Bengals 20, Texans 6: The Field Picked Houston Three Times Out of Four…and Lost

Cincinnati went into Houston and won 20-6. We tracked 21 graded pregame calls on this one — straight winner picks and full score predictions — and the field leaned hard the wrong way.

The verdict on the field

Across all 21 calls, 15 picked the Texans to win and only 6 picked the Bengals. Strip it down to the 13 straight winner picks and it's the same story: 9 for Houston, 4 for Cincinnati. Nearly seven in ten pregame calls we graded had this one backwards.

That matters beyond the box score because Houston was the home team, and picking the home team is the naive, zero-credit default our scoring measures against. A Bengals pick here wasn't just right — it was right against the crowd. Every pundit who called Cincinnati outright scored a perfect 100 accuracy and a 194 skill delta, the site's shorthand for "beat the coin flip AND beat the room."

The calls worth naming

Four pundits called it straight for Cincinnati before kickoff:

• Danny Parkins: "going five on the Bengals... I like Cincinnati outright."
• Jonas Knox: "I'll take Cincinnati. Not a believer in Houston at the quarterback position... I'll take the Bengals to win that game outright."
• Tyler Sullivan, writing for CBS Sports, flagged Houston as one of five Week 2 favorites on upset alert: "Cincinnati Bengals (+3) at Houston Texans ... If this is an early indication that the unit may regress (even slightly) from their stellar production last season, the Bengals could be live." That's a road underdog call built on a specific, checkable reason, and it landed.
• Lance Zierlein talked himself off Houston live on air, worried about the secondary and the pressure on C.J. Stroud without Nico Collins: "But without Nico Collins, I just think there's a tremendous amount of pressure on CJ, frankly, to be perfect." Also correct.

On score predictions, nobody nailed the number — Cincinnati's 14-point margin was bigger than anyone forecast — but two got the winner right. John Breech picked "Bengals 23-20 over Texans," a call that scored 73.81 accuracy and a 113.24 skill delta for at least having the right team up, even with the margin off by 11 points. A pick attributed to Colin Cowherd's show similarly had Cincinnati winning 31-27, which scored 76.19 accuracy and 117.86 skill — again correct on the team, wrong on how lopsided it would get.

Everyone else who put a number on it had Houston winning: Jared Dubin (26-24 and 24-24 versions, twice, each 21.43 accuracy), Pete Prisco (28-23, 28.57 accuracy), Jordan Dajani (27-23, twice, 26.19 accuracy), and Richard Sherman (24-17, 33.33 accuracy, the best of the wrong-team bunch simply for keeping the margin closer to reality).

The misses, straight

The winner-pick misses went to Colin Cowherd, Greg Jennings, Chris Broussard, and Nick Wright, all on Houston. Houston radio voice John Granato took the Texans three separate times across three different appearances — in May, and twice more in the days before kickoff ("I still got the Texans winning this one" and "I got the Texans. Because when times are the darkest, I'm still with our team.") All three missed. Lance Zierlein also has a Texans pick on the record here alongside his correct one: he says he'd already turned in Houston for an official picks contest before changing his mind on air, so the same game produced one hit and one miss under his name.

What this did to the record book

Zierlein enters the week ranked No. 1 on our overall leaderboard (47.39 MissedTakes Score, 64.77% accuracy across 43 graded calls), but game-outcome calls are actually his thin spot — just 4 graded, and this game accounts for half of them, at 50% accuracy. The flip-flop here is exactly why: he was right about the how, wrong about the paperwork.

Tyler Sullivan, ranked No. 2 overall (42.8, 59.17%), is a different case — his game-outcome record was already his strongest category (13 graded, 70% accuracy, 58.5 skill score) before this one, and a road-dog hit with a stated reason extends that rather than rescuing it.

John Granato's record, by contrast, is thin and rough: seven graded calls total, 9.05% overall accuracy, and 0% across six graded game-outcome picks. Three of those misses are the Texans calls above. That's not a fluke reading against a bigger sample — it's most of his sample.

John Breech's score-prediction record sits at 41.62% accuracy over 39 graded calls, below his own game-outcome number (43%, 25 graded). His 23-20 call here, at 73.81% accuracy, is well above both — a good week set against a career mark that runs cool.

Beyond this one game

We didn't find graded 2026 season win-total, division, or playoff calls tied specifically to either club yet, so it's too early to say how this result reshapes those bets. What we can say: anyone still holding a Texans-favorite line after this loss to a Bengals team not many picked has a harder case to make than they did a week ago. For the fuller picture on both franchises, see the Texans and Bengals pages, and check where these pundits stand overall on the leaderboard.