Giants 27, Texans 21: Only Two of 11 Pundits Picked the Upset
Drafted by a language model from MissedTakes' graded data, then reviewed and approved by Ben Garves before publishing. How our posts are made
The New York Giants beat the Houston Texans 27-21 on the road in Week 7 of the 2026 season, and the pundits we track leaned hard the other way. Nine of the 11 voices who picked a side took the Texans; only two called the Giants, and both of those were right. Every pundit willing to name a final score also had Houston winning, and the closest of those calls still missed the margin by 10 points.
Nine pundits picked Houston. The Giants won in Houston anyway.
The split on this game ran 9-2 for the Texans among the 11 pundits who made a pick, and the two who broke from the field were the only ones who got it right. Across the 12 winner and score calls that settled, average accuracy came in at 34 out of 100, a rough week for anyone who had Houston cruising. New York's win also means nobody picking the Texans beat the naive call of favoring the home team, since that's exactly the side they all chose.
Who picked the Giants before kickoff?
The Giants' two backers both made the call ahead of time, and one of them did it twice in different directions. First Things First's "Coach" used an Upset Alert segment on October 23 to go against Houston, saying: "I'm going with the New York Giants over the Houston Texans, and Houston's feeling great..." That's the only graded call First Things First has on the books so far, and it landed at 100% accuracy.
Tyler Sullivan is the more interesting case. Two days before the game, on October 21, he projected a Texans win in a score-by-score column: "Giants at Texans Projected score: Texans 27, Giants 20." That call missed, grading at 47.62% accuracy. But on October 22, in a separate upset-alert column, Sullivan flagged the Giants as a team that could pull off a surprise over the Texans, and that later pick is the one that counted toward his overall record for this game, scoring 100% accuracy. One pundit, two predictions, two different grades, and the one he published last was the one that beat the home-team default.
Among the five pundits who put a score on it, nobody favored the Giants, but Pete Prisco came closest. His "Texans 27, Giants 23" call had Houston winning by 4, missing the final margin by 10 points, the tightest of the five.
| Pundit | Score call | Texans margin called | Missed the final margin by |
|---|---|---|---|
| Pete Prisco | Texans 27-23 | +4 | 10 |
| John Breech | Texans 24-17 | +7 | 13 |
| Tyler Sullivan | Texans 27-20 | +7 | 13 |
| Richard Sherman | Texans 28-17 | +11 | 17 |
| Chris Simms | Texans 33-16 | +17 | 23 |
How far off were the Texans believers?
Nine pundits picked Houston to win outright, and all nine scored zero on that call, including Jared Koch, who wrote that there was "good reason to believe that the Texans will be able to walk out of this weekend and into the bye week having logged a third straight win against this John Harbaugh-led Giants squad." John Shipley and Henry McKenna both named this game as the first of four straight wins they expected Houston to log; neither came close.
Among the five who attached a score, Chris Simms' "Texans 33 to 16" call was the widest miss of the week for this game, 23 points off the final margin. Richard Sherman's "Houston Texans winning this one 28-17" wasn't far behind, missing by 17.
Does this fit Chris Simms' and Tyler Sullivan's track records?
Chris Simms' blowout call fits a pattern that's run through his season. Across 65 graded calls, Simms sits at 52.48% accuracy overall with a missed-takes score of -7.54, meaning his picks have run slightly worse than a naive default. His score predictions specifically average 51.3% accuracy and a -15.8 skill score over 63 graded calls, so this game's 23.81% accuracy and -18.75 skill reading isn't an outlier for him, it's close to his normal.
Tyler Sullivan's split result also lines up with how he's graded all season. Across 154 graded calls he sits at 63.02% accuracy overall, but his outcome picks (72.78% accuracy on 40 graded calls) have consistently out-performed his score predictions (62.92% accuracy on 101 graded calls). That's precisely the gap that showed up this week: his winner pick hit, his score prediction didn't.
Jared Koch's record cuts the other way. He's a strong roster-move pundit, 83.33% accuracy across 12 graded roster-spot calls, but this is now his second graded game-outcome pick and his second miss there, leaving him 0-for-2 picking winners even as his other work holds up.
What does this mean for the Texans' and Giants' season calls?
Houston went into this game widely projected to win the AFC South, with pundits across our tracked outlets picking double-digit win totals and division titles for the Texans well before the season started. Shanna McCarriston's column just two days before kickoff already had Houston as "the 2-4 Texans," arguing the team was better than that record. A loss to the Giants doesn't make that division case easier. The Giants, for their part, carried a much wider range of preseason win-total calls, from as low as 6-11 up to 11 wins, and neither end of that range is settled by one Week 7 result. The season still has to answer both questions; for where the two teams stand and how pundits have called their full seasons, see the Houston Texans and New York Giants team pages, and the pundit leaderboard for who's graded best so far this year.