WNBA 2026 Playoffs
Who wins the 2026 title?
Elo ratings and 20,000 Monte Carlo simulations of the rest of the season, re-run every morning when new results come in. Through games of Oct 31, 2026.
Golden State Valkyries
Champions
won the 2026 title
Correct picks
67.7%
2,000 games, 2019–2026, not used for tuning
Log loss
0.608
vs 0.688 always picking the home team (lower is better)
Games rated
7,023
every game since 1997
Live picks
19–4
recorded before tip-off this season
Playoff odds
Share of simulations in which each team reaches each round. Change is in percentage points since the forecast through Oct 26, 2026.
| Seed | Team | W–L | Elo | Semifinals | Finals | Title | Change |
|---|---|---|---|---|---|---|---|
| 2 | Golden State Valkyries | 32–12 | 1672 | 100% | 100% | ▲ 28 | |
| 1 | Minnesota Lynx | 33–11 | 1574 | 0% | 0% | ||
| 3 | Las Vegas Aces | 31–13 | 1577 | 100% | 0% | ||
| 4 | Atlanta Dream | 30–14 | 1687 | 100% | 100% | ▼ 28 | |
| 6 | Indiana Fever | 28–16 | 1576 | 0% | 0% | ||
| 5 | Washington Mystics | 28–16 | 1530 | 0% | 0% | ||
| 7 | Dallas Wings | 27–17 | 1552 | 0% | 0% | ||
| 8 | New York Liberty | 26–18 | 1569 | 100% | 0% |
Title odds over time
One panel per team, all on the same scale.
Golden State Valkyries100%
Minnesota Lynx0%
Las Vegas Aces0%
Atlanta Dream0%
Indiana Fever0%
Washington Mystics0%
Dallas Wings0%
New York Liberty0%
Show the data
| Through | GSV | MIN | LVA | ATL | IND | WAS | DAL | NYL |
|---|---|---|---|---|---|---|---|---|
| Sep 24, 2026 | 27% | 21% | 10% | 17% | 10% | 6% | 5% | 4% |
| Sep 27, 2026 | 29% | 7% | 18% | 19% | 4% | 7% | 5% | 10% |
| Sep 27, 2026 | 37% | 7% | 15% | 23% | 4% | 3% | 2% | 9% |
| Sep 29, 2026 | 38% | 0% | 10% | 24% | 8% | 3% | 2% | 15% |
| Sep 30, 2026 | 26% | 0% | 11% | 36% | 8% | 0% | 5% | 14% |
| Oct 1, 2026 | 26% | 0% | 20% | 35% | 0% | 0% | 5% | 14% |
| Oct 2, 2026 | 36% | 0% | 17% | 33% | 0% | 0% | 0% | 13% |
| Sep 27, 2026 | 30% | 7% | 18% | 18% | 4% | 7% | 5% | 10% |
| Oct 4, 2026 | 44% | 0% | 10% | 39% | 0% | 0% | 0% | 7% |
| Oct 9, 2026 | 55% | 0% | 0% | 45% | 0% | 0% | 0% | 0% |
| Oct 17, 2026 | 69% | 0% | 0% | 31% | 0% | 0% | 0% | 0% |
| Oct 19, 2026 | 43% | 0% | 0% | 57% | 0% | 0% | 0% | 0% |
| Oct 24, 2026 | 46% | 0% | 0% | 54% | 0% | 0% | 0% | 0% |
| Oct 26, 2026 | 72% | 0% | 0% | 28% | 0% | 0% | 0% | 0% |
| Oct 31, 2026 | 100% | 0% | 0% | 0% | 0% | 0% | 0% | 0% |
Live record
Every prediction on this page is saved before the game is played, so this is the model scored on games it had never seen. So far: 19 of 23 correct, log loss 0.604.
| Date | Game | Result | Pick | |
|---|---|---|---|---|
| Oct 31 | WNBA Finals Game 7 | ATL 63 @ GSV 68 | GSV 55% | ✓ Correct |
| Oct 29 | WNBA Finals Game 6 | GSV 69 @ ATL 80 | ATL 57% | ✓ Correct |
| Oct 26 | WNBA Finals Game 5 | ATL 55 @ GSV 64 | GSV 51% | ✓ Correct |
| Oct 24 | WNBA Finals Game 4 | GSV 67 @ ATL 62 | ATL 59% | ✗ Missed |
| Oct 21 | WNBA Finals Game 3 | GSV 71 @ ATL 97 | ATL 59% | ✓ Correct |
| Oct 19 | WNBA Finals Game 2 | ATL 79 @ GSV 65 | GSV 60% | ✗ Missed |
| Oct 17 | WNBA Finals Game 1 | ATL 64 @ GSV 68 | GSV 58% | ✓ Correct |
| Oct 9 | Semifinals Game 3 | GSV 87 @ LVA 77 | GSV 54% | ✓ Correct |
| Oct 9 | Semifinals Game 3 | ATL 85 @ NYL 83 | ATL 56% | ✓ Correct |
| Oct 7 | Semifinals Game 2 | LVA 81 @ GSV 83 | GSV 66% | ✓ Correct |
| Oct 7 | Semifinals Game 2 | NYL 98 @ ATL 101 | ATL 67% | ✓ Correct |
| Oct 4 | Semifinals Game 1 | LVA 60 @ GSV 71 | GSV 63% | ✓ Correct |
Is the model calibrated?
Parameters were tuned on 1997–2018 only. On the 2,000 games from 2019–2026, games are grouped by the predicted home win chance; a well-calibrated model sits on the diagonal.
Show the data
| Predicted home win | Actual | Games |
|---|---|---|
| 17% | 12% | 40 |
| 25% | 22% | 139 |
| 35% | 31% | 233 |
| 45% | 41% | 308 |
| 55% | 57% | 360 |
| 65% | 64% | 393 |
| 75% | 73% | 309 |
| 84% | 85% | 202 |
| 92% | 88% | 16 |
Home court is shrinking
Home teams won 60% of regular-season games from 1997 to 2019, and 54% since 2021 (2020 was played in a bubble). A model with a fixed home advantage overrates home teams now, so this one sets each season's home advantage from the seasons just before it.
Show the data
| Season | Home win rate |
|---|---|
| 1997 | 62.5% |
| 1998 | 60.0% |
| 1999 | 60.4% |
| 2000 | 59.8% |
| 2001 | 60.2% |
| 2002 | 60.2% |
| 2003 | 62.2% |
| 2004 | 60.6% |
| 2005 | 64.7% |
| 2006 | 60.5% |
| 2007 | 54.3% |
| 2008 | 63.9% |
| 2009 | 64.3% |
| 2010 | 58.8% |
| 2011 | 61.8% |
| 2012 | 58.3% |
| 2013 | 60.8% |
| 2014 | 57.8% |
| 2015 | 61.8% |
| 2016 | 55.4% |
| 2017 | 59.3% |
| 2018 | 54.2% |
| 2019 | 60.8% |
| 2020 | bubble, no home games |
| 2021 | 53.6% |
| 2022 | 54.6% |
| 2023 | 52.1% |
| 2024 | 52.5% |
| 2025 | 55.9% |
| 2026 | 53.9% |
How it works
- Data. Every WNBA game since 1997 (team and player box scores) from the league's stats API, validated and loaded into SQLite. The build checks that every game has two consistent sides and that player points add up to team scores.
- Ratings. Elo, updated after every game with a margin-of-victory multiplier, regressed halfway to the mean between seasons, with a home advantage that adapts as it changes.
- Simulation. Each of 20,000 runs plays out every remaining game from the win probabilities, with ratings updating inside the run, using the real schedule and playoff bracket. Series already under way start from their actual score.
- Automation. A scheduled job fetches new results every morning, rebuilds the database, re-runs the forecast if anything changed, and republishes this page.
Built with Python and SQLite. Data: stats.wnba.com. Source code on GitHub.