2026 Super Bowl picks
Three picks from three bands of our own simulation: a team from our top three by title odds, a pick from the tier below, and a genuine longshot. Each number is how often that team won the Super Bowl across two hundred thousand simulated seasons — our read, not a claim about anyone’s prices — and all three stay on this page until they settle in February, win or lose.
- Top pickOpen
Seattle Seahawks
23%Maiden gives the Seattle Seahawks a 23% shot at the 2026 Super Bowl — one of our top three teams by simulated title odds.
Two hundred thousand simulated seasons put the Seattle Seahawks on top in 23% of them — which is another way of saying even the best team in football usually doesn't win it.
- Seattle Seahawks: 22.8% to win the Super Bowl in 200,000 simulated seasons — our rank 1 of 32.
- 12.5 wins projected (±2.0), 58% to win the NFC West.
- 93% to reach the playoffs, 36% to win the NFC.
- ContenderOpen
Houston Texans
8%Maiden gives the Houston Texans an 8% shot at the 2026 Super Bowl — our pick from the tier below the favourites.
8% from the tier under the favourites: far enough down our board to be interesting, close enough to matter.
- Houston Texans: 7.7% to win the Super Bowl in 200,000 simulated seasons — our rank 4 of 32.
- 10.9 wins projected (±2.2), 44% to win the AFC South.
- 79% to reach the playoffs, 16% to win the AFC.
- LongshotOpen
San Francisco 49ers
3%Maiden gives the San Francisco 49ers a 3% shot at the 2026 Super Bowl — a genuine longshot, in our 1-3% band.
3% is exactly what it looks like — a small chance, taken deliberately, and we're calling it one.
- San Francisco 49ers: 2.6% to win the Super Bowl in 200,000 simulated seasons — our rank 11 of 32.
- 9.4 wins projected (±2.3), 10% to win the NFC West.
- 52% to reach the playoffs, 5% to win the NFC.
Where these numbers come from
We simulate the whole season — all 272 regular-season games plus the fourteen-team playoff — two hundred thousand times, playing every game through the same calibrated win model behind our weekly board. Each simulated season first draws a season-long strength offset for every team, so the run accounts for a team being better or worse than we rate them. Once games are played, a re-run carries the real results as facts. The title probability is simply how often a team finishes as champion.
Two limits, and they both push the same way. The spread we use for those season-long offsets is solved assuming our errors and the market’s are independent — they are somewhat correlated, so the real spread is wider than the one we use. And a team’s strength is held fixed inside a simulated season, so a midseason quarterback injury never happens in our runs. Both make the top of this board read higherthan reality. We are not claiming these are sharper than the market’s odds — we publish no price to compare against, and the tier word is the claim, not the decimal.
Common questions
- Who does Maiden pick to win the 2026 Super Bowl?
- We publish three: the Seattle Seahawks, the Houston Texans and the San Francisco 49ers. They sit in different bands of our own simulation — one of our top three by title odds, a pick from the tier below, and a genuine longshot in our 1-3% band — each with our championship probability attached, and all three are graded once the Super Bowl is played.
- What are the Seattle Seahawks' chances of winning the 2026 Super Bowl?
- Our season simulation puts the Seattle Seahawks at 23%. That comes from simulating the full 272-game schedule and the fourteen-team playoff two hundred thousand times, using the same calibrated game model we grade every week.
- How are these Super Bowl picks calculated?
- We simulate the whole season — all 272 regular-season games plus the fourteen-team playoff — two hundred thousand times. Each simulated season draws a season-long strength offset for every team and then plays every game through the same calibrated win model our weekly board uses, so the simulation and the graded weekly record come from one place. The title probability is simply how often a team finishes as champion across those runs. Once games are played, a re-run carries the completed results as facts.
- Are these better than the market's odds?
- We don't claim that, and we don't publish anyone's price to compare against. These are our own simulated probabilities and the tier word is the claim. Two known limits both push the top of our board higher than reality: the season-to-season strength spread we solve for is a lower bound, and a team's strength is held fixed inside a simulated season, so a midseason quarterback injury is not modelled.
- Is this betting advice?
- No. Maiden publishes probabilities and grades them in public. Nothing here is betting advice, no outcome is guaranteed, and we don't tell anyone what to back.
Maiden publishes probabilities, not bets. These three picks are graded once the Super Bowl is played.