GUIDES · BETSLIP DOCTOR
Why can a team's Pythagorean winning percentage differ from its actual record?
It uses a formula based on runs scored and allowed, while the actual record counts the games the team won and lost. A model-derived estimate and the observed record can differ because they summarize a season in different ways.
For Bettors and cappers learning to read baseball records · Updated
The formula uses runs, not the win column
Pythagorean winning percentage estimates a winning rate from runs scored and runs allowed: RS raised to an exponent divided by RS raised to that exponent plus RA raised to that exponent. MLB describes 2 as the original exponent; it notes that Baseball-Reference uses 1.83.
Because the chosen exponent and provider matter, identify them when reporting an estimate. The team’s official win-loss record remains the count of games actually won and lost.
Keep the estimate in its lane
A runs-based estimate is a descriptive model summary, not a retroactive standings change or a guarantee about another season. Different close-game results can leave the run totals and actual wins pointing to different summaries. Do not present the estimate as a certain future record.
A fictional season with an explicit exponent
Hypothetical 162-game season: a fictional team scores 800 runs and allows 750. Using the original exponent 2, its estimated winning percentage is 800 squared divided by (800 squared plus 750 squared), about .532, or roughly 86 wins over 162 games. The team’s invented actual record is 82–80. The estimated total and the played-game record can differ; the example is not a forecast or real team result.
Related questions
Are Pythagorean wins the official standings?
No. Standings use actual wins and losses. Pythagorean winning percentage is a separate estimate from runs scored and allowed, using a stated exponent.