This is the second entry in a running, season-long watching brief comparing SharpModels' own ratings against the market, game by game. The first entry set the model's Week 1 numbers before a snap was played; every team now carries one real result, and the ratings below reflect it. This entry lines the model up against the market again for Week 2, then goes game by game through where the two sides disagree and why.
A gap between model and market is worth chasing, not noise to shrug off. SharpModels' ratings are built purely from what happens on the field, so unlike the market they carry no offseason or injury news, only real results. Model probabilities are taken from the SharpModels matchups page; market probabilities are two-sided prices normalised to sum to 100. Away team is shown first, and diff is model minus market on the away side, so a negative number means the market is higher on the away team than the model is.
Week 2 Slate
| Game | Model away | Model home | Market away | Market home | Diff (away) |
|---|---|---|---|---|---|
| Lions@ Bills | 34% | 66% | 33% | 68% | +1 |
| Seahawks@ Cardinals | 77% | 23% | 65% | 35% | +12 |
| Panthers@ Falcons | 34% | 66% | 57% | 44% | −23 |
| Saints@ Ravens | 22% | 78% | 22% | 79% | 0 |
| Vikings@ Bears | 48% | 52% | 33% | 68% | +15 |
| Bengals@ Texans | 30% | 70% | 42% | 59% | −12 |
| Browns@ Buccaneers | 31% | 69% | 22% | 79% | +9 |
| Commanders@ Cowboys | 45% | 55% | 35% | 65% | +10 |
| Jaguars@ Broncos | 40% | 60% | 42% | 59% | −2 |
| Packers@ Jets | 64% | 36% | 64% | 37% | 0 |
| Colts@ Chiefs | 28% | 72% | 28% | 73% | 0 |
| Raiders@ Chargers | 25% | 75% | 28% | 72% | −3 |
| Steelers@ Patriots | 36% | 64% | 33% | 69% | +3 |
| Eagles@ Titans | 80% | 20% | 76% | 25% | +4 |
| Dolphins@ 49ers | 23% | 77% | 12% | 89% | +11 |
| Giants@ Rams | 22% | 78% | 26% | 74% | −4 |
Mean absolute deviation is 6.8 points and the median is 4.0, with 7 of 16 games disagreeing by more than 5. The signed mean is +1.3, a slight lean toward the model being more bullish on away teams than the market this week, not a systematic one.
Match by match
Lions @ Bills Thu, Sep 17, 2026 · 8:15 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Lions | 34% | 33% | +1 |
| Bills | 66% | 68% | −1 |
Detroit needed overtime to beat New Orleans at home in Week 1 (31–30), just barely justifying its 77% favourite tag, while Buffalo won on the road at Houston (36–31) as a market-favoured underdog on the model's own numbers, one of the model's clearest misses of the week. Both teams are now top ten in the ratings (Buffalo 2nd at 1672, Detroit 9th at 1594), and for once model and market barely disagree: Buffalo by roughly two touchdowns of win probability either way.
Seahawks @ Cardinals Sun, Sep 20, 2026 · 4:25 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Seahawks | 77% | 65% | +12 |
| Cardinals | 23% | 35% | −12 |
Seattle opened its season with the closest thing to a statement win, holding off New England at home (13–10, rating up 8 to a league-best 1716), and Arizona pulled a genuine surprise on the road at the Chargers (26–14, rating up 39, the single largest gain of the week). Seattle's rating is still the best in football and the model treats this as close to a blowout on paper (77–23); the market, presumably crediting Arizona's own big Week 1 result, is notably less sure, at 65–35.
Panthers @ Falcons Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Panthers | 34% | 57% | −23 |
| Falcons | 66% | 44% | +23 |
Both teams lost in Week 1: Atlanta on the road at Pittsburgh (13–20), Carolina blown out at home by Chicago (37–59). Both ratings fell as a result, Atlanta to 1435 (−14) and Carolina further to 1360 (−26), still the league's 29th-rated team. That's what makes this the widest and strangest gap on the board: the model, largely unmoved on its preseason view that Atlanta had the better roster, still makes the Falcons a two-to-one favourite at home (66–34). The market disagrees entirely, installing Carolina, the team that just lost by 22, as a 57% favourite instead. Atlanta's own Week 1 storyline was a quarterback change, with Tua Tagovailoa handed the job over incumbent Michael Penix Jr.; if that got off to a shaky start in a losing effort, it would explain a market swing the rating has no way to see, since Elo only ever reads the final score.
Saints @ Ravens Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Saints | 22% | 22% | 0 |
| Ravens | 78% | 79% | 0 |
Nothing to chase here. Baltimore posted the most one-sided result of the week among teams not covered above, routing Indianapolis on the road (41–23) to lift its rating to 1600 (+25, tied for the third-best gain in the league), while New Orleans lost in overtime at Detroit (30–31) and dropped three points to 1402. Model and market read the exact same gap this week: Ravens by roughly four-to-one.
Vikings @ Bears Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Vikings | 48% | 33% | +15 |
| Bears | 52% | 68% | −15 |
Minnesota beat Green Bay comfortably at home (39–22, rating up 24 to 1589) and Chicago blew out Carolina on the road by 22 (rating up 26 to 1547, the league's fourth-biggest riser), two of the best results of the week. The model still treats this as close to a coin flip on the road (48–52), while the market has fully bought into Chicago's performance and made the Bears a 68% home favourite. Elo deliberately caps how much a single game's margin can move a rating, so the market may simply be a week further ahead of the data than the model is allowed to be yet.
Bengals @ Texans Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Bengals | 30% | 42% | −12 |
| Texans | 70% | 59% | +12 |
Houston lost a shootout at home to Buffalo (31–36) and fell 19 points to 1605, still a top-five rating, while Cincinnati won at home over Tampa Bay (33–27) and climbed to 1495. The model, still leaning on Houston's higher base rating, makes the Bengals a 30% road underdog; the market is less convinced at 42%, splitting the two teams' opposite Week 1 trajectories more evenly than the rating does.
Browns @ Buccaneers Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Browns | 31% | 22% | +9 |
| Buccaneers | 69% | 79% | −9 |
Cleveland was thoroughly beaten on the road at Jacksonville (10–34, rating down 15 to 1369), and Tampa Bay lost a competitive road game at Cincinnati (27–33, down 17 to 1469) despite being the better-rated team going in. The model still gives Cleveland a below-average but not hopeless 31% here; the market is harsher at 22%, pricing Cleveland's blowout loss more heavily than the rating currently does.
Commanders @ Cowboys Sun, Sep 20, 2026 · 4:25 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Commanders | 45% | 35% | +10 |
| Cowboys | 55% | 65% | −10 |
Washington lost by two points on the road at Philadelphia (22–24, rating down just 6 to 1443, the smallest drop among Week 1's losers) and Dallas lost by eight on the road at the Giants (20–28, down 22 to 1425, one of the bigger falls of the week despite entering as market favourites). Two struggling teams, but the model still has Washington's underlying numbers ten points of win probability ahead of Dallas's; the market prices the Cowboys as home favourites regardless.
Jaguars @ Broncos Sun, Sep 20, 2026 · 4:05 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Jaguars | 40% | 42% | −2 |
| Broncos | 60% | 59% | +2 |
Jacksonville won convincingly at home over Cleveland (34–10, rating up 15 to 1580), while Denver was routed on the road at Kansas City (10–31, rating down 34, the third-largest fall of the week). Denver's own big offseason move, acquiring Jaylen Waddle in a trade with Miami, hasn't shown up in a raw scoreline yet. Model and market are close to agreement here, both making the Broncos short home favourites, which suggests both sides read the Chiefs loss as more Kansas City's Mahomes-return story than a genuine Denver form issue.
Packers @ Jets Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Packers | 64% | 64% | 0 |
| Jets | 36% | 37% | 0 |
Green Bay lost on the road at Minnesota (22–39, rating down 24 to 1527), and New York pulled the week's biggest straight-up surprise, winning on the road at Tennessee (23–10, rating up 29 to 1342, still the league's 30th-rated team). Despite the gap in direction, model and market both still rate Green Bay's roster well clear of the Jets, at 64% on the road.
Colts @ Chiefs Sun, Sep 20, 2026 · 8:20 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Colts | 28% | 28% | 0 |
| Chiefs | 72% | 73% | 0 |
Indianapolis was blown out at home by Baltimore (23–41, rating down 25 to 1426), and Kansas City throttled Denver (31–10, rating up 34 to 1559, its best single-week gain in years). Patrick Mahomes returning from the torn ACL that wrecked the Chiefs' 2025 season is the obvious explanation for a jump that size, and for once model and market are in lockstep on it: Kansas City a heavy 72–73% favourite either way.
Raiders @ Chargers Sun, Sep 20, 2026 · 4:05 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Raiders | 25% | 28% | −3 |
| Chargers | 75% | 72% | +3 |
Las Vegas pulled the week's other big surprise at home over Miami (27–13, rating up 32 to 1334) on the back of an offseason that brought in Kirk Cousins as a bridge quarterback and Tyler Linderbaum as the league's highest-paid center, while Los Angeles was upset at home by Arizona (14–26, rating down 39, the single largest fall in the league). Despite the Chargers' collapse, both model and market still make them clear favourites here, treating Arizona's win as the surprise rather than a genuine form change from Los Angeles.
Steelers @ Patriots Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Steelers | 36% | 33% | +3 |
| Patriots | 64% | 69% | −3 |
Pittsburgh won at home over Atlanta (20–13, rating up 14 to 1532), and New England lost the season opener at Seattle (10–13, rating down just 8, the smallest drop among Week 1's losing favourites). New England still rates ahead on the raw numbers, and the model reflects that by making the Patriots home favourites regardless of who won last week; the market agrees, just slightly more strongly.
Eagles @ Titans Sun, Sep 20, 2026 · 1:00 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Eagles | 80% | 76% | +4 |
| Titans | 20% | 25% | −4 |
Philadelphia survived at home against Washington by two points (24–22, rating up 6 to 1614, third in the league), and Tennessee was blown out at home by the Jets (10–23, rating down 29 to 1263, the league's lowest rating). Both sides treat this as the most lopsided game of the week, with Philadelphia a three-to-one favourite or better either way.
Dolphins @ 49ers Sun, Sep 20, 2026 · 4:25 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Dolphins | 23% | 12% | +11 |
| 49ers | 77% | 89% | −11 |
Miami lost on the road at Las Vegas in Week 1 (13–27, rating down 32) with Malik Willis starting at quarterback for an offense that lost Tua Tagovailoa and Jaylen Waddle over the offseason, while San Francisco produced the week's biggest result, winning by 20 on the road at the Rams (27–7, rating up 36, the league's second-largest gain). The model already had Miami as a heavy underdog before a snap of 2026 was played; the market has moved even further past it, down to 12%, suggesting the read is that Miami's rebuild is worse in practice than a rating built on one data point currently assumes.
Giants @ Rams Mon, Sep 21, 2026 · 8:15 PM ET
| Team | Model | Market | Diff |
|---|---|---|---|
| Giants | 22% | 26% | −4 |
| Rams | 78% | 74% | +4 |
New York beat Dallas at home (28–20, rating up 22 to 1403), and Los Angeles suffered the week's most one-sided upset loss, blown out at home by San Francisco (7–27, rating down 36 to 1602, still 6th in the league). The Rams remain the model's clear favourite here on the strength of their full-season rating; the market is only a little less convinced, giving the Giants a few extra points of credit for winning last week.