Fourth Down Data
The Read

This week’s analytical highlights

The numbers worth knowing before the slate kicks off — computed from our EPA, identity, and situational stack. Team profiles are from the 2025 season.

Forward-lookingWeek 2, 2026 · 86 games on the slate

Profile Mismatches

The week's most stylistically divergent matchups — where two teams' analytical fingerprints clash hardest. Air-raid vs. ground-and-pound, havoc defense vs. methodical offense. Distance is measured across 44 features via the same engine that powers The Archive.

Rice
American Conference
@
Notre Dame
FBS Independents
39.4%Off success52.3%
1.72Pts/drive3.41
21.6%Explosive28.9%
Oregon
Big Ten Conference
@
Oklahoma State
Big 12 Conference
3.20Pts/drive1.28
41.4%Def success48.7%
49.7%Off success40.6%
Charlotte
American Conference
@
Ole Miss
Southeastern Conference
1.06Pts/drive3.11
63.2 pl/gTempo75.5 pl/g
38.6%Off success50.0%
Old Dominion
Sun Belt Conference
@
Virginia Tech
Atlantic Coast Conference
22.3%Havoc14.8%
41.1%Def success47.9%
71.2 pl/gTempo65.4 pl/g
San Diego State
Pac-12 Conference
@
UCLA
Big Ten Conference
41.8%Def success53.3%
19.9%Havoc12.6%
34%Pass rate49%

The Long Haul

The week's longest road trips. Travel is one of the few purely descriptive schedule factors — these are the teams covering the most ground to play.

New Mexico State
at Hawai'i
3,210 mi
road trip
New Mexico State travels 3,210 miles to face Hawai'i.
California
at Syracuse
2,420 mi
road trip
California travels 2,420 miles to face Syracuse.
Louisiana
at USC
1,554 mi
road trip
Louisiana travels 1,554 miles to face USC.
Memphis
at Boise State
1,510 mi
road trip
Memphis travels 1,510 miles to face Boise State.
Oregon
at Oklahoma State
1,472 mi
road trip
Oregon travels 1,472 miles to face Oklahoma State.

Rest Watch

Where one team arrives on materially less rest than its opponent — a short week against a team coming off a bye, for instance.

No significant rest disadvantages this week. Rest Watch surfaces games where one team has five or more fewer days of rest than its opponent — typically after byes, short weeks, or midweek games.

Trend Lines

Teams whose late-season EPA trend diverged most from their season baseline — who was rising and who was fading as last season closed.

San José State
↓ trending down · vs Cal Poly
-0.279
3-wk EPA trend
San José State closed 2025 in decline — a 3-week EPA trend of -0.279 off its season baseline.
Wake Forest
↑ trending up · vs Purdue
+0.271
3-wk EPA trend
Wake Forest closed 2025 on an upswing — a 3-week EPA trend of +0.271 off its season baseline.
Charlotte
↓ trending down · vs Ole Miss
-0.220
3-wk EPA trend
Charlotte closed 2025 in decline — a 3-week EPA trend of -0.220 off its season baseline.
Georgia Tech
↓ trending down · vs Tennessee
-0.218
3-wk EPA trend
Georgia Tech closed 2025 in decline — a 3-week EPA trend of -0.218 off its season baseline.
Alabama
↓ trending down · vs Kentucky
-0.216
3-wk EPA trend
Alabama closed 2025 in decline — a 3-week EPA trend of -0.216 off its season baseline.
Coastal Carolina
↓ trending down · vs Fordham
-0.211
3-wk EPA trend
Coastal Carolina closed 2025 in decline — a 3-week EPA trend of -0.211 off its season baseline.

Reality Check Watch

Where the win-loss record and our per-drive model rank disagree most — teams the record over- or under-states relative to the analytical profile.

Arkansas
record understates the profile
-71
record vs model rank
Arkansas's record ranks #126 but our model ranks them #55 (2-10, 0 one-score wins) — the data likes them more than the W-L.
Western Michigan
record overstates the profile
+64
record vs model rank
Western Michigan's record ranks #25 but our model ranks them #89 (10-4, 3 one-score wins) — the data is cooler on them than the W-L.
Jacksonville State
record overstates the profile
+64
record vs model rank
Jacksonville State's record ranks #49 but our model ranks them #113 (9-5, 5 one-score wins) — the data is cooler on them than the W-L.
South Carolina
record understates the profile
-59
record vs model rank
South Carolina's record ranks #98 but our model ranks them #39 (4-8, 0 one-score wins) — the data likes them more than the W-L.
Mississippi State
record understates the profile
-50
record vs model rank
Mississippi State's record ranks #95 but our model ranks them #45 (5-8, 2 one-score wins) — the data likes them more than the W-L.

Public vs Analytical

The biggest gaps between the AP poll and our model — teams the public and the data see differently.

Tulane
AP #18 · model #53
-35
AP vs model gap
The AP ranks Tulane #18; our model has them #53 — the poll is higher on them than the data.
Houston
AP #22 · model #54
-32
AP vs model gap
The AP ranks Houston #22; our model has them #54 — the poll is higher on them than the data.
Navy
AP #23 · model #46
-23
AP vs model gap
The AP ranks Navy #23; our model has them #46 — the poll is higher on them than the data.
Virginia
AP #16 · model #32
-16
AP vs model gap
The AP ranks Virginia #16; our model has them #32 — the poll is higher on them than the data.
TCU
AP #25 · model #36
-11
AP vs model gap
The AP ranks TCU #25; our model has them #36 — the poll is higher on them than the data.

Historical Echoes

The upcoming matchups with the strongest historical analogs — past games whose team profiles most resemble this one. Descriptive context, not a prediction.

Duke @ Illinois
Closest historical analogs
2018: West Virginia @ Texas
Texas 41-42 West Virginia
2025: Tennessee @ Illinois
Illinois 30-28 Tennessee
Tennessee @ Georgia Tech
Closest historical analogs
2021: Wake Forest @ North Carolina
North Carolina 58-55 Wake Forest
2022: USC @ UCLA
UCLA 45-48 USC
Penn State @ Temple
Closest historical analogs
2025: Penn State @ Michigan State
Michigan State 10-28 Penn State
2025: Michigan @ Michigan State
Michigan State 20-31 Michigan
UCF @ Pittsburgh
Closest historical analogs
2020: East Carolina @ Tulsa
Tulsa 34-30 East Carolina
2018: Western Kentucky @ Middle Tennessee
Middle Tennessee 29-10 Western Kentucky