# Rich Eisen's Chiefs Arc and More | MissedTakes Daily

*Published 2026-08-29 — MissedTakes*

2,123 takes were graded in the last seven days, and the batch was weighted toward the categories punditry handles worst. Division Winners accounted for 351 of them, Conference Champions another 187 — and for both types, the number graded this week equals the number graded in the site's entire history. Same for Trade and Signing Calls (349), Sleeper Calls (308) and Fantasy Bust Calls (84). For those five categories, the site-wide type averages *are* this week's results.

Those averages are not flattering:

![Five categories whose entire graded history landed this week](/posts/visuals/eda24722-c972-460d-81a5-72493432dc8d.png)

[Division Winners](/predictions/division-winner) came in at 27.4 average accuracy across 351 calls; [Conference Champions](/predictions/conference-champion) at 25.1 across 187. Picking one team out of four, or naming a Super Bowl participant in August, is mostly a losing exercise. What saves the pundits is that the naive baseline is losing too: division picks still carry a mean skill delta of +19.64, and conference champion picks +6.66. Being wrong more often than not is not the same as being worse than a coin.

The reverse case is [Trade and Signing Calls](/predictions/player-movement): 51.3 average accuracy — the second-best figure of the five — but a skill delta of **-15.07**. Half of those calls landed, and the field would still have done better saying nothing.

## Lead time explains most of it

The site's lead-time split shows exactly where the division and conference numbers come from. Conference champion picks made inside six weeks of resolution average 60 accuracy (n=55). The same call made four to eight months out averages 7 (n=96). Division winners: 49 accuracy at six weeks to four months (n=63), 24 at four to eight months (n=253), 14 beyond eight months (n=28). The claims that graded in bulk this week were overwhelmingly the long-lead kind.

## Both lists, same article

The week's best-call and worst-call boards share names, and in two cases share the exact take. [Andrew Erickson](/pundits/andrew-erickson) posted a 100-accuracy Fantasy Bust Call and a 0-accuracy Sleeper Call [out of one article](/opinions/3809e587-6fa2-464b-ab98-fd96d353c1bd). [Mitchell Eisenstein](/pundits/mitchell-eisenstein) did the same thing with [a single Rams/Lions/Super Bowl segment](/opinions/7fedb68a-c73b-4823-9b46-158905a559ef): 100 on one graded claim, 0 on another. That is grading working as designed — a podcast riff with eight claims in it becomes eight rows, not one verdict — but it is worth remembering before anyone quotes a single 100 as a pundit's week. Elsewhere on the wrong end: [Cynthia Frelund](/pundits/cynthia-frelund)'s [Week 1 slate](/opinions/4bf7f315-5d0b-4e17-a65c-b783dd6b4a79) and [Ryan Wormeli](/pundits/ryan-wormeli)'s [rookie-receiver bets](/opinions/192d6e67-2e0e-49b3-bc82-4c8a77bd5950) both graded at 0.

## Nobody formally changed their mind — but Kansas City

The changed-their-mind feed is empty this week. The closest thing in the data is [Rich Eisen](/pundits/rich-eisen)'s 2025 Chiefs file, graded on August 13 and worth revisiting because Kansas City is also the field's single worst collective win-total miss (-6.3 across 13 calls, against an actual 6 wins). Eisen made five of those 13:

| Date | Number | Accuracy |
|---|---|---|
| May 30 | 12 wins | 0 |
| Jun 23 | double digits (10) | 20 |
| Jun 24 | 15-2 | 0 |
| Jun 26 | 15-2 | 0 |
| Nov 17 | 10 wins | 20 |

He moved from 12 to 10 and then back up to 15 inside four weeks, and the late-November revision down to 10 still finished four games high. Eisen's Season Win Totals record overall: 35.33 accuracy, -2.52 score on 18 graded calls — third of the four pundits who qualify in that category, behind [Colin Cowherd](/pundits/colin-cowherd) (21.92) and [Chris Brockman](/pundits/chris-brockman) (6.59).

![The field's biggest 2025 win-total misses](/posts/visuals/a070354a-13ae-4dea-954a-fdaaafb0dde8.png)

New England, Seattle and Denver were each underrated by exactly 5 wins by the field; Washington (-4.8) and Las Vegas (-4.5) were overrated. All six of the biggest misses come from 6 to 13 calls apiece, which is thin per team but consistent in direction: the field was slow on the risers and slow off the fallers.

The award ballots tell the same story from the other end. Matthew Stafford's MVP was named on 26 of 100 2025 picks — the plurality, and the field's best award result on the board. Tetairoa McMillan's Offensive Rookie of the Year was named 5 times in 29 picks, tied with Ashton Jeanty.

## Leaderboard

[David Harrison](/pundits/david-harrison) still leads at 48.77 on 131 graded takes, with [Marcus Mosher](/pundits/marcus-mosher) carrying the top ten's best raw accuracy (80.31). Nine of the ten leaders are still flagged "limited" confidence; only [Daniel Jeremiah](/pundits/daniel-jeremiah), at 277 graded takes, is "established". Full board on the [leaderboard](/leaderboard), and the scoring rules are on the [methodology page](/methodology).