Let the elements in a payoff matrix be denoted a_(i j), where the is are player A's strategies and the js are player B's strategies. Player A can get at least min_(j<=n) a_(i j) for strategy i. Player B can force player A to get no more than max_(j<=m) a_(i j) for a strategy j. The best strategy for player A is therefore max_(i<=m) min_(j<=n) a_(i j), and the best strategy for player B is min_(j<=n) max_(i<=m) a_(i j).

game | game saddle point | payoff matrix | strategy | value