Thirteen Sunday Games Finished 102.5 Points Below the Closing Totals

Thirteen of Sunday's fourteen games were final by early evening. Add up what they scored and you get 471 points. Add up what the market said they would score and you get 573.5.

That is 102.5 points of daylight across one afternoon, an average of 7.9 points per game, and a scoreboard that went 10-3 to the under.

The arithmetic is ours and it is easy to check. Every final came off ESPN's Week 2 scoreboard at roughly 6:20 p.m. ET, with Colts–Chiefs still in progress and Giants–Rams not played. Every closing total was taken from ESPN's posted Week 2 numbers and confirmed against FOX Sports' odds page, which matched on all thirteen. We are not using a proprietary line or a best-price shop; these are the numbers both outlets printed.

Game	Final	Total	Result
Cowboys–Commanders	30-20 (50)	50.5	Under
Seahawks–Cardinals	31-7 (38)	40.5	Under
Panthers–Falcons	34-3 (37)	43.5	Under
Vikings–Bears	9-3 (12)	47.5	Under
Eagles–Titans	24-20 (44)	39.5	Over
Patriots–Steelers	20-3 (23)	41.5	Under
Packers–Jets (OT)	20-17 (37)	44.5	Under
Browns–Buccaneers	23-19 (42)	41.5	Over
Saints–Ravens	24-17 (41)	45.5	Under
Bengals–Texans	20-6 (26)	45.5	Under
Broncos–Jaguars	20-13 (33)	45.5	Under
Raiders–Chargers	26-14 (40)	43.5	Under
49ers–Dolphins	35-13 (48)	44.5	Over

Include Thursday's Bills–Lions, which went 72 against a 54.5 and was the loudest over of the week, and the fourteen completed games sit 85 points under in aggregate at 10-4.

The number that frames it

Week 1 tied the post-merger record. The league averaged 49.44 points per game across sixteen games and scored 791 in total, matching 2012 exactly on both counts. Week 1 Sunday alone produced 693 points across thirteen games — 53.3 per game, the highest-scoring opening Sunday in NFL history.

Sunday's thirteen completed games produced 471. Same number of games, one week apart, 222 fewer points.

The market did not move from 53.3 to 36.2. It moved to a market-implied 44.1 points per game, which is roughly the midpoint between the record it had just watched and the league's ordinary base rate. That is what a market does when it is uncertain: it splits the difference. The problem is that splitting the difference is only correct if the truth is in the middle, and this week it was not in the middle. It was below both.

We said this in advance, and that cuts both ways

On September 13 this desk published a piece arguing that Week 1's scoring explosion was, functionally, one football game. The headline was that taking out Bears–Panthers left a 1.5-point overage across the remaining thirteen games, and that fourteen completed games had gone 7-7 against the posted numbers with 48.5 of the 50 total points of overage coming from that single 96-point outlier.

Sunday was the follow-through. Chicago, which hung 59 in that record game — the most the Bears have scored since 1980 — scored three points, gained 297 yards, and went 15-for-26 through Caleb Williams with an interception before he pulled up on a fourth-quarter rollout with a non-contact hamstring injury and was carted off. Carolina, the other half of the 96-point game, held Atlanta to three and forced five turnovers.

We want to be honest about what that does and does not prove, because the temptation here is obvious and we have just spent a week telling other people to resist it.

It does not prove scoring is down. Two weeks is not a sample. The strongest version of the argument against this entire piece is that we are doing in one direction exactly what we criticised the "scoring is back" crowd for doing in the other: taking a single Sunday, finding a pattern, and narrating it. Rain at Soldier Field is a real explanation. Two starting quarterbacks leaving with injuries — Williams' hamstring, and Jayden Daniels planting his left elbow awkwardly on the final play of the first half in Dallas, the same elbow he dislocated last season, with Marcus Mariota finishing — depresses scoring in a way that says nothing about league-wide offensive quality. Scott Kacsmar previewed the week on September 19 under the explicit heading of reversion, which means this was not merely foreseeable but foreseen.

All of that is right. So the claim we are actually making is narrower.

The claim

The totals market for Week 2 was anchored to Week 1 rather than to base rates, and the anchoring is visible in the individual numbers rather than in the aggregate.

Look at Vikings–Bears. Chicago closed as a 4.5-point favourite with the total at 47.5, a week after scoring 59. The game produced twelve points. Brian Flores blitzed on 77% of Williams' dropbacks and Will Reichard scored every point Minnesota had. A 47.5 on a Bears home game in September rain is not a number derived from how Chicago's offence had performed over any meaningful window; it is a number derived from the most recent, most vivid, least representative thing that offence had done.

The same shape shows up at Bengals–Texans, 45.5 into a 26-point game, and Saints–Ravens, 45.5 into 41 in a game Baltimore led 17-9 entering the fourth. The three overs were the three games where the total was already low or the favourite was enormous: Eagles–Titans at 39.5, Browns–Buccaneers at 41.5 cleared by half a point, and San Francisco laying 13.5 in a game that went 48.

When the highest numbers on the board are the ones that miss by the widest margins and the lowest numbers are the ones that cash over, that is a signature. It is the signature of a market that adjusted the top of its range upward after a record week and did not adjust the bottom.

What would falsify this

Week 3. If totals come in materially lower next week and the results land near them, the market has repriced and this was a one-week lag — which is ordinary, and which would mean the story here is simply that Week 2 was a correction and we happened to measure it. If Week 3 totals stay in the mid-forties and the games stay in the thirties, the anchoring is persistent and worth taking seriously.

What would genuinely undercut us is if Sunday's under sweep turns out to be driven mostly by the two quarterback injuries and the weather in Chicago. That is testable: strip Vikings–Bears and Cowboys–Commanders out entirely and the remaining eleven games still sit 66.5 points under their closing numbers at 8-3 — 6.0 points per game instead of 7.9. The effect shrinks. It does not disappear.

One more caveat, and it matters. Colts–Chiefs was still being played when we ran these numbers, and Giants–Rams has not kicked off. A 60-point Sunday night game does not change a 102.5-point gap across thirteen others, but the week's final ledger is not the one printed above, and anybody citing this should cite it as what it is: thirteen completed Sunday games, measured at 6:20 p.m. ET.