About the Pythagorean Win Expectation Calculator
This Pythagorean expectation calculator estimates how many games a team “should” have won based on how many runs or points it scored and allowed. Bill James introduced the idea for baseball, and analysts now use the same formula with different exponents for basketball, football and hockey.
Enter points for and against, games played and the team’s actual wins. The calculator returns the expected winning percentage, expected wins and losses, and the gap between real and expected wins — a quick measure of luck in close games that often evens out later in the season. It is a staple for fantasy players, bettors and fans projecting the rest of a season.
Choose a sport preset or type your own exponent. For baseball, the Pythagenpat option sets the exponent from the run environment ((RS + RA) ÷ G)^0.287, which fits high- and low-scoring leagues better than a fixed exponent.
With the default inputs, the expected winning percentage is 0.545. Change any value above to recalculate instantly.
How to use the pythagorean win expectation calculator
- 1Choose the sport preset, Pythagenpat, or a custom exponent.
- 2Enter total runs or points scored and allowed.
- 3Enter games played and the team’s actual wins.
- 4Read the expected winning percentage and expected wins.
- 5Check “actual − expected wins” to see whether the record looks lucky or unlucky.
Formula and method
The Pythagorean formula estimates winning percentage from scoring alone. Raising points scored and allowed to an exponent x and taking the ratio gives a value between 0 and 1; multiplying by games played gives expected wins. Bill James originally used x = 2 for baseball, and later work found about 1.83 fits MLB slightly better.
Higher-scoring sports with less randomness per point need far larger exponents: Daryl Morey’s 13.91 is the common basketball value and 2.37 is widely used for the NFL. Pythagenpat, developed by David Smyth and the analyst known as Patriot, sets the exponent from runs per game so it adjusts to the scoring environment. Ties are not modelled, so for sports with draws treat the result as points share rather than strict wins.
- RS
- Runs or points scored
- RA
- Runs or points allowed
- x
- Sport-specific exponent
- G
- Games played
Worked examples
MLB team: 750 runs scored, 680 allowed
750^1.83 ÷ (750^1.83 + 680^1.83) = .545. Over 162 games that is 88.2 expected wins, so the team’s 88 actual wins match its run differential almost exactly.
NBA team: 9,430 points for, 9,100 against
With the 13.91 exponent, a +330 point differential gives a .621 expected winning percentage, about 51.0 wins in 82 games. Winning only 44 suggests the team lost a lot of close games.
Pythagenpat for a 60-game season
The run environment is 560 ÷ 60 = 9.33 runs per game, so the exponent is 9.33^0.287 = 1.898. That gives a .568 expected percentage, or 34.1 wins — the team’s 36 wins are about two above its Pythagorean record.
Frequently asked questions
What is Pythagorean expectation?+
It is a formula created by Bill James that estimates a team’s winning percentage from runs scored and allowed: RS² ÷ (RS² + RA²). It is called Pythagorean because the squared terms resemble the Pythagorean theorem.
What exponent should I use for each sport?+
Common choices are 1.83 (or the original 2) for baseball, 13.91 for the NBA, 2.37 for the NFL and around 2 for the NHL. Pythagenpat calculates a baseball exponent from runs per game instead of using a fixed number.
Why does a team outperform its Pythagorean record?+
Teams that win many one-run or close games, or that lose a few blowouts, beat their projection. The gap tends not to persist from one season to the next, so run or point differential is usually a better guide to future results than the win-loss record itself.
How accurate is the Pythagorean formula?+
For a full MLB season the projection is typically within about four wins of a team’s actual total. Accuracy is lower for short seasons or small samples because a handful of close games can swing the record.
What is Pythagenpat?+
Pythagenpat sets the exponent to ((RS + RA) ÷ G) raised to roughly 0.287, so the exponent rises in high-scoring environments and falls in low-scoring ones. It was developed by David Smyth and the analyst known as Patriot.