Recruiting, Success and Rivals.com

Last summer we created a recruiting model that tracked 4 year accumulated recruiting talent using Scout.com’s data. The overall model was perhaps less than satisfactory due to Scout.com changing their ranking method in the 4th year (2007) we used. I created two models to address this – one that merely had the total accumulated points of the past 4 years, and one that weighted the 4th year to look more like the other 3. Neither would/could be perfect, as however Scout changed their ranking was unavailable to us.

Nevertheless, we had the following for the top 25 in “Model 1” and “Model 2” –

Model 1

1 Florida
2 USC
3 Texas
4 LSU
5 Tennessee
6 Georgia
7 Michigan
8 Miami (Fl)
9 Oklahoma
10 Florida State
11 Ohio State
12 Auburn
13 Notre Dame
14 Penn State
15 California
16 Alabama
17 South Carolina
18 Texas A&M
19 Pittsburgh
20 North Carolina
21 Nebraska
22 Oregon
23 Clemson
24 Virginia Tech
25 UCLA

Model 2

1 USC
2 Florida
3 Georgia
4 Texas
5 Michigan
6 LSU
7 Tennessee
8 Oklahoma
9 Florida State
10 Miami (Fl)
11 Ohio State
12 Auburn
13 Penn State
14 Notre Dame
15 California
16 Alabama
17 Texas A&M
18 South Carolina
19 Pittsburgh
20 Nebraska
21 North Carolina
22 UCLA
23 Virginia Tech
24 Clemson
25 Oregon

Of the two, I prefer Model 2 as the changes to Scout’s rankings greatly increased the number of points in the 2007 class, thus perhaps causing that class to have an outsized impact on Model 1. While I cannot be positive I correctly re-weighted 2007 in Model 2, it seems to correctly reflect USC’s 4 year recruiting strength against Florida’s stellar 2007 class.

Some commentators at the time suggested I simply use Rivals.com’s data. I have never been a huge fan of Rivals, especially as I suspected that they consistently overrated FSU recruiting classes. I even recall reading a piece that showed how recruit’s rankings rose or fell after they either committed to FSU or to another school.

Regardless, I decided to do a four-year trailing model using Rivals.com data. Included was every school that had a recruiting ranking for the past 4 years (FAU, for instance, has only 2 years). Since their scoring system has remained constant, no adjustments to the weightings were necessary.

Let’s just say the results did little to allay my suspicions about a FSU bias.

Rivals.com Four Year Accumulated Recruiting Rankings

1 Southern Cal
2 Florida
3 Florida State
4 LSU
5 Oklahoma
6 Georgia
7 Tennessee
8 Texas
9 Michigan
10 Miami-FL
11 Ohio State
12 Auburn
13 Alabama
14 Nebraska
15 Penn State
16 California
17 Notre Dame
18 South Carolina
19 Texas A&M
20 Clemson
21 Oregon
22 Maryland
23 Ole Miss
24 Virginia Tech
25 Arkansas

Obviously the Seminoles stand out as jumping from either 10th or 9th in the Scout models to 3rd in the Rivals model. While differences of opinion are to be expected, the variation for FSU is rather startling, with no other team having nearly the same magnitude of variance.

When we change the model to average recruiting class of the schools over the past 4 years, FSU again comes out very favorably –

Rivals.com Four Year Average Recruiting Class Rankings

1 Southern Cal
2 Florida State
3 Florida
4 Oklahoma
5 LSU
6 Georgia
7 Michigan
8 Texas
9 Miami-FL
10 Tennessee
11 Ohio State
12 Alabama
13 Auburn
14 Penn State
15 Nebraska
16 California
17 Texas A&M
18 Notre Dame
19 South Carolina
20 Maryland
21 Clemson
22 Oregon
23 Ole Miss
24 Arkansas
25 Virginia Tech

Are we to believe that, over the past 4 years, the Seminoles have averaged the 2nd best recruiting classes in the nation?

If we are to so believe, then the Florida State story is one of enormous failure. Let’s take a look at those top 10 average classes again, and the two year records that would be the results of the majority of that talent –

1 Southern Cal: 11-2 Pac 10 Champs, 10-2 Pac 10 Champs
2 Florida State: 7-6, 7-5
3 Florida: 13-1 SEC, MNC Champs, 9-3
4 Oklahoma: 11-3 Big 12 Champs, 11-2 Big 12 Champs
5 LSU: 11-2, 11-2 SEC and possible MNC
6 Georgia: 9-4, 10-2
7 Michigan: 11-2, 9-3
8 Texas: 10-3, 9-3
9 Miami-FL: 7-6, 5-7
10 Tennessee 9-4, 9-4

If Rival’s rankings are correct we have what amounts to coaching malpractice being performed in Tallahassee (and perhaps Coral Gables).

I think it is more likely Rivals does indeed overrate FSU players. That they do this troubles me not in the least, as I think it hurts the overall image of the Florida State program significantly.

However, it doesn’t much help me with the origin of this project – locating an unbiased, reliable data source for my recruiting models. I’m not giving up quite yet, and remain open to ideas.