We rebuilt the model from weeks 1–4 only — 64 completed games, nothing after — wrote down what it had to say about each week-5 matchup, and then opened the week to see whether those statements survived. Every scoreable week of 2025 is here; pick one from the chart below.
No prices are involved. Our line archive is append-only and begins on 29 July 2026, so there are no archived 2025 closes — and inventing them from another source would be exactly the fabricated data this product exists to refuse. Nothing here says whether a bet would have won.
This is not a third backtest. Two pre-registered attempts to beat NFL game markets both found no edge. That budget is only spent by comparing a forecast to a price, and this replay never sees one.
What it does measure: whether a statement like “this defense has allowed the 2nd-most receiving yards to receivers” was still true the following week — the defense staying on the same side of the league median.
A single week is a few dozen statements — far too few to separate a real effect from one good or bad Sunday. It is shown beside the season figure, never instead of it.
The interval clears 50, so the descriptions do carry week to week — but only just. A matchup note is a real tilt, not a reason to be confident.
Every confidence percentage on THE LINE links here. It is not a win probability and not a projection — it is how often a statement of that exact class was still true the week after we made it, measured across all 624 scoreable statements of the 2025 season.
What a chip on a row is actually quoting. The classes are not equal, and until now the site showed one average for all of them.
“generous to WRs”
63/105 held · 95% CI 50.4–68.9
60%
“stingy against WRs”
57/107 held · 95% CI 43.9–62.4includes 50 — no measured edge
53.3%
“generous to TEs”
57/102 held · 95% CI 46.2–65.1includes 50 — no measured edge
55.9%
“stingy against TEs”
62/106 held · 95% CI 49–67.4includes 50 — no measured edge
58.5%
“generous to RBs”
48/100 held · 95% CI 38.5–57.7includes 50 — no measured edge
48%
“stingy against RBs”
54/104 held · 95% CI 42.4–61.3includes 50 — no measured edge
51.9%
The calibration question: does a statement about the most extreme defense in the league hold up better than one about the eighth? Directionally yes — but read the intervals before believing it.
Top 3 in the league
135/234 held · 95% CI 51.3–63.8
57.7%
Ranked 4th–8th
206/390 held · 95% CI 47.9–57.7includes 50 — no measured edge
52.8%
The spread between our best and worst class is real: “generous to WRs” at 60% against “generous to RBs” at 48%. If you only trust one kind of matchup note here, trust the first.
Sharper claims did hold more often — 57.7% against 52.8% — but the two intervals overlap heavily, so we cannot claim that gap is real. It is reported because it is what we measured, not because it flatters the model.
A class whose 95% interval includes 50 is marked in amber and is not distinguishable from a coin flip. We leave those on the site, labelled, rather than quietly dropping the boards that produce them — a reliability table that only ever confirms the product is marketing.
Each note is what the model would have said going in, with the sample size it rested on. the description still held that week · it did not · not scoreable (bye or no qualifying plays).
DAL37@
NYJ22DAL had allowed the 1st-most receiving yards to WRs — good news for NYJ.
through week 4 · 4 games · 206.8 rec yds/gm (league median 145)
Week 5: 117 allowed · league median 171
DEN21@
PHI17PHI had allowed the 3rd-fewest receiving yards to TEs — good news for DEN.
through week 4 · 4 games · 24.3 rec yds/gm (league median 47.4)
Week 5: 44 allowed · league median 51.5
DET37@
CIN24CIN had allowed the 2nd-most rushing yards to RBs — good news for DET.
through week 4 · 4 games · 123.5 rush yds/gm (league median 89.3)
Week 5: 119 allowed · league median 81.5
CIN had allowed the 7th-most receiving yards to WRs — good news for DET.
through week 4 · 4 games · 164.8 rec yds/gm (league median 145)
Week 5: 121 allowed · league median 171
HOU44@
BAL10BAL had allowed the 5th-most receiving yards to TEs — good news for HOU.
through week 4 · 4 games · 66.8 rec yds/gm (league median 47.4)
Week 5: 88 allowed · league median 51.5
BAL had allowed the 5th-most rushing yards to RBs — good news for HOU.
through week 4 · 4 games · 116 rush yds/gm (league median 89.3)
Week 5: 121 allowed · league median 81.5
HOU had allowed the 5th-fewest receiving yards to TEs — good news for BAL.
through week 4 · 4 games · 32 rec yds/gm (league median 47.4)
Week 5: 47 allowed · league median 51.5
KC28@
JAX31KC had allowed the 6th-fewest receiving yards to TEs — good news for JAX.
through week 4 · 4 games · 33 rec yds/gm (league median 47.4)
Week 5: 32 allowed · league median 51.5
JAX had allowed the 3rd-fewest rushing yards to RBs — good news for KC.
through week 4 · 4 games · 63.8 rush yds/gm (league median 89.3)
Week 5: 89 allowed · league median 81.5
LV6@
IND40LV had allowed the 4th-most receiving yards to WRs — good news for IND.
through week 4 · 4 games · 173.8 rec yds/gm (league median 145)
Week 5: 148 allowed · league median 171
IND had allowed the 5th-most receiving yards to WRs — good news for LV.
through week 4 · 4 games · 169.8 rec yds/gm (league median 145)
Week 5: 130 allowed · league median 171
MIA24@
CAR27MIA had allowed the 2nd-most receiving yards to TEs — good news for CAR.
through week 4 · 4 games · 72.8 rec yds/gm (league median 47.4)
Week 5: 22 allowed · league median 51.5
CAR had allowed the 3rd-most receiving yards to TEs — good news for MIA.
through week 4 · 4 games · 71.3 rec yds/gm (league median 47.4)
Week 5: 116 allowed · league median 51.5
MIA had allowed the 6th-most rushing yards to RBs — good news for CAR.
through week 4 · 4 games · 110.5 rush yds/gm (league median 89.3)
Week 5: 228 allowed · league median 81.5
CAR had allowed the 7th-most rushing yards to RBs — good news for MIA.
through week 4 · 4 games · 110 rush yds/gm (league median 89.3)
Week 5: 14 allowed · league median 81.5
MIA had allowed the 4th-fewest receiving yards to WRs — good news for CAR.
through week 4 · 4 games · 105.3 rec yds/gm (league median 145)
Week 5: 141 allowed · league median 171
CAR had allowed the 1st-fewest receiving yards to WRs — good news for MIA.
through week 4 · 4 games · 99 rec yds/gm (league median 145)
Week 5: 110 allowed · league median 171
MIN21@
CLE17CLE had allowed the 8th-fewest receiving yards to WRs — good news for MIN.
through week 4 · 4 games · 121.3 rec yds/gm (league median 145)
Week 5: 191 allowed · league median 171
CLE had allowed the 4th-fewest receiving yards to TEs — good news for MIN.
through week 4 · 4 games · 30.5 rec yds/gm (league median 47.4)
Week 5: 73 allowed · league median 51.5
MIN had allowed the 3rd-fewest receiving yards to WRs — good news for CLE.
through week 4 · 4 games · 104.3 rec yds/gm (league median 145)
Week 5: 66 allowed · league median 171
CLE had allowed the 2nd-fewest rushing yards to RBs — good news for MIN.
through week 4 · 4 games · 60.5 rush yds/gm (league median 89.3)
Week 5: 70 allowed · league median 81.5
NE23@
BUF20NE had allowed the 1st-most receiving yards to TEs — good news for BUF.
through week 4 · 3 games · 84 rec yds/gm (league median 47.4)
Week 5: 123 allowed · league median 51.5
BUF had allowed the 8th-most rushing yards to RBs — good news for NE.
through week 4 · 4 games · 108.8 rush yds/gm (league median 89.3)
Week 5: 59 allowed · league median 81.5
BUF had allowed the 7th-fewest receiving yards to WRs — good news for NE.
through week 4 · 4 games · 110.5 rec yds/gm (league median 145)
Week 5: 206 allowed · league median 171
NE had allowed the 4th-fewest rushing yards to RBs — good news for BUF.
through week 4 · 4 games · 65.3 rush yds/gm (league median 89.3)
Week 5: 60 allowed · league median 81.5
BUF had allowed the 1st-fewest receiving yards to TEs — good news for NE.
through week 4 · 4 games · 16.8 rec yds/gm (league median 47.4)
Week 5: 46 allowed · league median 51.5
NYG14@
NO26NYG had allowed the 2nd-most receiving yards to WRs — good news for NO.
through week 4 · 4 games · 183.3 rec yds/gm (league median 145)
Week 5: 191 allowed · league median 171
NYG had allowed the 4th-most rushing yards to RBs — good news for NO.
through week 4 · 4 games · 116.3 rush yds/gm (league median 89.3)
Week 5: 68 allowed · league median 81.5
SF26@
LA23LA had allowed the 8th-fewest rushing yards to RBs — good news for SF.
through week 4 · 4 games · 72.8 rush yds/gm (league median 89.3)
Week 5: 68 allowed · league median 81.5
SF had allowed the 6th-fewest receiving yards to WRs — good news for LA.
through week 4 · 4 games · 110.3 rec yds/gm (league median 145)
Week 5: 278 allowed · league median 171
TB38@
SEA35SEA had allowed the 7th-fewest rushing yards to RBs — good news for TB.
through week 4 · 4 games · 71 rush yds/gm (league median 89.3)
Week 5: 44 allowed · league median 81.5
TB had allowed the 5th-fewest receiving yards to WRs — good news for SEA.
through week 4 · 4 games · 107 rec yds/gm (league median 145)
Week 5: 266 allowed · league median 171
SEA had allowed the 2nd-fewest receiving yards to WRs — good news for TB.
through week 4 · 4 games · 103 rec yds/gm (league median 145)
Week 5: 272 allowed · league median 171
TB had allowed the 1st-fewest rushing yards to RBs — good news for SEA.
through week 4 · 4 games · 54 rush yds/gm (league median 89.3)
Week 5: 122 allowed · league median 81.5
TEN22@
ARI21TEN had allowed the 3rd-most rushing yards to RBs — good news for ARI.
through week 4 · 4 games · 117.5 rush yds/gm (league median 89.3)
Week 5: 143 allowed · league median 81.5
TEN had allowed the 6th-most receiving yards to WRs — good news for ARI.
through week 4 · 4 games · 167.8 rec yds/gm (league median 145)
Week 5: 134 allowed · league median 171
ARI had allowed the 7th-most receiving yards to TEs — good news for TEN.
through week 4 · 4 games · 64 rec yds/gm (league median 47.4)
Week 5: 82 allowed · league median 51.5
ARI had allowed the 8th-most receiving yards to WRs — good news for TEN.
through week 4 · 4 games · 156.5 rec yds/gm (league median 145)
Week 5: 172 allowed · league median 171
TEN had allowed the 8th-fewest receiving yards to TEs — good news for ARI.
through week 4 · 4 games · 36.3 rec yds/gm (league median 47.4)
Week 5: 51 allowed · league median 51.5
WAS27@
LAC10WAS had allowed the 4th-most receiving yards to TEs — good news for LAC.
through week 4 · 4 games · 70.8 rec yds/gm (league median 47.4)
Week 5: 44 allowed · league median 51.5
LAC had allowed the 7th-fewest receiving yards to TEs — good news for WAS.
through week 4 · 4 games · 34 rec yds/gm (league median 47.4)
Week 5: 9 allowed · league median 51.5
LAC had allowed the 6th-fewest rushing yards to RBs — good news for WAS.
through week 4 · 4 games · 69 rush yds/gm (league median 89.3)
Week 5: 124 allowed · league median 81.5
This is the whole product, on real football: context you can check, with its sample size attached, and no claim we haven’t measured. When the 2026 season starts, the same notes appear on the matchup pages — beside a line archive that records what the price did.
See this season’s matchups