Yay: Michigan 10th worst in Recruit-to-Performance Rankings

Submitted by JamesBondHerpesMeds on

http://regressing.deadspin.com/chart-which-ncaa-football-teams-outplay-…

They've taken the average recruit ranking less the average BCS ranking (using Massey's poll) amongst all CFB teams. With an average BCS ranking of 39.8 and recruiting class ranking of 12.4, the -27.4 differential is 10th worst in college football.

Who is worse? Kansas (obvs!), Maryland, Colorado, Tennesse, Virginia, Cal, Illinois, Memphis, Ole Miss. Good company.

Our friends at Boise State, with a differential of +52.6, are 3rd best. Crushin' it. Michigan State is at +6.6; Ohio State at -6.4. Wisconsin is the Big Ten frontrunner with a +31.1 differential.

 

alum96

October 3rd, 2014 at 1:32 PM ^

It's a 5 year average so the 2011 record (11-2) really helps. If 2011 was 8-5 or something I think we'd be top 4ish.  Someone like TN is almost impossible to beat - they get top 25 classes and have had multiple 4-5-6 win seasons.

Some of the teams below us (above us?) recruit "ok" but have had multiple 3, 4, 5 win seasons.  We recruit "great" and have had 7-8 win seasons plus one outlier. 

Tater

October 3rd, 2014 at 2:04 PM ^

AFAIC, the problem is that the offensive scheme doesn't put players in a position to fully utilize their talents.  It is so bad that one of the popular performance indices has Michigan as an underdog in five of the next seven games, according to ESPN.  

I feel really sorry for Devin Gardner.  He deserved a lot better.  I hope fans will forever hold him in high esteem and be grateful for how he has represented the program.

alum96

October 3rd, 2014 at 1:24 PM ^

Thanks for the link - data like this is fun to pour over for us data nerds.

I was going to get snarky and say "give us 6 weeks and we could climb into the top 5" but this is a rolling 5 year average so 2011 alone should keep us out of that storied level. 

Wisconsin is the #1 Big 10 team at +31.1

Others of interest:

  • TCU: +30.0
  • Stanford +23.8
  • Kansas State: +23.1
  • Oregon: +15.9

 

On the bad side:

  • Maryland: -45.3
  • Colorado: -44.2 (people Wolverine Devotee's age dont realize Colorado was a NC in  90 and had a Heisman winner in 94!)
  • Tenn: -40.4
  • Virginia: -40
  • Illinois: -36.4
  • Michigan: -27.4
  • UCLA: -25.7
  • Minn:  -24.1
  • Miami (YTM): -23.7
  • Texas: -19.4

 

No surprises there - as Kelly has improved Notre Dame, Miami has come down to replace ND as a fallen angel... and joined UM and Tenn as the 2 major tire fires relative to recruiting and the previous 20 years of performance.

M-Dog

October 3rd, 2014 at 2:51 PM ^

The interesting thing about the data is that it confirms the eye test of which programs are generally viewed as "well coached" and which teams are generally viewed as "poorly coached".

If you're in the market for a new coach (hint, hint), look at some of those guys coaching those +20 and above teams. 

FreddieMercuryHayes

October 3rd, 2014 at 1:25 PM ^

Are these adjusted for youth in any way?  Because that's a big confounder.  A senior 5 star is a lot different from a freshman 5 star.  It would also be interesting to see how the data changes for different positions in which some are less likely to predict future success than others.

alum96

October 3rd, 2014 at 1:31 PM ^

Here is how they did it.  It is fair - rolling 5 year average Rivals ranking for EACH season.  Can't complain about it. 

We compared how teams recruited to where they ended up in computer polls at the end of each of the last five seasons. To measure on-field success, we used Kenneth Massey's ranking composite. Massey is a statistician whose work contributed to the BCS computer rankings; his composite index averages dozens of rankings including the six computers used in the BCS, the AP Poll, and the USA Today Coaches Poll.

Rivals rankings were used to measure recruiting. For each season, we used an average of the five previous recruiting classes. Even though upperclassmen generally contribute more than underclassmen, we avoided weighted averages because upperclassmen also transfer schools, declare for the NFL draft early, and have career-ending injuries.

LSAClassOf2000

October 3rd, 2014 at 1:40 PM ^

Here are the averages and differential for the Big Ten, taken right from their aggreggate table. The average differential for a Big Ten team is -9.65, based on this:

Team BCSAvg  RecAvg  Diff 
Wisconsin 18.4 49.5 31.1
Northwestern 55.4 68.2 12.8
Michigan St 26.8 33.4 6.6
Iowa 37.6 43.9 6.3
Nebraska 22 21.8 -0.2
Penn St 33.4 28 -5.4
Ohio St 16.6 10.2 -6.4
Indiana 87.6 74.7 -12.9
Rutgers 59.2 43.3 -15.9
Purdue 80 62.1 -17.9
Minnesota 71.2 47.1 -24.1
Michigan 39.8 12.4 -27.4
Illinois 77.2 40.8 -36.4
Maryland 78.8 33.5 -45.3

 

JamesBondHerpesMeds

October 3rd, 2014 at 1:54 PM ^

I recalculated the numbers eliminating the 2009 and 2010 BCS rankings (but retained the recruiting rankings since '09 recruits would be hypothetically contributing to the team in 2011).

The results do lean more in our favor, as we had a performance average of 24.3. Less our recruiting average, that gives us a differential of -11.9. That would put us roughly around the same spot as Florida State. 

Avon Barksdale

October 3rd, 2014 at 1:55 PM ^

I'm really disappointed with that ranking. I'd prefer us to be first just so it can be yet another sucker-punch to the jaw - no pun intended towards C'Sonte York of course.

gwkrlghl

October 3rd, 2014 at 2:18 PM ^

There's only one other school who recruits at our level who is worse: Tennessee. So our thoughts of "We're not #1?" are pretty much true.

  1. Kansas
  2. Maryland
  3. Colorado
  4. Tennessee
  5. Virginia
  6. Cal
  7. Illinois
  8. Memphis
  9. Ole Miss
  10. Michigan

A lot of those schools are middling Big 5 schools who have been terrible lately. Only Michigan and Tennessee can really boast that they can take teams of 4* players and make them look like they belong in the MAC

M-Dog

October 3rd, 2014 at 3:07 PM ^

Tennessee has been a tire fire for years.  I'm not happy to be joining them.

UCLA also has a similar profile, but they look to be coming out of it because of the right coach.

That's the good news . . . there are certain perennial powers that can wander in the wilderness for years, but that can be fixed simply by hiring the right coach.  Michigan is one of them.

Others that fit this profile:

  Texas

  FSU

  Penn State

  USC

  ND

  Auburn

That's good company to be in.  

Hannibal.

October 3rd, 2014 at 3:09 PM ^

By my count, 10 of the 14 B1G teams are below the line.  Even Ohio State.  That's saying a lot considering that Big Ten recruiting was already garbage to begin with.  Our coaches suck ass collectively at developing players as well.

alum96

October 3rd, 2014 at 4:01 PM ^

Yes but the issue is a lot of these coaches are already gone... or a guy like Bill Snyder who is never leaving for another school...or its a service academy.  A lot of guys at these schools are now in their 1st or 2nd year as other schools snapped up their coaches. 

A school like Cincy has hired Dantonio, Kelly, and Butch Jones - an impressive array.  That AD should be heading our coaching search.  But they currently have a stop gap coach rather than an up and comer (Tommy Tuberville who is 60 years old).   Northern Illinois is another good one - they had Jerry Kill and Dave Doeren,both now in Power 5).  Boise State just lost their coach to Washington and has a mostly miss record of sending their coaches to Power 5s and succeeding.  TCU's coach has been there 16 years or so. Navy and Air Force are there because I bet almost none of their guys have any stars.  Stanford is Stanford.  BYU has an alum coaching who is probably never leaving.  Nevada and Utah State are on coaches beginning their 2nd year so not responsible for those good records.  And so on and so forth.

Lower on the list you have Baylor (Briles), OK State (Gundy), and Oregon.  Near them are Tulsa (Todd Graham) and Houston (Sumlin).  

michelin

October 3rd, 2014 at 6:34 PM ^

I do not buy their explanation for not giving different weights to different years of seniority.  The authors said they did that because some of the class members would no longer be around when they are seniors.  True, but if these members are not around and you still include them in the recruiting rankings--which they did--- then teams with coaching changes and high attrition will almost surely underperform these spurious rankings of the the class.

I do not have time to correct for such attrition, but I did at least redo the analysis with different weights different to given years (more weight proportionately given to a recruiting class in predicting performance four years later, less given to it in predicting immediate performance). 

One can divide the teams into better or worse than average performance teams.

Consider first only the group of above average BCS performers.  Four teams clearly separated themselves out as underperforming their recruit rankings: Tenn again was an outlier and by far the worst; and the next worst group included Miss, Mia, and Fla.  The next group did include UM as well as Tex, USC, Auburn, WVa, Pitt, UCLA, NC, Ark. 

Among the lower than average BCS performers, there were 19 teams worse than UM.  So UM was not among the ten worst teams or even among the worst 25 (it was about #32)  

You might attribute these teams' underperformance to poor player development, but there were a lot of coaching changes in the underperformers.  Plenty of player attrition also occurred during the transitions.  eg Fla had enormous attrition in the early transition between coaching regimes at least thru the 2013 class, which was #3 or #4 nationally in Rivals.  Since they did not possess many of the recruits initially ranked so highly, one would expect them not to do as well as might be otherwise expected from the rankings of the (absent) recruits.

What makes UM diff from some of underperformers, like Fla, however, is that it largely retained and even strengthened the 2013 class (especially if you consider Ty Isaac, the #1 Rivals RB as a redshirted member of that class).  In fact, UM was either #2 or #3 in the nation—better than Urban’s best class at Ohio-- if you redo the Rivals rankings based on the former recruits who are still around.

If there is again a coaching change again at UM, however, it would be critical to get the new guy in right away to prevent attrition.  I wonder if this might be an argument for Schiano.  I know that the Pats coach thought very highly of him and drafted an inordinate number of his players from Rutgers.  His is immediately available and some think this week’s game at Rutgers could facilitate his recruitment.

AlbanyBlue

October 3rd, 2014 at 11:54 PM ^

I guess this is a decent place to put this, as there's no need to start ANOTHER thread about it.

In the last few weeks, I've watched many non-Michigan CFB games, and I'm seeing offenses click all over the place. OLs open holes, RBs run with burst and vision, and QBs make different throws requiring combinations of touch, timing, and arm strength. Teams like Utah State (that I am watching now, just an example) do all these things at various times.

In no way does USU (for example) have the caliber of recruits that Michigan has, so it just reinforces to me how piss-poor our offensive coaching/teaching/development is. Hoke's old-boy network combined with the coaches that were retained in the changeover are truly not power-5 level.

So much potential talent being wasted. I just can't believe it's being allowed to continue -- but it is. Sad and frustrating!!