Perfect-Season Odds Calculator
An 82-0 season only happens if you win every single night — all 82 games, no back-to-back letdown, no cold shooting slump, no April rest day gone wrong. Set how good a favorite your roster is each game, and this tool shows the odds of running the entire table, plus the math, so you can see exactly why no NBA team has ever come within nine losses of it.
For reference: the greatest teams ever project around 80–88% per night. The 73-9 Warriors and 72-10 Bulls — the two best regular seasons on record — still landed near an 89% single-game win rate, and both lost nine or ten times anyway.
By the numbers: the math in one line
If your roster wins each game with probability p, and you treat the games as independent, the chance of winning all 82 is simply:
P(perfect) = p82
That exponent is the whole story. Because you multiply a number below 1 by itself 82 times, the result collapses fast — this is the same compounding that keeps net rating and win probability honest. To have even a coin-flip (50%) shot at a perfect 82-0 season, you would need to be a 99.2% favorite every single game — a per-night win rate no team in NBA history has sustained across a full 82-game marathon.
The expected-wins reality check
Flip the model around and it gets brutal. Expected wins are just p × 82. A juggernaut that grades out as a +10 net rating team — right at the ceiling the all-time-great rosters live at — projects to roughly an 85% favorite on a typical night. Multiply that across the schedule and the expectation isn't 82 wins. It's about 70, meaning even that historic team is expected to lose around a dozen games. Push the roster to an unheard-of 90% per night and you still project to roughly eight losses. The margin between "greatest team ever" and "perfect" is enormous, and it lives entirely inside that exponent.
Sensitivity table
How the perfect-season odds change with your per-game win rate, over a full 82-game season:
| Win probability per game | Chance of 82-0 | Odds |
|---|---|---|
| 60% | 6.4e-17% | 1 in 1.6 × 10^18 |
| 70% | 2.0e-11% | 1 in 5.0 × 10^12 |
| 75% | 5.7e-9% | 1 in 1.8 × 10^10 |
| 80% | 1.1e-6% | 1 in 8.8 × 10^7 |
| 85% | 1.6e-4% | 1 in 613,265 |
| 90% | 0.018% | 1 in 5,651 |
| 95% | 1.5% | 1 in 67 |
| 98% | 19.1% | 1 in 5.2 |
| 99% | 43.9% | 1 in 2.3 |
Why the real odds are even longer
This calculator assumes a constant win probability and independent games. Real basketball breaks both assumptions, and every way it breaks them makes perfection harder: fatigue and back-to-backs erode a thin roster as the season wears into its second half, injury luck can quietly drop one night's rating, altitude and travel tilt road games, and a single flat shooting night on the second leg of a road trip is exactly the kind of coin flip that ends unbeaten runs. The in-game engine on 82-0 models those forces directly — read the full methodology to see how the net-rating math works, then draft your own five and take them for a run.