Patriots 20, Steelers 3: Everyone Picked New England, Nobody Called the Blowout

The Patriots beat the Steelers 20-3 at Gillette Stadium in Week 2, and the result settled 18 graded calls we'd tracked heading into kickoff — 10 outright winner picks and 8 score predictions. The field leaned hard toward New England, and the lean paid off. But not one of the pundits who put a number on it saw anything close to a 17-point margin.

The verdict: lopsided, not unanimous

Fifteen of the 18 calls backed the Patriots to win. Three went the other way — Nick Shook, John Middlekauff and Colin Cowherd all took Pittsburgh. That's not a unanimous field, but it's close to it, and the majority was right about who'd win.

Here's the catch: New England was the home team and the favorite, which our scoring treats as the naive, no-skill-required pick. Every one of the outright winner calls that took the Patriots — Chris Broussard, Rich Eisen, Connor "Boston Connor" Campbell, Jonas Knox, Richard Sherman (twice, on Sept. 5 and again Sept. 17), John Breech and Mark Wahlberg — graded out at zero added skill. They called the chalk and the chalk covered. Correct, but not the kind of call that tells you anything about who's sharp.

The calls that actually separated the field were the score predictions, because the "naive" baseline there wasn't just picking New England — it was picking New England by half a point. Getting the margin right required seeing something the market didn't.

The calls worth naming

Nobody did. Jared Dubin came closest, calling "Patriots (-5.5) 21-13 over Steelers" in one column and repeating the exact 21-13 score in another — an 8-point margin that graded out at 78.57 accuracy and +39.34 skill, the best mark anyone posted on this game. Pete Prisco (Patriots 21, Steelers 14) and Breech (27-20) both called 7-point Patriots wins, each worth 76.19 accuracy and +34.72 skill. All three undersold the actual margin by nine to ten points, and all three still finished as the sharpest calls on the board — a low bar, given what the rest of the field guessed.

Tony Fusco, writing satire, picked "Patriots 3, Steelers 0" and Jordan Dajani twice called a nervy 16-13 New England win. Both landed at 66.67 accuracy — closer to right than wrong on the winner, further off on the number.

The misses

Shook: "I would take the Steelers straight up... just because of how well their defense performed in week one." Middlekauff: "I'm taking Pittsburgh in this spot. Pittsburgh's defense looked pretty solid." Both graded at 0 accuracy and -194 skill — a heavier penalty than the naive-pick calls received for being right, because going against the favorite and losing costs more than going with it and winning gains.

The worst number on the board belongs to Cowherd, who called it "a road upset" and predicted Pittsburgh 24-21. That graded 16.67 accuracy and -80.76 skill — wrong on the winner and off by 41 points on the margin.

Whose record moved

Dubin's strong call here bucked his own pattern. Across 27 graded score predictions overall, he's at 53.76% accuracy and essentially break-even skill (-0.53). This one graded well above both marks — a good week, not (yet) a trend, on a thin score-prediction sample.

Middlekauff's miss is a real outlier rather than a pattern. In 15 graded game-outcome calls he's hitting 75.93% accuracy; picking Pittsburgh here is one blemish against an otherwise solid outright-winner record, on a season where he's already picked up a Pundit of the Week nod for accuracy.

Cowherd's miss fits more comfortably into his own numbers. Across 74 graded score predictions he sits at 62.36% accuracy — respectable — but this particular call, both the team and the margin, was a clean whiff against that baseline, not a shock relative to a shakier overall record (55.31% accuracy across 454 graded calls of all types).

What it touches beyond this game

New England win-total calls we've graded have not been kind to forecasters generally — the field's Patriots season-win projections we've tracked sit at just 10% accuracy across 16 graded calls, even as Patriots playoff-berth picks have hit 93.3% of the time over 30 graded calls. Steelers playoff-berth calls we've tracked are a perfect 29-for-29 on accuracy. A 17-point Week 2 win over Pittsburgh doesn't resolve any of those disputes on its own, but it's one more data point for anyone still deciding whether to believe the New England hype or the New England history. Full team pages for the Patriots and Steelers, and the season-long leaderboard, have the rest.