Findings, methodology notes, and the occasional caught mistake.All posts
September 2026 · History
The Big 12's Power Shift Was Bigger Than the Big 12
In the last post, we watched
the Big 12's power slide from its North to its South in about three
seasons, 2002 to 2004. Which made us wonder: was that just one
conference's drama, or was the whole country shifting? So we zoomed
out. Way out. Every FBS team since 1965, more than 45,000 games, 61
seasons, sorted by where on the map each team lives.
The short version: the Big 12 was the epicenter of
something bigger. Texas, Oklahoma, Louisiana and Arkansas went from
the weakest part of the country in the 1990s to one of the strongest
by the late 2000s, and the sharpest jump came in 2003. The Plains,
home to most of the old Big 12 North, faded at the same time. And the
Southeast, which everybody would guess is the story of this era? It
had already been the boss since the early 1980s. What changed in the
SEC era wasn't that the region got stronger; it's that its power got
squeezed into fewer programs.
South Central top-25 teams per season, 1990s2.5
… in 2006–20155.7
South Central wins vs. other regions35% → 55%
Seasons studied61
First, a map
The Census Bureau's map of U.S. regions and divisions, which we borrowed to sort every team.
Map: U.S. Census Bureau (public domain), via
Wikimedia Commons.
Conferences change every few years (see: everything since 2010), so
instead we sorted teams by geography. We used the Census Bureau's
nine divisions, squished into seven so every region has enough teams:
Northeast: New England plus New York, New Jersey and Pennsylvania
Great Lakes: Ohio, Indiana, Illinois, Michigan and Wisconsin (yes, Notre Dame counts)
Plains: Iowa, Kansas, Minnesota, Missouri, Nebraska and the Dakotas
Southeast: the Atlantic coast from Delaware to Florida, plus Kentucky, Tennessee, Alabama and Mississippi
South Central: Texas, Oklahoma, Louisiana and Arkansas
Mountain: Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah and Wyoming
Pacific: California, Oregon, Washington, Alaska and Hawai'i
That means some teams land in surprising spots. Colorado, a Big 12
North team, counts as Mountain. LSU and Texas A&M, both SEC
schools today, count as South Central.
How do you compare a 1968 team to a 2018 team?
Carefully. We built a rating for every team in every season using
nothing but final scores, adjusted for how good each opponent was and
for home field. It's not perfect, but it works the same way in every
era, and it agrees closely with SP+, a much fancier rating, in every
season where both exist (a typical correlation of 0.94, where 1 is a
perfect match). Then we asked a simple question each season: which
regions do the 25 best teams come from?
Sixty-one seasons, seven tiny charts
Share of the nation's 25 best-rated teams, by region
NortheastNew England, NY, NJ, PA
Peak 14% (1966) · now 5%
Great LakesOH, IN, IL, MI, WI
Peak 20% (1995) · now 15%
PlainsIA, KS, MN, MO, NE, ND, SD
Peak 12% (1997) · now 5%
SoutheastAtlantic coast from DE to FL, plus KY, TN, AL, MS
Peak 42% (2019) · now 32%
South CentralTX, OK, LA, AR
Peak 26% (1977) · now 21%
MountainAZ, CO, ID, MT, NV, NM, UT, WY
Peak 13% (1983) · now 9%
PacificCA, OR, WA, AK, HI
Peak 17% (1990) · now 12%
Five-season averages. In each panel, the blue line is that region and the gray lines are the other six. The shaded band is 2002–2004, when the Big 12 flipped. Hover a line for a season's value.
See the numbers by five-year stretch
% of the top 25
1965–69
1970–74
1975–79
1980–84
1985–89
1990–94
1995–99
2000–04
2005–09
2010–14
2015–19
2020–24
Northeast
13
12
10
7
6
8
7
3
8
1
5
5
Great Lakes
17
17
18
12
14
14
19
14
11
17
19
15
Plains
6
6
6
8
6
7
12
10
7
6
4
6
Southeast
25
24
21
34
32
37
33
36
27
30
39
38
South Central
19
23
26
18
17
8
12
17
23
22
18
20
Mountain
10
6
7
12
10
10
10
8
10
9
5
8
Pacific
10
11
13
10
14
15
6
13
13
14
10
8
Here's the story those lines tell, era by era.
1965–79 The Big Ten and Texas run the show
Michigan Stadium on Nov. 22, 1969, the day Michigan knocked off an 8–0 Ohio State, 24–12. The Great Lakes owned this era.
Photo: University of Michigan News and Information Services, University of Michigan Libraries,
CC BY 4.0, via
Wikimedia Commons.
Back then, the powers were the Great Lakes (the Big
Ten plus Notre Dame) and South Central (Texas,
Oklahoma, Arkansas and the old Southwest Conference). Between 1965 and
1979 they piled up 37 and 32 AP top-10 finishes. Two weeks after that
Michigan game, Texas beat Arkansas 15–14 and went on to win the
national title. South Central peaked at 26% of the top 25 in
1975–79.
1980s–90s The Southeast takes the throne
Around 1981, the Southeast put about three more
teams a season into the top 25, and it has had the biggest share of
the top 25 in every five-year stretch since. Forty-plus years at the
top. That's not a streak; that's a dynasty.
South Central, meanwhile, fell off a cliff. By 1990–94 it had
just 8% of the top 25, and in the 1990s it won only 35% of its games
against teams from other regions. That's the worst record of any
region in any decade since 1965. (The old Southwest Conference, home
of most of those Texas schools, broke up after the 1995 season.)
2000s–10s Texas and friends come roaring back
This is the national version of the Big 12 story, and it gets its
own section below.
2020s The Big Ten strikes back
Plot twist for the current era: the Great Lakes is
back. It had 16 AP top-10 finishes in 2020–24, the most for the
region in any five-year stretch since 1965, and national titles for
Michigan (2023), Ohio State (2024) and Indiana (2025). Indiana. Let
that sink in.
So what changed around 2003?
Let's compare the decade before the Big 12 flip (1990–99) with
the decade after it (2006–15), skipping the messy years in
between:
Region
Top-25 teams per season
AP top-10 finishes per season
Win % vs. other regions
Northeast
1.9 →0.9
0.5 →0.3
43% →42%
Great Lakes
4.2 →3.8
1.6 →1.7
47% →50%
Plains
2.4 →1.5
1.2 →0.5
53% →49%
Southeast
8.7 →7.3
4.6 →3.4
66% →53%
South Central
2.5 →5.7
0.5 →2.2
35% →55%
Mountain
2.6 →2.5
0.9 →0.5
56% →41%
Pacific
2.7 →3.3
0.7 →1.5
47% →55%
South Central wins "most improved" in a landslide. It went from 2.5
to 5.7 top-25 teams a season and from 0.5 to 2.2 AP top-10 finishes,
and its record against other regions went from 35% to 55%. Its
average team went from 2.7 points per game worse than an
average FBS team to 0.8 points better. And the year with the
sharpest jump? 2003, right when the Big 12 South was taking over.
The other half of the Big 12 story shows up nationally too. The
Plains dropped from 2.4 to 1.5 top-25 teams a season and from 1.2 to
0.5 AP top-10 finishes. The Mountain region, which includes Colorado,
went from winning 56% of its games against other regions to 41%.
Ouch.
It wasn't just the Big 12 South
LSU's Tiger Stadium. No program in the region climbed further: from 2 top-25 seasons in the 1990s to 9 in 2006–15.
Photo: getmahesh, CC BY-SA 2.0, via
Wikimedia Commons.
Top-25 seasons, 1990–99 vs. 2006–15
South Central
LSU2 → 9
TCU0 → 7
Oklahoma4 → 10
Oklahoma State0 → 6
Arkansas2 → 6
Texas5 → 4
Texas A&M8 → 4
Plains
Missouri1 → 6
Iowa5 → 2
Kansas State6 → 2
Nebraska10 → 4
Southeast
Clemson3 → 7
Ole Miss0 → 4
South Carolina0 → 4
Alabama6 → 8
Florida State10 → 6
Florida9 → 5
Miami9 → 1
Tennessee10 → 2
Seasons in our rating's top 25 in each ten-year stretch. Blue bars are gains, red bars are losses.
South Central's comeback was a group project. LSU
went from 2 top-25 seasons to 9, winning the BCS title after the 2003
season and the AP title in 2007 along the way. TCU, which didn't even join the
Big 12 until 2012, went from 0 to 7. Oklahoma went
from 4 to 10, Oklahoma State from 0 to 6, and
Arkansas from 2 to 6. Texas was good in both decades
(5 and 4). The one program that went the wrong way was Texas A&M,
the region's best of the 1990s by this count, which slid from 8 to 4.
Up in the Plains, Nebraska went from 10 top-25
seasons to 4, Kansas State from 6 to 2 and Iowa from 5 to 2. Missouri
was the rebel, climbing from 1 to 6.
The Southeast was already the boss
Fans leaving Alabama's Bryant–Denny Stadium after the 2010 spring game. By the 2010s, the Southeast's power ran through a few programs like this one.
Photo: Carol M. Highsmith, Library of Congress (public domain), via
Wikimedia Commons.
Ask any fan to name the big regional story of the 2000s and 2010s and
they'll say "the SEC," probably loudly. But by our numbers, the
Southeast didn't rise in that period. It was already on top, and it
actually slipped. From the 1990s to 2006–15, its top-25
teams per season fell from 8.7 to 7.3 and its record against other
regions from 66% to 53%, even as its national titles went up from 6
to 8.
How can a region win more titles while getting weaker? Depth. In the
1990s, four Southeast programs were in the top 25 almost every year:
Florida State and Tennessee (10 seasons each) and Florida and Miami (9
each). That's 38 top-25 seasons between them. By 2006–15, those
same four had 14. Alabama (6 to 8) and Clemson (3 to 7) rose, along
with Ole Miss and South Carolina (0 to 4 each), but the region's power
got packed into a smaller group at the very top. The Southeast also
added a lot of new FBS programs, growing from about a quarter of all
FBS teams in the 1990s to about a third today. The 13 newcomers since
1995 averaged 4.5 points a game worse than an average FBS team
in 2006–15. Leave them out, and the rest of the Southeast was
exactly as strong as it had been in the 1990s. The old guard didn't
get weaker on average; the region just got bigger and more lopsided.
Data: every FBS score, final AP polls and SP+ ratings (1970 on) from
CollegeFootballData.com. The AP ranked only 10 teams through 1967, so
we counted top-10 finishes throughout. Our scores-only rating gives
nearly the same regional results as SP+ (r = 0.97). Records against
other regions include home-and-home and "buy" games, which bigger
programs tend to win, but bowl games alone, played mostly on neutral
fields, point the same way: South Central won 39% of its bowls
against other regions in the 1990s, 48% in the 2000s and 59% in the
2010s. The year-by-year timing is a description, not a formal test;
61 seasons isn't enough data to prove exactly when a shift began.
_at_Night.jpg){kind=link}
