How do they turn out?

Submitted by Ron Utah on January 29th, 2013 at 12:40 PM

An informational post about the Rivals 100 players Michigan has recruited since 2002 got me thinking, and in this relatively quiet period, I decided I wanted to dig a bit deeper.

The question I set out to answer: How do these guys turn out?  At what rate do top recruits become top players in our program?  And how does that compare to other programs?

Given limited time, I compared us to only one other program: Ohio.  I used Rivals 100 data for position, stars, and rank.  The "Impact" data point is my subjective interpretation of a player's career impact; 3 is a high impact player (Solid starter to All-B1G type), 2 is a role player (contributor to starter), and 1 is a low impact player (did not produce for whatever reason).  These ratings are NOT based on talent or careers at other schools--only the player's impact where they signed their LOI.  Players who have not yet had the opportunity to demonstrate a rating are designated "n/a".  Players with an asterix have not yet signed.   And yes, some of you will argue with me, but my overall ratings are close enough to make some good starting points for conversation.  Here is the data, followed by conclusions:

Chad Henne QB 5 13 2004 PA 3
LaMarr Woodley LB 5 14 2003 MI 3
Brandon Graham LB 5 15 2006 MI 3
Donovan Warren DB 5 25 2007 CA 3
Stephen Schilling OL 5 26 2006 WA 3
Gabe Watson DT 5 33 2002 MI 3
Shawn Crable DE 4 39 2003 OH 3
Jonas Mouton DB 4 45 2006 CA 3
Mario Manningham WR 4 45 2005 OH 3
Prescott Burgess DB 5 6 2003 OH 2
Will Campbell DT 5 26 2009 MI 2
Tim Jamison DE 4 40 2004 IL 2
Darryl Stonum WR 4 41 2008 TX 2
Greg Mathews WR 4 93 2006 FL 2
Ryan Mundy DB 4 96 2003 PA 2
Terrance Taylor DT 4 96 2005 MI 2
Ryan Mallett QB 5 4 2007 TX 1
Kevin Grady RB 5 22 2005 MI 1
Justin Turner DB 4 35 2009 OH 1
Marques Slocum OL 4 37 2005 PA 1
Carlos Brown RB 4 39 2006 GA 1
Boubacar Cissoko DB 4 44 2008 MI 1
Antonio Bass ATH 4 49 2005 MI 1
Dann O'Neill OL 4 49 2008 MI 1
Justin Boren OL 4 64 2006 OH 1
Clayton Richard QB 4 71 2003 IN 1
Alex Mitchell OL 4 80 2004 MI 1
Cory Zirbel OL 4 83 2005 KY 1
Jim Presley LB 4 89 2003 MI 1
Adam Patterson DT 4 91 2006 SC 1
Toney Clemons WR 4 91 2007 PA 1
Matt Gutierrez QB 4 96 2002 CA 1
Brett Gallimore OL 4 96 2004 MO 1
James McKinney DT 4 98 2005 KY 1
Doug Dutch WR 4 98 2004 DC 1
Cullen Christian DB 4 99 2010 PA 1
Derrick Green (*) RB 5 8 2013 VA n/a
Ondre Pipkins DT 5 14 2012 MO n/a
Kyle Kalis  OL 5 22 2012 OH n/a
Henry Poggi (*) DT 4 70 2013 MD n/a
Erik Magnuson OL 4 78 2012 CA n/a
Shane Morris (*) QB 4 81 2013 MI n/a
Patrick Kugler OL 4 82 2013 PA n/a
Justice Hayes RB 4 85 2011 MI n/a

Let's start by looking at Michigan's "gets".  There are some definite correlations.  A higher national rank does indeed give a player a higher likelihood of making an impact.  Of the 36 players who received a rating, nine were 3's (high impact), eight were 2's (role players), and 19 were...not so good.  That gives Rivals 100 players during this period a 25% chance of being great, a 22% chance of being okay to good, and about a 53% chance of not being helpful at all.  Basically, it's about 50/50 on whether or not these kids make a positive impact at Michigan.

That said, of the nine players who were 3's, 6 were five-star players.  Two more five-star players made a 2 rating (Burgess & Campbell), and many would argue Burgess was a 3 (erroneously, but they would argue).  That means roughly 80% of your five-star players end-up solidly contributing, and of the two that didn't--Mallet and Grady--only Grady was a complete bust, as Mallet went on to SEC stardom.

Of the 20 players who were 1's, 10 were ranked 80th or lower nationally, and only six were ranked higher than 40th.

I think it's important to consider that this time period includes two tumultuous coaching changes and a year of "lame-duck" coaching from Carr.  I do not believe it will be representative of our success going forward, but it's the data we have.

Terrelle Pryor QB 5 1 2008 OH 3
Theodore Ginn, Jr DB 5 2 2004 OH 3
Chris Wells RB 5 3 2006 OH 3
Mike Adams OL 5 3 2008 OH 3
Michael Brewster OL 5 12 2008 FL 3
Alex Boone OL 5 20 2005 OH 3
DeVier Posey WR 5 21 2008 OH 3
Donte Whitner DB 4 27 2003 OH 3
Marcus Freeman LB 4 31 2004 OH 3
Corey Brown DB 5 31 2009 PA 3
Braxton Miller QB 4 34 2011 OH 3
Maurice Clarett RB 5 37 2002 OH 3
Brandon Saine RB 4 50 2007 OH 3
Andrew Norwell OL 4 59 2010 OH 3
Doug Datish OT 4 68 2002 OH 3
Quinn Pitcock DT 4 72 2002 OH 3
Doug Worthington DE 4 80 2005 NY 3
Robert Rose DE 5 17 2006 OH 2
Etienne Sabino LB 4 46 2008 FL 2
Garrett Goebel DT 4 64 2008 IL 2
J.B. Shugarts OL 4 87 2008 TX 2
Mike D'Andrea LB 5 29 2002 OH 1
Lamaar Thomas ATH 4 33 2008 MD 1
Louis Irizarry TE 4 38 2003 OH 1
Justin Zwick QB 4 40 2002 OH 1
Connor Smith OL 4 55 2006 OH 1
Kyle Mitchum OL 4 56 2004 PA 1
Jaamal Berry RB 4 56 2009 FL 1
Eugene Clifford DB 4 60 2007 OH 1
Kenny Hayes DE 4 68 2011 OH 1
Raymond Small WR 4 88 2006 OH 1
Duron Carter WR 4 90 2009 FL 1
Dorian Bell LB 5 33 2009 PA 1
Jamie Wood DB 4 74 2009 OH 1
Curtis Grant LB 5 2 2011 VA n/a
Noah Spence DE 5 9 2012 PA n/a
Adolphus Washington DE 5 25 2012 OH n/a
Mike Mitchell* LB 5 26 2013 TX n/a
Jalin Marshall* ATH 4 35 2013 OH n/a
Cameron Burrows DB 4 39 2013 OH n/a
Michael Bennett DT 4 41 2011 OH n/a
Joey Bosa* DE 4 47 2013 FL n/a
Marcus Hall OL 4 52 2009 OH n/a
Tommy Schutt DT 4 64 2012 IL n/a
Roderick Smith RB 4 65 2010 IN n/a
Ezekiel Elliott* RB 4 84 2013 MO n/a
Eli Apple DB 4 89 2013 NJ n/a
Evan Lisle* OL 4 90 2013 OH n/a
Se'von Pittman DE 4 95 2012 OH n/a

Ohio's data gives us 35 rateable recruits to our 36.  They show a similar correlation, with higher rankings and five-star players much more likely to be 3's.  Of their 35 rated players, 17 were 3's, 4 were 2's, and 13 were 1's.  That means roughly half (49%) of their rated players were 3's, and about 37% were 1's.  Interestingly, many of their 1's were players who had trouble with the law--an issue that was much less prevalent with Wolverines.

The comparisons are pretty obvious: Ohio has gotten much more production out of their top recruits.  This is, no doubt, partially attributable to mostly consistent coaching through the period by one of the best in the game (even if was a lying cheater).  Ohio also had higher-ranked recruits--their average national ranking is 45.9 to Michigan's 55.2--and were much more geographically concentrated in Ohio and the midwest than Michigan's players.

Another interesting bit of data is that position does not seem to make much of a difference.  LBs are probably the most successful recruits, but it matters very little.  National ranking seems to correlate with impact regardless of position.

Going forward, my expectation is that roughly two-thirds (60-66% would be good) of Rivals 100 recruits end-up as solid contributors or better for Michigan, with about half becoming impact players.  Unfortunately, the lower rankings of this year's four Top 100 recruits (Morris is 81 and Kugler 82) would suggest they have a smaller chance of being successful, while Poggi is most likely to be at least a contributor and Green has a 50/50 chance of being great.  If Green finishes his career as a 3, and we get two 2's out of the other three, it will have been a very good year.  If there are two 3's, it's a great year, and if there are two or three 1's, things didn't go so well.

I do believe our success with top talent will say a lot about or staff and look forward to revisiting this in 2016, when Hoke has had a full five-year cycle to demonstrate how well he can maximize talent.


EDIT: After some honest thought and good criticism, I bumped Will Campbell up to a "2".  It's a "meh" difference statistically, but he probably earned it this year.



January 29th, 2013 at 1:53 PM ^

I agree with the overall conclusion and found it quite interesting. I have one concern and its in your ranking system. I think a 4 point system provides a more accurate comparison 4 being all conference, 3 solid starter, 2 starter, 1 low impact. On your list I see carlos brown and campbell being rated as 1's even though they provided quality depth and started their senior years, (campbell all year brown sporadically) and provided an impact on the team. even boren and cissoko were starters for a year and did well before they transfered and were kicked off the team respectively. I get that you are trying to encompass all 4 years and evaluate the overall impact but i feel in doing so you fail to take into account the development of players as even the highly rated may need time/coaching to get college ready (campbell) or be blocked by others at their position (brown).

Ron Utah

January 29th, 2013 at 4:22 PM ^

I don't disagree with any of your criticisms about the structure of the data.  In fact, I looked at a four-tiered rating, but found that there were very few "middle" players (2's and 3's), and the ratings got even more muddled.

I also agree that I may have been a bit harsh on Campbell.  He may very well be a 2.  But remember that he only had 14 tackels in 2011, and wasn't on the field much.  As for Brown, Cissoko, and Borern--none of those guys made enough of an impact to earn a 2, IMO.  Cissoko played a lot but was terrible; Brown was inconsistent at best, and Boren left before he was really good.  You really earn a 2 by postiviely contributing to our win-loss record; I'm not sure any of those guys achieved that.

And Campbell's problem wasn't development or coaching, he was fat and lazy.  And Brown just wasn't good enough.  I see your argument, I just agree to disagree.

EDIT: Bumped Campbell to a "2".  


January 30th, 2013 at 12:35 AM ^

which is very...nitpicky, is that I think you could control some for percentages of success by counting people who became good contributors elsewhere, as that speaks to their performance/upside in general. I understand it's the upside for us we want to consider mainly, but it would be a more accurate representation of the actual quality of the recruit to assess how they did as football players. I would imagine finding the data for the ones who left might make it harder for you.

Ron Utah

January 30th, 2013 at 9:59 AM ^

I appreciate the idea.  Actually, it's not hard to find the data for the guys who leave.  But I'm not trying to measure our ability to evaluate talent, I'm trying to examine our ability to maximize the talent we bring in.  Part of that process is keeping your talent.

Michigan does not bring in many JUCOs or transfers, so when we see a guy leave, he's not often replaced by a similarly-talented player.  So guys we lose that become contributors elsewhere are still busts for us, even if they end-up being good elsewhere.

The Ryan Mallets and Justin Borens of the world are great players, but that doesn't help us win games.