What's new
AllBuffs | Unofficial fan site for the University of Colorado at Boulder Athletics programs

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

  • Prime Time. Prime Time. Its a new era for Colorado football. Consider signing up for a club membership! For $20/year, you can get access to all the special features at Allbuffs, including club member only forums, dark mode, avatars and best of all no ads ! But seriously, please sign up so that we can pay the bills. No one earns money here, and we can use your $20 to keep this hellhole running. You can sign up for a club membership by navigating to your account in the upper right and clicking on "Account Upgrades". Make it happen!

Recruitment Rankings vs. On Field Success

TimmyDUBs

Dirty haole
Club Member
v8oln0qsi9qkmfs3jg0d.jpg


Linked from Deadspin: http://regressing.deadspin.com/char...ource=deadspin_facebook&utm_medium=socialflow

Kind of interesting.
 
depressing validation of our general evaluations of both our recruiting and coaching since 2005. The nice thing is that even average coaching is going to do a whole lot for us.
 
Last edited:
Our performance on the field has been worse than our recruiting.

It's like picking out the least rotten apple, but it is what it is. CU football has under-performed its talent in recent years.
 
Interestign that the SEC on the whole does less wqith more from a recruiting ranking standpoint. Only Mizzu is clear of the line and: Georgia, T, Ole Miss, Uk, and Auburn are all well below it.
 
Interestign that the SEC on the whole does less wqith more from a recruiting ranking standpoint. Only Mizzu is clear of the line and: Georgia, T, Ole Miss, Uk, and Auburn are all well below it.

It also looks like there's something off. P5 schools are under-represented on the "better on the field" side of the line.

Maybe the recruiting services over-rate P5 recruiting classes.

Maybe the ranking system under-rates the SOS of P5 programs.

Probably a combination of the two.

Take CU for example. If we took our same recruiting classes and played in the Sun Belt, Buffs would have been winning 8+ games a year during our recent struggles. Our exact same recruiting classes would have been ranked lower, though. But our power ranking would have been higher. That would have not meant that Hawkins and Embree were worse recruiters and better coaches, but I think this analysis would have painted them that way.
 
Our performance on the field has been worse than our recruiting.

It's like picking out the least rotten apple, but it is what it is. CU football has under-performed its talent in recent years.

Isn't that the hope we have, that MM will coach up the under-utilized talent we have, and use that to leverage into better talent? Seems to be working so far. Just a longer process that most hoped for.
 
The chart basically shows that you perform as well as you recruit. Not many outliers.
 
It also looks like there's something off. P5 schools are under-represented on the "better on the field" side of the line.

Maybe the recruiting services over-rate P5 recruiting classes.

Maybe the ranking system under-rates the SOS of P5 programs.

Probably a combination of the two.

Take CU for example. If we took our same recruiting classes and played in the Sun Belt, Buffs would have been winning 8+ games a year during our recent struggles. Our exact same recruiting classes would have been ranked lower, though. But our power ranking would have been higher. That would have not meant that Hawkins and Embree were worse recruiters and better coaches, but I think this analysis would have painted them that way.

Nik - its not wins against competition hey are looking at it is team rank in a similar fashion to the BCS so some of that is off-set for-instance an 8-3 sunbelt team wouldn't register you'd have to be Boise or NIU good to get recognized. I agree it doesn't eliminate the bias but it reduces it significantly beyond what your asserting.
 
The chart basically shows that you perform as well as you recruit. Not many outliers.

Disagree but the article says it, there are out-liers like: Snyder, Boise, Wiscy, Stanford, TCU, Cincy and on on the up side. On the Down look at schools like Tenn/Cal/Memhis/us. We should be somewhere in the 60-80 range with how we've recruited ye we manged to touch the 100+ range, Cal and Tenn have pulled repeated top 25 classes of late and cant even finish .500
 
The combination of inflated recruits that shouldn't have had the rating they did, attrition issues, plus bad coaching all pretty much make it look like we recruited well yet never played well. I'd say we both recruited horribly and had horrible coaching, and I doubt that is a controversial statement....
 
Disagree but the article says it, there are out-liers like: Snyder, Boise, Wiscy, Stanford, TCU, Cincy and on on the up side. On the Down look at schools like Tenn/Cal/Memhis/us. We should be somewhere in the 60-80 range with how we've recruited ye we manged to touch the 100+ range, Cal and Tenn have pulled repeated top 25 classes of late and cant even finish .500

I'm not bored enough to do it but if you fit the data, you'd find that the line in the middle is probably pretty dead on. I doubt there are many programs that fall outside 2 standard deviations (there are some, of course).
 
Last edited:
I'm not bored enough to do it but if you fit the data, you'd find that the line in the middle is probably pretty dead on. I doubt there are many programs that fall outside 2 standard deviations (there are some, of course).

Welcome to statistics 101. In a normal distribution, approx. 95% of observations always fall within 2 standard deviations of the mean. Because the way you calculate the standard deviation is based on the amount of variation in the data set. So, the more variation, the larger the standard deviation.

There is is a good linear correlation here, and also a very large amount of variance not explained by recruiting rank.
 
I'm not bored enough to do it but if you fit the data, you'd find that the line in the middle is probably pretty dead on. I doubt there are many programs that fall outside 2 standard deviations (there are some, of course).

2 standard deviations is a pretty ****ing high bar.
 
Welcome to statistics 101. In a normal distribution, approx. 95% of observations always fall within 2 standard deviations of the mean. Because the way you calculate the standard deviation is based on the amount of variation in the data set. So, the more variation, the larger the standard deviation.

There is is a good linear correlation here, and also a very large amount of variance not explained by recruiting rank.
Too bad the study didn't go a step further and with a regression analysis
 
Too bad the study didn't go a step further and with a regression analysis
Be interesting to see a multiple regression. I'd like to see a few more independent variables, like AD budget, prior year record, SOS etc.
 
Be interesting to see a multiple regression. I'd like to see a few more independent variables, like AD budget, prior year record, SOS etc.
The Massey rating incorporates SOS, but yeah there's a lot of other variables that could be significant that are missing
 
There is also a huge difference between recruiting class rankings and the actual quality of the class.

CU had some classes under Hawk and Embree that were perfect examples of this. On LOI day they ranked fairly highly based on a number of highly ranked players. Six months later the guys at the bottom of the class were all in camp but missing were a number of 4* or high 3* guys who weren't admitted or had other problems like Katoa. Other guys had their injury issues surface like MTM and others showed why they were overrated like Scott.

An interesting study would be to correlate the schools that see the majority of the top ends of their classes end up contributing and wins. Bama would be a good example of this. Off the top of my head I'd bet that over time Michigan State, Wisconsin, Stanford, and some others would head this list.
 
The chart basically shows that you perform as well as you recruit. Not many outliers.

Or you recruit as well as you perform. Not clear which drives the other. There's an argument to me made -- not one I necessarily agree with -- that Alabama wins so they recruit well and their recruits are rated higher because the Tide wins so much. You can certainly over- or under-perform your recruiting. Boise obviously has the reputation.

Also, if you just graph the P5 teams and Norte Dame by total wins (especially conference wins) and recruiting ranks, I bet there are fewer teams far from the line. Teams like Colorado, Maryland, Illinois and Tennessee are going to have higher recruiting rankings than most MAC and MWC teams even though many of them will win more games and, as a result, get higher computer rankings.
 
Be interesting to see a multiple regression. I'd like to see a few more independent variables, like AD budget, prior year record, SOS etc.

Good idea. Got any recommended convenient sources of this kind of info?

An interesting study would be to correlate the schools that see the majority of the top ends of their classes end up contributing and wins. Bama would be a good example of this. Off the top of my head I'd bet that over time Michigan State, Wisconsin, Stanford, and some others would head this list.

How would you be able to gather that data objectively / not stacking the deck towards a favored conclusion?

if you just graph the P5 teams and Norte Dame by total wins (especially conference wins) and recruiting ranks, I bet there are fewer teams far from the line. Teams like Colorado, Maryland, Illinois and Tennessee are going to have higher recruiting rankings than most MAC and MWC teams even though many of them will win more games and, as a result, get higher computer rankings.

This would be pretty easy to validate.

I wouldn't mind crunching some of these hypotheses. Maybe I'll open up a bottle of rye whiskey and give it a whirl.
 
Welcome to statistics 101. In a normal distribution, approx. 95% of observations always fall within 2 standard deviations of the mean. Because the way you calculate the standard deviation is based on the amount of variation in the data set. So, the more variation, the larger the standard deviation.

There is is a good linear correlation here, and also a very large amount of variance not explained by recruiting rank.

Sorry, you're right, of course. I'm too old to remember back that far.

The deviation in that data just doesn't look very large to me.
 
How would you be able to gather that data objectively / not stacking the deck towards a favored conclusion?
.

Part would be subjective. On the other hand you could do a recruiting ranking using the same methods they use on signing day but do it on the opening day of the first season and second season subtracting off the kids who failed to be admitted or otherwise aren't on the teams. Beyond that you could subtract the kids who either not participated or been on injured list for injuries that they are documented as having prior to signing. I.E. by the time MTM got to CU he had already missed a season of HS due to a full knee reconstruction. In less than a year at CU he had re-injury of the same knee. He was highly ranked coming out of HS but his first choices including USC backed off because of the injury and fear of re-occurance. His high rating bumped up our ranking in that class but he never contributed much, unfortunately.

You could make the same discussion about Shane Dillon who had known injuries before he ever signed, turned out that due at least in part to those injuries he never played for us but he counted in the class.

A non-injury example would be Ed Nuckols, 4* DT who was so bad academically that he had trouble staying eligible at his JC where he landed but he gave our LOI day ranking a big boost. Again other schools looked at him, a number of major schools offered but then pulled the offer when it became obvious that he wasn't going to qualify.
 
I wish we were as average as Oregon State and Louisville.

That's kind of what I expect out of hcmm ceiling wise unless he proves he can recruit at a higher level. That's where cu needs to get to before taking the next step
 
That's kind of what I expect out of hcmm ceiling wise unless he proves he can recruit at a higher level. That's where cu needs to get to before taking the next step

I'm terrified that we will have 5 years of 7 and 8 win seasons, maybe hit 10 once and then decide we need a coach to get us to the next level so we part ways with Mac. We will then the wrong coach and suck again.
 
I'm terrified that we will have 5 years of 7 and 8 win seasons, maybe hit 10 once and then decide we need a coach to get us to the next level so we part ways with Mac. We will then the wrong coach and suck again.

Don't worry, he'll leave for a better job before that happens.
 
Our facility's will be some of the best in the country. If he can get CU to 8 wins we are a top20 job nationally again
 
I'm terrified that we will have 5 years of 7 and 8 win seasons, maybe hit 10 once and then decide we need a coach to get us to the next level so we part ways with Mac. We will then the wrong coach and suck again.
I am more worried about him getting tired of losing and leaving for greener pastures
 
Back
Top