# Jeff Kerr's Perfect Week on the Eagles' Roster Bubble | MissedTakes Weekly

*Published 2026-10-12 — MissedTakes*

Three calls, three hits. Jeff Kerr swept both Pundit of the Week awards for 2026-W41, taking Accuracy (100) and MissedTakes Score (133.33) on a sample of three graded calls, all of them about the Philadelphia Eagles' roster.

![A perfect three-for-three, with one easy call mixed in](/posts/visuals/b94dfe75-6042-4e77-bb6f-f08c1a107982.png)

A three-call week is a small one, and Kerr's season as a whole shows why that number matters. Across 184 graded calls (mostly Eagles roster projections), he carries a 75.58 accuracy and a MissedTakes Score of 1.86, almost exactly break-even against a naive baseline. His roster-spot calls specifically, 181 of them, average 77.01 accuracy but a negative 6.96 score, meaning when he's right it's often the easy kind of right. This week was the exception, and the grading on each call shows why.

The highest-value call was about backup quarterback Andy Dalton. Back in August, Kerr argued "the Eagles just can't trust Dalton to play a football game against a first-team defense" and predicted he'd leave Carolina for Philadelphia. The naive expectation was that Dalton simply stayed where he was; he didn't, Kerr called the move, and the take graded the full 200 points on the skill scale. His second call, that running back Carson Steele had "played his way onto the roster" and would beat out Cameron Latu for a depth spot, also returned the full 200, since Steele spent time on the practice squad before cracking the 53. His third call, that lineman John Ojukwu would make the roster, was also right, but it graded a skill delta of zero: Ojukwu had already made a 53-man roster the year before, so predicting he'd do it again was the naive pick, not an insightful one. Two hard calls and one easy one, all correct, is a rare combination, and it's what separated this week from his season-long average.

### The week in the aggregate

Across the site, 1,156 takes were graded in the week, averaging 51.8 accuracy and a modest +14.89 skill score against the naive baseline. Score Predictions (388 calls) and Game Outcome Picks (272) carried the bulk of the volume, with Mock Draft Picks close behind at 251.

![Award picks were the week's hardest calls to get right](/posts/visuals/4157edfb-39f7-47ce-ad1b-1545290f0f85.png)

Accuracy by category split sharply. Pro Bowl and All-Pro calls topped the week at 62.0 accuracy, Score Predictions and Game Outcome Picks followed in the mid-50s to low 60s, and Trade and Signing Calls and Roster Spot Calls landed near the middle at 52.2 and 50.0. Mock Draft Picks sat at 36.5, Season Win Totals at 40.5, and Award Picks brought up the rear at 19.2, the week's toughest category on just 26 graded calls.

That Award Picks number lines up with how lopsided last season's MVP ballot actually was: of 68 picks the field made for the 2025 award, 24 went to the eventual winner, Matthew Stafford, with the rest scattered across ten other names including Drake Maye (10), Baker Mayfield (7) and Lamar Jackson (6). The Offensive Rookie of the Year ballot was even more split, with Tetairoa McMillan winning on just 5 of 29 picks, tied with Ashton Jeanty. Season-ending award calls tend to get graded in a cluster once a season closes, which likely explains why the category looks rough in any single week.

Season Win Totals told a similar story of a field that missed in both directions on 2025: the collective call on the Chiefs overshot their actual win total by 6.4 wins across nine calls, Washington's field missed high by 5.2, while Denver, New England and Seattle were each underestimated by around 5 wins across their respective calls.

### Best and worst of the week

The week's best-graded calls were about as decisive as they get. John Granato's game-by-game walk through the Texans' schedule (predicting wins in nearly every game and just one loss, at Baltimore) graded a perfect 100 accuracy and +200 skill. Ross Jackson's game-by-game Saints projection and Nick Wright's slate of season win-total, playoff and Super Bowl picks did the same.

The low end was rougher. Rich Eisen appears twice among the week's worst calls, once for a Jaguars season projection and "game wreckers" ranking, once for a preseason power ranking that had the Rams and Chargers meeting in the Super Bowl. Bill Huber's Packers projection, Adam Chudwin's 12-5 call on the Bears, and Douglas Clawson's bold-predictions piece (headlined by a Ravens-Giants Super Bowl and a winless Dolphins season) each landed at the bottom of the scale as well: 0 accuracy and a full -200 on the skill measure.