How accurate are UFCalendar's MMA fight predictions?
How the model's pre-fight probabilities actually fared, graded against every decided fight in the database. Nothing is cherry-picked: these are the same numbers shown on bout pages before each fight, recomputed live.
UFCalendar's Power Index has correctly picked the winner in 62.6% of the 19470 MMA fights graded so far, with a Brier score of 0.239. It is well calibrated: when the model gives a fighter a 50%-55% chance, that favorite wins 56.9% of the time.
- Favorite accuracy
- 62.6%
- Fights graded
- 19470
- Brier score
- 0.239
When the model says 70%, does the favorite actually win about 70% of the time? Predicted vs. observed, by confidence bucket.
| Model says | Avg predicted | Favorite won | Fights |
|---|---|---|---|
| 50%–55% | 52.2% | 56.9% | 6874 |
| 55%–60% | 57.4% | 58.4% | 5285 |
| 60%–65% | 62.4% | 66.1% | 3773 |
| 65%–70% | 67.3% | 72.3% | 2241 |
| 70%–75% | 72.2% | 78.2% | 1330 |
| 75%–80% | 77.2% | 78.1% | 570 |
| 80%–100% | 82.9% | 82.4% | 267 |
| Year | Fights | Accuracy |
|---|---|---|
| 2026 | 1023 | 61.1% |
| 2025 | 1815 | 60.6% |
| 2024 | 1775 | 61.7% |
| 2023 | 1779 | 63.1% |
| 2022 | 1496 | 64.2% |
| 2021 | 1439 | 63.1% |
| 2020 | 1051 | 64.1% |
| 2019 | 1463 | 60.4% |
| 2018 | 1250 | 61.8% |
| 2017 | 1049 | 63.1% |
| 2016 | 955 | 64.5% |
| 2015 | 780 | 60.8% |
| 2014 | 754 | 65.1% |
| 2013 | 504 | 64.9% |
| 2012 | 407 | 61.4% |
| 2011 | 335 | 61.2% |
| 2010 | 275 | 63.6% |
| 2009 | 230 | 60% |
| 2008 | 241 | 64.3% |
| 2007 | 195 | 66.2% |
| 2006 | 171 | 65.5% |
| 2005 | 92 | 70.7% |
| 2004 | 46 | 50% |
| 2003 | 39 | 59% |
| 2002 | 52 | 67.3% |
| 2001 | 38 | 50% |
| 2000 | 42 | 71.4% |
| 1999 | 39 | 59% |
| 1998 | 23 | 69.6% |
| 1997 | 29 | 65.5% |
| 1996 | 35 | 62.9% |
| 1995 | 27 | 70.4% |
| 1994 | 18 | 77.8% |
| 1993 | 3 | 66.7% |
Every number on this page is recomputed live from the stored pre-fight probabilities, the same ones printed on bout pages before each fight happens. Draws are excluded (no verdict to grade) and dead-even 50/50 prices grade no pick. Ratings and probabilities span every organization we track, back to 1993.