What are Pythagorean wins in the NBA?
The same Bill James idea ported to basketball with an exponent of 14 (Basketball-Reference's standard): W = G × PF^14 / (PF^14 + PA^14). Because basketball scores are high and consistent, the formula is remarkably sharp — point differential almost always beats record at predicting future wins.
How the calculator works
Enter points for and against over any span, add actual wins, and the tool reports the luck gap — the wins a team has banked beyond its differential.
Common use cases
- Mid-season over/under analysis
- Fantasy: buy players on unlucky teams, sell the lucky ones
- Debating which conference contenders are real
Formula & caveats
Clutch-time performance and blowout management create real, persistent overachievers — but they are rarer than fans hope. Sustained 5-win luck is a regression flag.
Frequently Asked Questions
Why exponent 14?
It empirically fits NBA win totals to point differential; Basketball-Reference uses it for SRS-era analysis.
Is it better than win pace?
For prediction, yes — differential is more stable than record over small samples.
Can a team be truly 'clutch'?
Some maintain positive close-game records for seasons, but most extreme overachievers regress.
Privacy & Your Data
StatPlay runs entirely in your browser. The numbers you enter never leave your device — there is no server, no account and no upload. We store only a single local preference flag when you dismiss the cookie notice.