# Lance Zierlein's 123.5-Point Week Tops a Roster Spot-Heavy Slate | MissedTakes Weekly

*Published 2026-09-29 — MissedTakes*

## Pundit of the Week: MissedTakes Score

[Lance Zierlein](/pundits/lance-zierlein) takes the MissedTakes Score award for the week of September 21–27, averaging **123.5** across five graded calls spread over four articles. That's a huge number by the site's own math — a MissedTakes Score of 0 means a pundit did no better than a naive baseline (picking the home team, assuming a player repeats last year's role), so 123.5 means Zierlein's average call that week beat the baseline by more than a full baseline's worth of accuracy.

Two calls did the heavy lifting. Reviewing the Bengals-Texans matchup, Zierlein initially backed off a Houston pick, worrying aloud that without Nico Collins "there's a tremendous amount of pressure on CJ, frankly, to be perfect" — but noted he'd already turned in the Texans as his official pick. That created two graded predictions from the same moment: the spoken hedge toward Cincinnati graded as a full hit (100 accuracy, works out to a +194 skill swing against the naive Houston pick), while the official on-record Texans pick graded a miss (0 accuracy) once Cincinnati won 20-6. Split feelings, split results.

The rest of the week's haul came from roster calls tied to the Texans' 53-man roster: Zierlein's calls that Lewis Bond would make Houston's roster scored full marks twice (100 accuracy, +200 skill each), and a call on Jaylin Noel also graded 100 accuracy, though it carried no skill credit since keeping Noel was already the obvious, expected outcome.

Zierlein's season-long record puts this week in context: 43 graded calls overall, a 64.77 accuracy rate and a 47.39 MissedTakes Score that already ranks him No. 1 on the [full leaderboard](/leaderboard) — but the site flags that record as a limited sample, so one big week moves the number more than it would for a pundit with hundreds of calls.

## Pundit of the Week: Accuracy

[John Middlekauff](/pundits/john-middlekauff) wins the accuracy award at **88.89%** across six graded calls in three articles. The graded picks we can point to are game-outcome calls from his Week 2 and Week 3 previews: he called Philadelphia to beat Tennessee "by like 20 points," and the Eagles won 24-20 for a full-accuracy hit worth +194 skill against a naive pick of Tennessee. He also called the 49ers to beat Miami — San Francisco won 13-35, another full-accuracy hit, though with no skill credit since San Francisco was already the expected winner. Not every pick landed: he took Pittsburgh over New England ("Pittsburgh's defense looked pretty solid"), and the Patriots won 20-3, a miss worth -194 skill.

Middlekauff's season record is a fuller picture than Zierlein's: 106 graded calls overall, an established sample, with a 62.88 accuracy and an 18.5 MissedTakes Score. His mock-draft calls (76.12 accuracy on 19 graded) and game-outcome calls (75.93 accuracy on 15 graded) are his strongest lines; division-winner and Super Bowl picks (20 accuracy each) drag the overall number down, a reminder that a strong single week doesn't rewrite a full season of long-shot calls on bigger, harder questions.

![Graded Takes by Type, Trailing 7 Days](/posts/visuals/103f4a8a-3cf4-4e49-ab54-f6861b6cb779.png)

## The Week in Review

Across the site, 1,776 takes were graded in the trailing seven days — an unusually roster-heavy stretch. Of those, 1,489 were Roster Spot Calls, dwarfing the 154 Game Outcome Picks and 113 Score Predictions graded in the same window. That's the tail end of preseason cutdown coverage settling: pundits picked apart 53-man rosters position by position, and the results came in almost entirely as full hits or complete misses.

The best calls of the week were all roster projections graded at 100 accuracy and the maximum +200 skill: [Blaine Grisak](/pundits/blaine-grisak) nailed a full Rams 53-man projection down to the practice-squad cut; [Nathan Jahnke](/pundits/nathan-jahnke) went a perfect four-for-four on Pittsburgh, Chargers, Saints and other depth-chart calls off Week 3 preseason tape; [Noah Compton](/pundits/noah-compton) called several Colts roster outcomes correctly, including two rookie defensive ends making the 53; and [Patricia Traina](/pundits/patricia-traina) projected the Giants' full initial roster correctly, including a Dominic Zvada kicker call.

The worst calls landed at the other extreme — 0 accuracy and -200 skill, the floor. [Randy Gurzi](/pundits/randy-gurzi)'s Cowboys practice-squad projection whiffed, including a surprise addition of released Giants tackle Evan Neal that didn't happen. [Brian Peacock](/pundits/brian-peacock)'s detailed 49ers 53-man breakdown missed across the board. [Grisak](/pundits/blaine-grisak) also shows up twice in the misses column — a reminder that volume on one team's roster cuts both ways, since predicting five or six outcomes off the same preseason game means more chances to be wrong alongside the chances to be right. [Tom Jensen](/pundits/tom-jensen)'s 49ers cutdown projection rounds out the miss list.

The pattern across roster calls this week is stark: they resolve as near-total hits or near-total misses, with almost nothing in between — a binary sport where "he makes the team or he doesn't" leaves no room for partial credit.

There were 286 new takes logged in the same seven-day window, and nothing yet in this week's "changed their mind" tracking — no pundit publicly reversed a call that we caught.