Interesting read for recruitniks

#2
#2
Fascinating. Thanks Justin.

So what do recruits want? According to the model, they usually will pick the BCS-conference school nearest their hometown that has the biggest on-campus stadium and won the most games last season. Not the past five seasons, mind you. "You can make the argument that recruits may be a little short-sighted," DuMond said.
 
#5
#5
"Jeannette, Pa., quarterback Terrelle Pryor will sign with Ohio State.
According to the model, there is a 40.2 percent chance Pryor will choose Ohio State, compared to a 37.9 percent chance he will choose Michigan."

Pryor is going to completely debunk this prediction model when he signs with the Vols.

You heard it here first...the ink is already dry...my aunt is good friends with his mother...if I'm wrong, I'll never post here again...etc, etc.
 
#7
#7
Jaytrain5 said:
"Jeannette, Pa., quarterback Terrelle Pryor will sign with Ohio State.
According to the model, there is a 40.2 percent chance Pryor will choose Ohio State, compared to a 37.9 percent chance he will choose Michigan."

Pryor is going to completely debunk this prediction model when he signs with the Vols.

You heard it here first...the ink is already dry...my aunt is good friends with his mother...if I'm wrong, I'll never post here again...etc, etc.
Click to expand...

So your the ExVolLinebacker of Volnation?? :p
 
#13
#13
Doesn't this model suggest something about the quality of recruiting a school does?

Doesn't CPF and UT pretty much defy their model everytime they sign a top 10 class?

Thinking of last year... Didn't almost all the top recruits pass on BCS schools with good win totals, large stadiums, and closer to home?
 
#14
#14
sjt18 said:
Doesn't this model suggest something about the quality of recruiting a school does?

Doesn't CPF and UT pretty much defy their model everytime they sign a top 10 class?

Thinking of last year... Didn't almost all the top recruits pass on BCS schools with good win totals, large stadiums, and closer to home?
Click to expand...

Apparently not entirely, they hit the Rivals100 with 72.5% accuracy for the 2007 recruiting season.
 
#15
#15
that's the claim but here are guys UT signed and schools that recruited them and seem more likely choices according to the factors laid out in that article:

Berry- UGA
Martin- OSU
O'Neal- Cal
G Jones- OU among others
Langley- PSU
Creer- Tex, OU
McKenzie- Tex, OU, others
Sawtelle- Mich, OSU
Paige- FSU, UF, Tex
Walker- LSU

Without these guys, what would have been left last year?
 
#17
#17
sjt18 said:
that's the claim but here are guys UT signed and schools that recruited them and seem more likely choices according to the factors laid out in that article:

Berry- UGA
Martin- OSU
O'Neal- Cal
G Jones- OU among others
Langley- PSU
Creer- Tex, OU
McKenzie- Tex, OU, others
Sawtelle- Mich, OSU
Paige- FSU, UF, Tex
Walker- LSU

Without these guys, what would have been left last year?
Click to expand...

How many of those were Rivals100s?
 
#18
#18
The article didn't limit it to Rivals 100.... However, Jones, Berry, Martin, Donald, and Creer were.
 
#21
#21
sjt18 said:
The article didn't limit it to Rivals 100.... However, Jones, Berry, Martin, Donald, and Creer were.
Click to expand...

Yeah...but the question is what data set did they use to train their model? They may have used the Rivals100 to train the model..that is, generate the statistical weighting factors for each of their variables.

It doesn't really matter, though. I was just wondering how many of those were Rivals100....we could compare back to our other Rivals100 and see what percentage they hit for us. My guess is that we could be an outlier with the fact that generally have to rely strongly on nation-wide recruiting (more so than some other schools).
 
#23
#23
sjt18 said:
The article didn't limit it to Rivals 100.... However, Jones, Berry, Martin, Donald, and Creer were.
Click to expand...

The model predicted the following for these guys:

Berry (Auburn)
Martin (OSU)
Creer (OU)
G. Jones (TTU)
Donald (UT)

So, as we suspected, Tennessee was definitely an outlier school in this model. Other schools were predicted much better - which, I think, tells us what we already know - the staff can't rely on a slew of in-state or nearby talent to bring in recruits.
 
#24
#24
You know...I started looking at some of their data, and I think that it really speaks to the strength of our recruiting class that the model got so many of our commits wrong. This means that the staff basically had to go against prevailing tendencies to get some of these kids. I know that there are some special situations like Eric Berry that the model didn't take into account (familial connections - which I think that it should). Here are some numbers though. The "right" means that the model predicted the recruit would go to that school and the recruit did. The "wrong" means that the model predicted that the recruit would go to another school but the recruit in fact came to this school.

Florida: 9 right, 1 wrong -- 90% right
Georgia: 3 right, 0 wrong -- 100% right
S. Carolina: 4 right, 1 wrong -- 80% right
LSU: 5 right, 2 wrong -- 71% right
USC West: 7 right, 3 wrong -- 70% right
Cal: 1 right, 0 wrong -- 100% right
Notre Dame: 4 right, 3 wrong -- 57% right
Tennessee: 1 right, 4 wrong -- 20% right

I find that interesting. It would be even more interesting to examine it over a number of years. But, alone, it makes me feel good about our recruiting efforts (granted, this was a great class - that's why I would like to look at other years).
 
#25
#25
Give credit where credit is due. CPF is a master recruiter and if he had no other value as a HC... that might well be so critical that he's the best fit for UT.

One thing is certain... few if any will sell UT with more of his heart.
 
New posts
You must log in or register to reply here.

VN Store



Back
Top