Let (X1, Y1) and (X2, Y2) represent two random observations from a bivariate distribution. The concordance probability, K, is the conditional probability that Y2 is not less than Y1, given that X2 is not less than X1. See also coefficient of concordance.