Frozen 2026-08-28 · immutable17 over · 15 under · graded at season's end
2026 NFL Win Totals
Our number vs the market, one pick per team. The board was frozen before the season and does not change; every pick grades in January against actual regular season wins.
Each row: the team, the pick with its gap (our projected wins minus the line), our number against the book's line, the approximate chance the over hits, then a bar for the gap and what the pick needs. Rows marked * are where the book moved after our freeze; the fine print below has the details.
AFC East
BUFUNDER −0.310.2 vs 10.545%
o −120 · u +1000-0 · needs 7 losses of 17
NEOVER +0.610.1 vs 10.5*58%
o +115 · u −1400-0 · needs 10 of 17
DraftKings' line is 10.5; our pick was frozen against 9.5 and grades against that number. At the current line the model would sit on the other side. The pick stands as frozen.
NYJOVER +1.26.7 vs 5.567%
o −120 · u +1000-0 · needs 6 of 17
MIAOVER +2.86.3 vs 3.580%
o −140 · u +1100-0 · needs 4 of 17
AFC North
BALUNDER −0.310.2 vs 11.5*46%
o +115 · u −1400-0 · needs 7 losses of 17
PITOVER +0.89.3 vs 8.561%
o +110 · u −1300-0 · needs 9 of 17
CINUNDER −1.78.8 vs 10.528%
o +115 · u −1400-0 · needs 7 losses of 17
CLEOVER +1.87.3 vs 5.572%
o −125 · u +1050-0 · needs 6 of 17
AFC South
HOUOVER +1.410.9 vs 9.570%
o −146 · u +1240-0 · needs 10 of 17
INDOVER +0.47.9 vs 7.555%
o −140 · u +1150-0 · needs 8 of 17
JAXUNDER −1.77.8 vs 8.5*28%
o −140 · u +1160-0 · needs 8 losses of 17
TENUNDER −0.26.3 vs 6.547%
o +100 · u −1200-0 · needs 11 losses of 17
AFC West
DENOVER +0.910.4 vs 9.563%
o −115 · u −1050-0 · needs 10 of 17
LACUNDER −1.28.3 vs 9.533%
o −140 · u +1150-0 · needs 8 losses of 17
KCUNDER −2.77.8 vs 10.518%
o +115 · u −1400-0 · needs 7 losses of 17
LVOVER +0.76.2 vs 5.560%
o −150 · u +1250-0 · needs 6 of 17
NFC East
PHIUNDER −0.510.0 vs 10.544%
o +115 · u −1400-0 · needs 7 losses of 17
DALUNDER −1.28.3 vs 9.535%
o +100 · u −1200-0 · needs 8 losses of 17
NYGOVER +0.47.9 vs 7.555%
o −110 · u −1100-0 · needs 8 of 17
WASUNDER −0.17.4 vs 7.549%
o −105 · u −1150-0 · needs 10 losses of 17
NFC North
DETOVER +0.010.5 vs 10.550%
o −115 · u −1050-0 · needs 11 of 17
MINOVER +0.89.3 vs 8.560%
o −120 · u +1000-0 · needs 9 of 17
GBUNDER −1.38.2 vs 9.533%
o −130 · u +1100-0 · needs 8 losses of 17
CHIUNDER −1.48.1 vs 9.531%
o +105 · u −1250-0 · needs 8 losses of 17
NFC South
TBOVER +0.08.5 vs 8.551%
o +115 · u −1360-0 · needs 9 of 17
NOOVER +0.78.2 vs 7.559%
o −120 · u +1000-0 · needs 8 of 17
ATLOVER +0.37.8 vs 7.554%
o +120 · u −1400-0 · needs 8 of 17
CARUNDER −1.16.4 vs 7.535%
o +115 · u −1360-0 · needs 10 losses of 17
NFC West
LAUNDER −0.710.8 vs 11.540%
o −130 · u +1100-0 · needs 6 losses of 17
SEAOVER +0.310.8 vs 10.554%
o −115 · u −1050-0 · needs 11 of 17
SFUNDER −1.78.8 vs 9.5*29%
o −146 · u +1200-0 · needs 7 losses of 17
ARIOVER +3.16.6 vs 3.585%
o −136 · u +1150-0 · needs 4 of 17
Fine print
Our number is the mean projected win count from the preseason season simulation, run over the same margin model the rest of the site uses. It is static: it will not update until 2026 games complete, and this frozen board never updates at all. The percentage column is an approximate chance of the over, read off the simulation's win distribution; treat it as a rough number, not a price.
Lines and prices are DraftKings' own season win-total markets, captured August 28, 2026. Books move. Four teams (NE, BAL, JAX, SF) moved a full point between our freeze and that capture, and each carries the * marker. The picks themselves were frozen on August 28 against a multi-book consensus from VegasInsider: the median of three books, with a fourth column excluded because it carried the same uniform line on every team, which is a different market. Book identities in that table were header images, so the consensus is recorded by column position rather than by book name. New England is the one team where the line move would have flipped the pick; its row says so instead of quietly changing.
Scope, stated before someone corrects it: these are season win totals, a count of games a team wins. They are not the per-game point totals this site declines to publish; that restriction covers the game scoring model, which has a known defect, and does not apply to a wins market built on the season simulation.
This market has no graded history on this site, so this board is a learning record, not an edge claim. The picks are the sign of model minus line, nothing more. Grading is against the frozen line, and a tie is not a win.
