re: 247's 2026 College Football Team Talent Composite Rankings Are Out

Thought it'd be interesting to see the 2025 talent rankings vs final AP rankings. A bit difficult as 9 of the final top 25 weren't in the top 50 talent composite rankings.

Final ranking over(+) or under(-) achievement compared to talent:
Indiana +50
BYU +39
Vanderbilt +35
Virginia +34
Tulane +32
James Madison +31
Utah +30
Houston +28
Navy +27
North Texas +26
Iowa +25
Texas Tech +22
Ole Miss +18
Miami +13
TCU +7
Oregon +1
Oklahoma +1
Texas A&M 0
Notre Dame -1
Ohio State -2
USC -3
Georgia -5
Alabama -7
Texas -8
Michigan -10
Hall of Shame (Not in final Top 25 but in top 25 talent) - LSU, Clemson, Penn St, Florida, Auburn, Tennessee, South Carolina, Florida St, Nebraska, Missouri, Arkansas, Miss St, SMU

2025 Actual Final Rankings vs Talent Composite Rankings, ordered by Final AP ranking:
Rank Team
1 Indiana Not in top 50 (+50)
2 Miami #15 (+13)
3 Ole Miss #21 (+18)
4 Oregon #5 (+1)
5 Ohio State #3 (-2)
6 Georgia #1 (-5)
7 Texas Tech #29 (+22)
8 Texas A&M #8 (0)
9 Alabama #2 (-7)
10 Notre Dame #9 (-1)
11 BYU Not in top 50 (+39)
12 Texas #4 (-8)
13 Oklahoma #14 (+1)
14 Utah #44 (+30)
15 Vanderbilt Not in top 50 (+35)
16 Virginia Not in top 50 (+34)
17 Iowa #42 (+25)
18 Tulane Not in top 50 (+32)
19 James Madison Not in top 50 (+31)
20 USC #17 (-3)
21 Michigan #11 (-10)
22 Houston Not in top 50 (+28)
23 Navy Not in top 50 (+27)
24 North Texas Not in top 50 (+26)
25 TCU #32 (+7)

Can't you just go anytime to their team recruiting page and click on the average star score to get this list?

List can't be correct. No Ole Miss in the top 10

quote:

---

Would like to see the avg age of starters for each team.

---

Here's what I have so far. I mostly used the depth charts from Ourlads.com, collected starting in early summer, have updated some of them since camp started, but probably still not accurate and I don't plan to complete that process until actual opening game starters are available. I don't have data for Miss State and Arkansas yet. The numbers are based on 247 composite HS ratings only. This is stats for 22 starters on offense and defense, no special teams (K/P/LS).


Team-------Avg--------Median
ALA--------0.9339------0.9303
AUB--------0.8831------0.8799
FLO--------0.9116------0.8976
GEO-------0.9448------0.9559
KEN--------0.9031------0.8997
LSU--------0.9233------0.9109
MZZ--------0.8949------0.9086
OKL--------0.9028------0.9031
OLE--------0.8982------0.8976
SCA--------0.8829------0.8860
TAM-------0.9140------0.9074
TEN--------0.8990------0.8933
TEX--------0.9275------0.9375
VAN--------0.8681------0.8691


Note: If a player (usually a transfer or former walk-on) was unrated out of HS, I assign them a default HS rating of 0.8000. The impact on aggregate team numbers is not great because very few transfers go from unrated to high 4 and 5 star transfer rating. Comparing HS Recruit ratings to Transfer ratings is apples to oranges, IMO, and you'd have to rerank the entire teams to get a valid comparison.

When I look at teams' entire roster, the HS and transfer ratings of their transfers tend to average out to about the same. At least for SEC level teams, whether you use transfer ratings or not it doesn't have much impact when looking at team averages. When looking at 22 starters, it could have a substantial impact if they had multiple unrated starters, but so far the effect is pretty small.

Some of the lower rated teams have a large number of transfers with a significant amount of college experience at various levels. I do try to incorporate that into what I call an adjusted rating. I just started using that last season and I'm still tweaking the algorithm. As of right now, the RtgAdj gives an interesting look at HS rating adjusted by years in college, but doesn't incorporate data on snaps or starts, etc.