Broncos 30, Rams 26: Picking Denver Wasn't the Hard Part — Nailing the Margin Was
Denver beat the Rams 30-26 at home in Week 3, and the pregame record on this one settled 21 graded calls — the widest one-game slate we've covered so far this season for a single matchup. The headline isn't that the field got the winner right. It's that getting the winner right turned out to be the easy, low-value part.
The verdict: leaned Broncos, cashed Broncos
Across the 19 winner and score calls we tracked, 13 picked Denver and 6 picked Los Angeles. That's not unanimous — six pundits went out on the Rams — but it's a clear lean, and the lean paid off. Two more calls (both about rookie back Jonah Coleman starting Denver's backfield) had nothing to do with the winner and missed on their own terms; Coleman didn't take a snap.
Here's the catch: because Denver was the home team, picking the Broncos to simply win was the site's naive default. Every one of the ten pundits who made a plain winner pick on Denver — Tracy Sandler, Carmen Vitali, Kevin Wildes, Chris Brockman, Aniello Piro, John Breech, Cody Rourke, Sayre Bedinger, O'Shea Jackson Jr. and Gavin Carlson — graded a perfect 100 accuracy and exactly zero skill. Correct, but no better than the coin the site already flips for the home side. Carmen Vitali put real reasoning behind hers before kickoff — "Road favorites. I don't buy it. I don't buy it. I think the Denver has figured some things out" — which reads sharper than a shrug, even if the scoring treats it the same as everyone else's.
Who actually beat the baseline
The real separation showed up in the score predictions, where three pundits went beyond "who wins" to "by how much" — and got rewarded for it. Colin Cowherd's show called it Denver 24-21 ("I got Denver pulling out a win at home, 24-21"), missing the final margin by one point and posting the best grade of the game: 97.62 accuracy, +11.39 skill. Jordan Dajani explicitly framed his pick as a road dog upset — "With the game being at Mile High in prime time, give me the Broncos in an upset... Prediction: Broncos 26, Rams 24" — and John Breech's separate score call ("PICK: Broncos 23-21 over Rams") landed the same grade as Dajani's: 95.24 accuracy, +6.89 skill apiece. Those three didn't just win the coin flip the way the plain winner-pickers did; they beat the field's own baseline for how big a Denver win should be.
On the other side, three pundits backed the Rams to cover a real margin and paid for it. Pete Prisco and Jared Dubin both called Rams wins by 7 (30-23 and 24-17), and both graded identically at the bottom of the game: 42.86 accuracy, -92.11 skill. Tyler Sullivan's Rams-by-3 call (23-20) graded a little less badly at -83.11 skill, still a heavy loss for backing the wrong side by design.
The single worst take on the board
Nobody had a rougher night than Richard Sherman's pick on the NFL Network set: "This secondary, along with that, I got the Rams winning this game. In a defensive battle of sorts... I'd say 17-20." Denver scored 30. That call graded 0 accuracy and -189 skill, the single worst mark this game produced. It's also a step down from Sherman's own norm — his game-outcome picks run 73.95 accuracy across 42 graded calls, and his season-win-total and conference-champion calls have actually been above-average for him. This one didn't extend a slide; it's an outlier against a record that's mostly been decent when he sticks to naming winners.
By contrast, Colin Cowherd's +11.39 here is a clean outperformance of his own history in this category — his score predictions average just +1.65 skill across 77 graded calls, making this one of the sharper individual reads of his season. Pete Prisco's miss, meanwhile, sits well below his own típica line: his score predictions average +14.74 skill across 55 graded calls, so -92.11 here is a genuine outlier low, not a pattern.
What it touches beyond one game
Denver's win totals have run hot before — the field collectively undersold the Broncos' 2025 season by 4.9 wins across seven tracked calls, and this result keeps that same directional story going into 2026: pundits still adjusting to a Denver team that keeps beating its own number. The Rams' loss, meanwhile, lands on a club that pundits have picked to win 68.5 accuracy on straight game-outcome calls this season across everything we've tracked on them — nothing that one Week 3 loss on the road undoes, but a data point worth filing before the next big swing on either club's division or playoff outlook. For the season-long view on both teams, see the Broncos and Rams pages, and check the leaderboard for who's actually earning trust across a full slate rather than one Sunday.