Pythagorean Wins Calculator

Bill James discovered that a team's win percentage tracks remarkably closely with the ratio of runs scored to runs allowed raised to a fixed exponent — not the raw run differential, but scores^exponent divided by (scored^exponent + allowed^exponent). This calculator computes exactly that: Win% = RS^exp / (RS^exp + RA^exp), multiplied by total games to get expected wins. The exponent is what makes the formula sport-specific, and it isn't arbitrary — it reflects how much scoring variance exists in a given sport's games. Baseball uses James's original 1.83; basketball's much higher 13.91 (Hollinger's refinement) reflects that NBA games are high-scoring and low-variance, so small point differentials translate to lopsided win rates; football (2.37) and hockey (2.05) sit in between.

A custom exponent option lets you apply the same framework to a sport or dataset not in the preset list. Beyond the headline win percentage, the calculator reports run/point differential per game and "wins above .500" — the gap between Pythagorean-expected wins and an even .500 record, useful context for the team's actual standing. The real value of this metric is diagnostic: when a team's actual win total diverges sharply from its Pythagorean expectation, that's usually a signal of unsustainable luck in close games (bullpen performance in baseball, clutch shooting in basketball) rather than true team quality, and such teams tend to regress toward their Pythagorean number over a longer sample.