Chiefs 31, Broncos 10: 31 of 45 Pre-Game Calls Took Denver, and the Best Score Pick Was Off by 18
Kansas City beat Denver 31-10 on Monday night to open the 2026 season. That result settled 45 graded predictions we had on file — straight-up winner picks and exact-score calls, made anywhere from late July to the Sunday before kickoff.
The field was not close to right. Of those 45 calls, 31 took Denver and 14 took Kansas City. Average accuracy across the whole set came to about 30 out of 100.
And here is the part that stings more than the split: being right paid almost nothing. In our scoring, a correct pick is measured against the naive default of taking the home team, and in this game the home team was Kansas City. So the 11 straight-up Chiefs picks that scored a perfect 100 earned exactly zero credit for skill — they matched the default rather than beating it. The only calls in the entire set that did better than the default were three score predictions, and they got there by a hair.
The three calls that beat the default
Pete Prisco had Chiefs 20, Broncos 17 — in his CBS Sports Week 1 column and again on game week, where he was the lone expert in the roundup on Kansas City. Mike Florio had the same number: "Chiefs 20, Broncos 17."
Each of those calls graded at 57 accuracy and a small positive skill mark (+10.42) — right side, right home team, margin of three against an actual margin of 21. Eighteen points off, and still the best anyone did on this game.
That is the cleanest way to say what the field missed. Nobody predicted a Chiefs win by more than three points. The biggest margin anyone forecast in this game was 10 — Matt Hamilton at Broncos 27-17, and Cody Rourke, who said Denver would "win by at least 10 points". Both had the wrong team.
Prisco's score-prediction record now runs 61.16 accuracy over 17 graded calls, Florio's 61.46 over 16. Both are early records rather than long ones, so read them as a lean, not a verdict — but in this game they were the only two people whose numbers earned anything.
The misses, with the quotes
The Denver side of the ledger was loud and specific.
Shannon Sharpe took the Broncos with Nik Bonitto down for 2.5 sacks: "I got the Broncos. They game, too... It ain't going to beat that defense the Broncos got." Danny Parkins went five chips on Denver — "Another upset. Outright winner for Denver." Eric Mangini hedged the stake but not the side: "as much as I wanna go with Patrick, because I love him, I'm going with Denver too, but only with two." Tyler Sullivan flagged Denver +2.5 as an upset candidate and separately called it 24-20 Denver, noting he was 13-2 on the week. Conor Orr, Albert Breer, Mitch Goldich, Gilberto Manzano and Eva Geitheim all had Denver in their straight-up grids. So did Nick Ferguson, Aniello Piro, Sayre Bedinger and Leonard Fournette. Every one of those winner picks graded zero, and each landed below the home-team default.
Jared Dubin printed Broncos 21-20 three separate times — 2.38 accuracy apiece, the lowest score-prediction grade in the set. Jordan Dajani had Broncos 17-14 twice. Richard Sherman put it at 21-13 Denver — though in his AFC preview a few days earlier he had asked whether Denver could "sneak out of there with a win in Arrowhead? Maybe, but probably not with Patrick Mahomes healthy," which graded as a hit.
Two records this game actually moved
John Breech is the most interesting case, because he had it right first. In August he wrote that "it's hard to pick against a Chiefs team that will be opening the season at home in primetime" — a hit. By Labor Day he had Broncos 26-20, and by game week Broncos 23-21, twice. Four Denver calls, one Chiefs call, and the Chiefs call was the old one. His record overall is still short — 37 graded calls — but it reads rough: 41.73 accuracy, with game picks (38.89) and score picks (25.15) both well below the naive baseline.
Nick Wright was the other pole. He said the Chiefs would win this game four separate times across late summer — "Week one, Monday Night Football with the stadium on fire for Patrick's return against the Broncos. Win" is the tidiest version — and all four graded as hits. That is a streak extended, not started: Wright has 418 graded calls with us at 57.02 accuracy, though his game-outcome line (57.55 accuracy, and a skill mark of −7.03) says his winner picks have historically run a touch behind the default.
What it touches next
This is one game, and the season has answered almost nothing else yet. But there is a lot riding on Denver: we have 28 calls on the Broncos winning the AFC West still open, including a CBS Sports staff vote projecting a repeat, against 12 open calls for Kansas City. The Denver win totals we've logged run from 10 wins up to 13-4 (Rourke and Bedinger, in July). Greg Auman was already out the next day saying the division "goes through Kansas City again."
Worth remembering while those grade out: Kansas City was the field's single biggest collective win-total miss of the 2025 season, 6.3 wins off across 13 calls, with Denver among the next-largest at 5. This division has been punditry's blind spot two years running.
Follow it on the Chiefs and Broncos pages, or see who is actually beating the baseline on the leaderboard.