Skip to content
NFL EDGES

Live game edges β€” model vs. Kalshi, ranked

Gridiron Edge Β· 5-min refresh

NFL 2026

Power rankings, futures & win totals β€” all 32 teams

COMBO EDGE

Correlation-aware Kalshi combos β€” free

Home/NFL/Schedule/Predictor
Gridiron Edge272 games Β· 79 played Β· Week 6 on the board

NFL Schedule Predictor 2026: Every Game, Every Week Picked

Model as of Β· prices live

The model’s straight-up winner and win probability for every game on the 2026 schedule β€” this week first, then all 18 weeks β€” and where the season lands: projected record and playoff odds for all 32 teams from a 10,000-season simulation. Tap any game for its model line, the Kalshi price and the edge between them.

Quick Answer

As of October 15, 2026, across all 272 regular-season games the model's straight-up pick is the home team 195 times and the road team 77 times, with 54 games inside 55/45 on the model; of the 47 games Kalshi has priced so far it sides with the market favorite in 44 and calls 3 upsets.

This week (Week 6, 14 games): 11 market favorites and 1 upset straight up β€” biggest upset call Panthers over Eagles at 54% (Kalshi 49.5Β’).

Strongest call still to play: Rams over Cardinals in Week 6 at 87%.

Season simulation: San Francisco 49ers lead the league at a projected 12.2–4.8 (89% to make the playoffs) over 10,000 simulated seasons.

NFL Week 6 predictions β†’

This week on the boardNFL Week 6 predictions β†’Week 7 predictions β†’Week 5 results β†’

Week 6 gamesSeahawks at BroncosTexans at JaguarsBears at FalconsRavens at BrownsTitans at ColtsJets at PatriotsSaints at GiantsPanthers at EaglesSteelers at BuccaneersCardinals at RamsChargers at ChiefsBills at RaidersCowboys at PackersCommanders at 49ers

This week's straight-up winners

Kalshi has priced 12 of these 14 games and our published probability is anchored to the market on each, so the straight-up pick matches the market favorite by construction β€” that agreement is the method, not a finding. Widest disagreement this week: our own ratings line sits 4.0 points closer to the Eagles than the market's, a read we grade separately and never publish as an edge. 2 games are held until both starting quarterbacks are confirmed, so they carry no call yet.

Winner = the side with the higher published win probability. Once Kalshi prices a game that probability is market-anchored β€” it is the market’s number, which is why the Model and Kalshi columns sit within a point of each other and why no upset is labelled. Before a market opens it is the model’s own read. The model’s separate opinion is the lean, in points, graded on its own and never published as an edge. Kalshi = the mid price of the winner’s contract, in cents. A straight-up call is not the play β€” the play on each game is the edge side on its matchup page.

Full board with the play and the edge on every game: NFL Week 6 predictions Β· NFL Week 7 predictions

Go Pro Β· 26% off your first year

Every NFL game priced against the live Kalshi line, every week β€” $150/yr

Pro is the layer around the free game edges: fair value against the live Kalshi price on every market, edge alerts in Discord the moment a gap opens, and the full win-total, championship and projection boards.

Go Pro β†’

$111 for the first year, then $150/yr. Code applies automatically. Ends Dec 31. Monthly is $14.99/mo.

Every signal graded in public, losers left on the board. See the record β†’

Every week, every game picked

All 272 regular-season games, winner and model win probability. Weeks already played show the call the model published before kickoff β€” no result is claimed here. Kalshi prices fill in as each week’s markets list.

Week 1played16 pre-game callsShow picks β–Ύ
Week 2played16 pre-game callsShow picks β–Ύ
Week 3played16 pre-game callsShow picks β–Ύ
Week 4played16 pre-game callsShow picks β–Ύ
Week 5played15 pre-game callsShow picks β–Ύ
Week 6this week11 favorites Β· 1 upsetShow picks β–Ύ
Week 711 favorites Β· 0 upsetsShow picks β–Ύ
Week 814 games Β· not priced yetShow picks β–Ύ
Week 915 games Β· not priced yetShow picks β–Ύ
Week 1014 games Β· not priced yetShow picks β–Ύ
Week 1113 games Β· not priced yetShow picks β–Ύ
Week 1216 games Β· not priced yetShow picks β–Ύ
Week 1314 games Β· not priced yetShow picks β–Ύ
Week 1415 games Β· not priced yetShow picks β–Ύ
Week 1516 games Β· not priced yetShow picks β–Ύ
Week 1616 games Β· not priced yetShow picks β–Ύ
Week 1716 games Β· not priced yetShow picks β–Ύ
Week 1816 games Β· not priced yetShow picks β–Ύ

Winner = the side with the higher DAEPA model win probability. Kalshi = the mid price of that side’s contract, in cents, once the market lists. An upset is a model winner priced under 50Β’ β€” that is a straight-up call, not the play; the play on each game is the edge side on its matchup page.

Season simulation: projected records and playoff odds

Full standings & playoff odds β†’

Where the schedule lands: the mean of 10,000 simulated seasons on the same power ratings, with 79 played games fixed as results. Picked = games the predictor above hands the team straight up (a count of picks, not a record).

Proj W–L = simulated mean wins and the losses that leaves over 17 games. Playoffs = share of simulated seasons reaching the postseason by any route. Seven teams per conference qualify β€” the gold line marks the seventh seed. Projections, not results.

Every Thursday before kickoff: the week's picks, our number vs Kalshi, and the season outlook. Free.

No spam. Unsubscribe anytime.

NFL schedule predictor 2026 β€” frequently asked

How does the NFL schedule predictor pick each game?

Every game runs through DAEPA (Defense-Adjusted Expected Points Added), the play-by-play engine behind our PWR power ratings β€” PWR is points per game above an average team on a neutral field, PWR = Off PR + Def PR + ST PR. The rating gap plus home field gives an expected margin, the margin gives a home win probability, and the straight-up winner is simply the side above 50%. The projected score splits that margin around a ratings-implied total (league baseline 42.58 points plus both offensive ratings, minus both defensive ratings). Backtested on 3,266 games from 2013–2024, DAEPA picks the straight-up winner 59.6% of the time.

How often does the predictor update?

Power ratings refresh weekly once games are played, so every unplayed game on this page re-prices after each week; the current week also re-runs against the latest Kalshi prices through kickoff. The stamp under the title is the newest model or simulation run on the page β€” never the time the page was built. Games already played keep the pre-game call they were published with, labelled as such, rather than being rewritten after the fact.

What is the difference between the straight-up winner and the play?

The straight-up winner is the side with the higher model win probability β€” the team the predictor expects to win. The play is different: it is the side where the model probability and the Kalshi price disagree by 5 points or more, and it is routinely the underdog. A row tagged Upset is a straight-up winner priced under 50Β’ on Kalshi. Each game links to its matchup page, which shows the play, the model line and the edge against the market.

What is the 2026 season simulation?

The season table is the mean of 10,000 simulated seasons run on the same power ratings. Each simulated season plays the full 17-game schedule with home field, rating uncertainty and the real NFL tiebreaker procedure, and games already played are fixed as results (79 so far). Projected W–L is the average wins across those seasons and the losses that leaves; Playoffs is the share of seasons in which the team reached the postseason. "Picked" is a different number: how many games on the schedule the game-by-game predictor above hands that team straight up β€” a count of picks, not a projected record.

More NFL model boards

Keep exploring

Prediction markets carry risk. Contracts can settle at $0. Informational only β€” not financial advice. Trade responsibly.