Skip to content

二〇二六年

NFL Poll of Polls: Week 6

By

  • data-viz
  • nfl
  • statistics

Week 5 finally produced a clean answer at the top. San Francisco went to Seattle and won 32–29, and the panel rewarded it: the 49ers are 5–0 and take the No. 1 spot with seven of nine first-place votes.

Two other results mattered. Green Bay beat Chicago 31–7, moving the Packers up eight places and dropping the Bears five. Dallas lost to Buffalo in overtime and fell seven, the week’s biggest slide. And the Giants and Commanders tied 27–27, the season’s first.

Every one of the 32 teams changed position this week. That has not happened before in this poll.

This is the same poll of polls as previous weeks: the week’s rankings standardised and pooled, every team’s move charted, and each team’s record alongside its ranking.

Here is where the panel landed after Week 5.

2026 season · Week 6 · poll of polls

One consensus.

Each row is a team. The score is measured in rank standard deviations, so a value of +1.5 is roughly one and a half rank-spreads above the median team. The strip shows where each outlet in the panel placed that team, from first on the left to 32nd on the right. Click a row to see its individual ballots.

1
2

Select any row to see that team's ballots.
# Team Score 95% CI Mean Spread across the panel Panel Field P(No. 1)
interquartile spread of the panel one outlet's rank Scale: 1 → 32, left to right
The panel — agreement with the consensus
Scores are pooled and shrunk toward the panel mean (empirical Bayes). Rank intervals and P(No. 1) come from 4,000 bootstrap resamples over the nine sources. Club marks are the official team logos, used here for identification only. Full derivation, code, and the weekly replication checklist are documented at the end of this post.

Week 2 → Week 3

Movement

This week’s matchups, priced

Fourteen games in Week 6: Cincinnati, Detroit, Miami and Minnesota are on bye. Every game is priced below from the consensus alone, with a sandbox for building any matchup you like.

2026 season · Week 2 · head to head

What the consensus says about this week.

Every game, priced from the poll of polls alone. Win probability is the shaded side. The market column is shown for comparison and was not an input — on this slate the two agree on the favourite in 15 of 16 games.

Or build any matchup
Schedule from nflverse. Closing lines shown for reference only. How these numbers are derived, and how well they hold up, is in the methodology below.

Path to the Super Bowl

Simulated brackets

Loading…

Wild card round

These six games are fixed by the seeding, so their prices are computed directly rather than sampled. The higher seed hosts.

Most likely to reach the Super Bowl

Seed Team Reach SB Win SB

Most likely Super Bowl matchups

    9 outlets this week. Mean distance from the consensus order, in rank positions:

    OutletAnalystMean gapρ vs consensus
    Bleacher ReportB/R NFL staff1.250.982
    FOX SportsRalph Vacchiano1.380.976
    Sporting NewsVinnie Iyer1.440.976
    CBS SportsPete Prisco1.500.975
    Sports IllustratedConor Orr1.560.964
    theScoretheScore NFL desk1.560.971
    Sharp Football AnalysisRaymond Summerlin1.690.966
    NFL.comNick Shook1.880.960
    USA TodayNate Davis2.690.932

    Nine ballots is the fullest panel since this poll began, but agreement loosened slightly (mean pairwise Spearman ρ = 0.932, split-half 0.991) — adding a ninth voice that sits furthest from the consensus will do that. USA Today is the outlier at 2.69 positions; every other outlet is inside 1.88.

    Sources and method

    Every number above comes from one pipeline. Each outlet’s weekly 1–32 ballot is standardised, pooled in a measurement-error model, shrunk toward the panel mean, and published with a range reflecting how much the outlets actually disagree. Records are the standings each team carried into the week, computed from the same results feed — which is why the Giants and Commanders show 3-1-1 and 1-3-1 after their tie. The matchups are priced from the consensus score alone, using constants fitted on 6,719 completed regular-season games.

    The method, the calibration and the week-by-week replication checklist live with the code rather than in every post:

    • analysis/nfl-power-rankings/SOURCES.md — one section per outlet: URL pattern, cadence, markup and the specific traps. Follow this file to reproduce a week.
    • analysis/nfl-power-rankings/METHODOLOGY.md — both derivations, and what the method does not claim.
    • analysis/nfl-power-rankings/README.md — the pipeline and how to run it.

    This week’s panel — nine ballots, every one verified complete:

    OutletAnalystThis week’s ballot
    NFL.comNick Shooknfl.com
    CBS SportsPete Priscocbssports.com
    FOX SportsRalph Vacchianofoxsports.com
    Sports IllustratedConor Orrsi.com
    USA TodayNate Daviscentraloregondaily.com (syndicated)
    Sporting NewsVinnie Iyersportingnews.com
    Sharp Football AnalysisRaymond Summerlinsharpfootballanalysis.com
    theScoretheScore NFL deskthescore.com
    Bleacher ReportB/R NFL staffbleacherreport.com

    NFL.com came back

    Nick Shook’s column returns after two weeks of 404s and is a complete ballot, which puts the panel back at full strength. Its own ranking was the second furthest from the consensus but only slightly, so the absence looks like a scheduling gap rather than a change in the series.

    How USA Today is read now

    usatoday.com returns HTTP 402 Access Restricted to ordinary browser agents and serves only a Googlebot user agent. Last week the pipeline took that route and I said so. I am not going to keep doing it — presenting a non-Google crawler as Googlebot to get past a block is not something I want this pipeline doing quietly every Tuesday.

    The column is nationally syndicated, and a partner site carries it verbatim under the same headline. That copy returns HTTP 200 to an ordinary request, so Nate Davis’s ballot is back in the panel, fetched the way any reader could fetch it. The tradeoff is real and worth stating: syndicated copies carry no datePublished metadata, so the freshness gate cannot date this one. It is included on the strength of an independent cross-check of the original — same headline, same No. 1 (49ers) and No. 32 (Titans) — not on a verified timestamp. The pipeline logs it as freshness unverified on every run, which is the honest label.

    One trap this week, avoided

    The Reynolds columns are still the cautionary tale. NFL.com’s year-less slug …neil-reynolds-week-6-power-rankings returns HTTP 200 with a 2020 article — it was updated to 2020-10-11 — while the correct 2026 slug 404s. That is now four consecutive weeks without the column (weeks 3–6), so the fair description is that it has stopped rather than that it is running late.

    Every other ballot passed the freshness gate against its own datePublished / dateModified inside the October 13–19 window for Week 6.

    ESPN remains absent permanently: every URL returns an empty HTTP 202 to a plain request.

    One candidate found and rejected: Pro Football Talk (NBC) publishes a weekly 1–32, but its rendered page omits entries 12 through 14 entirely, so it fails the completeness check. Worth re-testing next week.