Broncos 20, Jaguars 13: The Field Leaned Upset, Nick Wright Picked It Three Times
Denver beat Jacksonville 20-13 in Week 2, and the result settled 34 graded calls we tracked on this game. The field didn't see it coming — a majority of the picks were on the Jaguars, and every single one of them missed.
The verdict: more voices had the upset than the favorite
Of the 34 graded calls, 19 backed Jacksonville and 15 backed Denver. Every Denver pick hit. Every Jacksonville pick missed. There's no split-decision nuance here — this wasn't a case of some right and some wrong within a side; the side itself was the whole story, and the majority side was wrong.
The score calls that actually landed
Five separate score predictions had Denver winning by exactly three — Carl Dumler (24-21), Keith Cummings (30-27), Bob Morris (23-20), Jared Dubin (24-21) and John Breech (30-27) — and all five graded out at 90.48 accuracy with a skill bump of +16.24 against the naive home-team baseline. The actual margin was seven, so none of them nailed it exactly, but calling a competitive Denver win instead of a blowout or an upset was worth real credit.
Chad Jensen (18-16) and Lance Sanderson (23-21), both with a two-point margin, scored just behind at 88.1 accuracy. Zack Kelberman picked the closest margin of anyone — Broncos by one, 20-19 — but graded lowest among the Denver score-callers at 85.71, a reminder that the model is grading the whole scoreline, not just who covers.
Several pundits also picked Denver outright without a score attached, including Greg Jennings ("I'm gonna take two chips and I'm gonna put it on the Denver Broncos... but I believe in the Broncos"), Jonas Knox, Brady Quinn, Randy McKay, Rich Eisen, Cody Rourke and Sayre Bedinger. All six graded 100 accuracy — and all six scored zero skill. Picking the home favorite is exactly what a naive default would do, so the site credits it as correct but not skillful.
The misses, and how loud they were
Nick Wright backed Jacksonville outright in three separate appearances across three days — "The Jags are better. They went to Denver last year and curb stomped them... They're gonna win outright" — and each instance graded 0 accuracy with a skill hit of -194. Pete Prisco did something similar with a score pick, running "Jaguars (+2.5) 23-20 over Broncos" in three different columns; each landed at 40.48 accuracy and -80.76 skill.
Not every Jacksonville pick was a total loss. Carter Bahns called Jaguars 24-17 — wrong winner, but the seven-point margin matched the real gap almost exactly, just pointed the wrong way, which was enough for 50 accuracy, the best mark anyone picking Jacksonville managed. Richard Sherman (Jaguars 24-16) and Dylan Von Arx (Jaguars 24-20) also came reasonably close on the scoreline while getting the winner wrong.
The most expensive miss belonged to Aniello Piro, whose preseason call in July — "the first game I have them lose is week two to Jacksonville" — was already flagged among the week's worst calls sitewide, at -200 skill. Locking in a loss for Denver two months out cost the maximum the scoring allows.
Whose record this actually moved
Wright's night added three identical zeroes to a game-outcome ledger that already sits at 57.41 accuracy and -7.93 skill across 149 graded picks — not a disaster for him overall (he's stronger in mock-draft calls, where he's hitting 82.26 accuracy), but a rough night specifically in the category where he's already below average.
Prisco's misses here look different in context: his score-prediction accuracy across 40 graded calls sits at 61.8, well above the roughly 40 he posted on this game three times over. This was a bad week inside a category he's generally handled well.
The one that stings most on paper is Tyler Sullivan, who sits second on our overall leaderboard. His Jacksonville upset-alert pick — "going into Denver and pulling out a win isn't foreign to the Jaguars" — graded 0 accuracy and -194 skill, a clean miss inside the one category where he's actually been sharp: 70% accuracy and +58.5 skill across his other 13 game-outcome calls.
Beyond this one game
Season win-total calls we've tracked for the Broncos are hitting just 22.2% accuracy so far on a small sample of nine graded calls, and the Jaguars aren't much better at 21.8% on eleven — both numbers thin enough to treat as early reads rather than verdicts, but they're consistent with a week where the field misjudged both clubs. For the fuller picture on either team, see the Jacksonville Jaguars and Denver Broncos pages, and check the leaderboard to see who's actually earning trust so far this season.