Free tool · NFL
Model win probability for every team, every remaining week — and the optimized one-team-per-week path that saves your best teams for the weeks that need them.
Teams you've already used (0)
Optimized path · week 1 on
Best distinct-team path through 18 weeks survives them all 1.1% of the time by the model's math.
That number is small because survivor is hard — stacking even strong favorites for months multiplies risk. Honest math, not a defect.
| Wk | Pick | Win prob | Alive thru | Next-best options |
|---|---|---|---|---|
| 1 | LAC vs ARI | 77% | 77% | DET 76% ·wk14JAX 75% ·wk12PHI 74% ·wk9 |
| 2 | LAR vs NYG | 82% | 63% | LAC 79% ·wk1SEA 78% ·wk10HOU 74% ·wk7 |
| 3 | SF vs ARI | 79% | 50% | DET 82% ·wk14SEA 70% ·wk10BUF 68% ·wk18 |
| 4 | CHI vs NYJ | 77% | 38% | BAL 80% ·wk16HOU 73% ·wk7SEA 73% ·wk10 |
| 5 | NE vs LV | 82% | 32% | HOU 74% ·wk7SEA 71% ·wk10DET 69% ·wk14 |
| 6 | IND vs TEN | 74% | 23% | LAR 83% ·wk2NE 82% ·wk5PHI 77% ·wk9 |
| 7 | HOU vs NYG | 81% | 19% | SEA 76% ·wk10LAR 74% ·wk2PHI 73% ·wk9 |
| 8 | CIN vs TEN | 72% | 14% | SEA 75% ·wk10PIT 72%GB 71% ·wk15 |
| 9 | PHI vs NYG | 81% | 11% | SEA 87% ·wk10SF 81% ·wk3KC 76% |
| 10 | SEA @ LV | 80% | 8.9% | BUF 75% ·wk18LAR 72% ·wk2JAX 68% ·wk12 |
| 11 | DEN vs LV | 85% | 7.6% | LAC 79% ·wk1BUF 76% ·wk18KC 74% |
| 12 | JAX vs TEN | 80% | 6.1% | MIA 71%BUF 70% ·wk18IND 70% ·wk6 |
| 13 | MIN vs CAR | 73% | 4.5% | SEA 80% ·wk10DEN 75% ·wk11PHI 70% ·wk9 |
| 14 | DET vs TEN | 83% | 3.7% | SEA 86% ·wk10DEN 74% ·wk11PHI 72% ·wk9 |
| 15 | GB vs MIA | 67% | 2.5% | DEN 75% ·wk11LAR 74% ·wk2BUF 69% ·wk18 |
| 16 | BAL vs CLE | 75% | 1.9% | DET 80% ·wk14NE 71% ·wk5MIN 70% ·wk13 |
| 17 | DAL vs NYG | 69% | 1.3% | SEA 73% ·wk10JAX 69% ·wk12MIN 68% ·wk13 |
| 18 | BUF vs NYJ | 85% | 1.1% | HOU 84% ·wk7KC 77%NE 72% ·wk5 |
An amber ·wk tag on an option means the optimizer is saving that team for a later week — taking it now re-routes the rest of your path (toggle it used after you lock a pick and re-plan).
More like this — daily graded slate + new free tools
In season, get each week's win probabilities and the updated path outlook in your inbox before Thursday kickoff. Check-ins start when the season does, in September.
On the drawing board
Multiple entries, saved plans, and weekly re-optimized paths by email, for one flat $19 — no subscription, no auto-renew. We'll build it if enough people want it: join the waitlist to vote. Nothing is charged today, and the free planner on this page stays free either way.
It shows the model's win probability for every NFL team in every remaining week of the 2026 season, then solves for the best one-team-per-week path — never reusing a team — so strong teams get saved for the weeks that need them. Mark the teams you've already used and it re-plans instantly.
From our team Elo ratings with home-field included — the same numbers as the Matchup Predictor and our graded game picks. Before the season they carry forward from last season's results. Every model pick is graded in public on our accuracy page, losses included.
Because no free, reliable source for survivor pick distribution exists, and we don't fabricate data. Win probability is one honest input into a survivor pick, not the whole answer — in big pools, differentiating from the crowd matters too, and this tool can't see the crowd.
No. Even a well-built path survives a full season only a small percentage of the time — that's the nature of multiplying week after week of risk. The probabilities are model estimates, not promises.