Part of the Sports & Games suite · 39 calculators

Pythagorean Expectation Calculator

A team’s expected win percentage from the points, runs or goals it has scored and conceded — Bill James’s Pythagorean expectation — and how many wins luck has added or taken away.

Teams win in proportion to scored^k ÷ (scored^k + conceded^k), where the exponent k is fitted to the sport: about 1.

Results update as you type
Results
Expected win percentage
56.1 %
Expected wins
Actual minus expected
Scored ÷ conceded
Reviewed September 2026. Sporting arithmetic: fixed by the rules of the game, not by where it is played. In the US "batting average" means hits per at-bat — the baseball definition.
No account required · Google Analytics off unless allowedCalculator arithmetic runs in your browserResults update as you type
All calculations run 100% in your browser. The calculator code does not submit your figures to GlobalCalc to obtain a result.
About pythagorean expectation

How the pythagorean expectation calculator works

Teams win in proportion to scored^k ÷ (scored^k + conceded^k), where the exponent k is fitted to the sport: about 1.83 for baseball, 2.37 for American football, 2 for ice hockey, 13.9 for basketball (scores are high and close) and 1.3 for football. A team whose actual record is well above its expectation has been lucky in close games and tends to regress.

Formula: win% = S^k ÷ (S^k + A^k)

Worked examples

InputsExpected win percentageNote
Baseball: 800 for, 700 against, 162 games56.1 %56.1% — about 91 wins
Basketball: 9,000 for, 8,700 against61.6 %61.6%
Hockey: 250 for, 250 against50 %50%

Frequently asked questions

Why does basketball have such a big exponent?

Because scores are high and every game is decided by a small share of the points, a small edge in scoring margin turns into a big edge in wins. Baseball’s 1.83 reflects low, variable scores where a run means a lot.

Is the exponent fixed?

It is fitted from history and varies a little by era and league; some versions make it depend on the total scoring (Pythagenpat). The values here are the widely used ones; use the custom option for your own league.

What does “luck” mean here?

The gap between actual and expected wins is mostly the record in one-run or close games, which is largely random. Teams several wins above expectation tend to fall back the next season.

Does it work for football (soccer)?

Less well, because draws exist and scores are low; the 1.3 exponent gives a rough win share. It is better used for points-per-game style comparisons than for exact predictions.

Where these figures come from

Last checked: September 2026. Formulas follow the governing bodies where one publishes them — the World Handicap System for golf, the ICC playing conditions for cricket run rates.