounding this, the players could be of different levels: if the players are within a few levels of each other, the odds should map closely to wins/losses. However, as the levels diverge, the odds of the lower level player winning reduce. As a swag:

Player 1 - W:5 L:5 - 1:1 odds

  • Against a level 8 - 1:2
  • Against a level 9 - 2:3
  • Against a level 10 - 1:1
  • Against a level 11 - 3:2
  • Against a level 12 - 2:1

These are just estimates, my sense is that there is a math formula out there which will calculate that - can anyone out there point me to what this could be?


1 Answer 1


Sounds like http://en.wikipedia.org/wiki/Elo_rating_system might be a good starting point: it's a system which ranks players based on their win/loss record, weighted by their rank. The gap between players' Elo rankings can then be used to calculate the relative odds of each winning.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .