2014 Signing Day By The Numbers

Submitted by The Mathlete on February 5th, 2014 at 2:46 PM

Comparing individual classes of recruits can be a very challenging exercise. Due to each school’s different position each year and internal standards, classes can vary from mid-teens to over 30 signees for a given class at a given school. Last year I introduced my best take on the subject with the Nth best recruit approach.

Nth best recruit takes each player in a class and gives them a rating from 0-99 and then places those rankings in order, high to low. This way you can see how one class compares to another at each level. With Jabril Peppers in the mix, Michigan is going to compete with everyone at the top of the class but then drop into very good range as the recruits progressively move from high four star territory (Drake Harris) to high three star (Brandon Watson) with former gray shirt candidate Brady Pallante pulling in the final spot.

One change for this year is that I have normalized the classes so that they all show how the class is dispersed as if they were a 25 person class. You lose the quantity estimate, but over time, the spread of recruit’s rankings are more indicative of a team’s recruiting prowess than the number of offers they have in a year. If Michigan had five more offers, the odds are their curve would look very similar to what it does now.

*All signee lists were updated as of late Tuesday night and don’t reflect any signing day action

Michigan Under Hoke


Michigan’s last three classes have been highly consistent in terms of recruit quality from top to bottom. Last year’s class was the strongest through the top half and this year’s class is nearly identical in ratings to 2012’s class with Pepper’s the welcome exception.

On an average basis, Michigan’s classes have landed them roughly in lower part of the top 10 nationally. The improvement in these classes will begin to show up this year as my prior studies have shown that player in their third year or more on campus are far and away the biggest predictors of success. The 2012 class enters that zone this year and Michigan should move near the top 10 in terms of overall roster talent+experience this season and move into the top 10 indefinitely beginning in 2015.

The Big Ten Leaders Legends East!


After a very close comparison last season, Michigan’s 2014 class is clearly lower rated than Ohio State’s. Michigan’s class falls behind immediately after Peppers and maintains a similar gap until the final few players.

Michigan sill is quite a bit ahead of the rest of the division. Penn St and Michigan St are in the next tier. There is a consistent gap between them and Michigan and Penn St’s class is currently slightly higher rated than Sparty’s across the first few spots.

The top third of Maryland’s class is in line with Michigan St and Penn St but they quickly fall into line with the bottom tier of Rutgers and Indiana.

The Race For #1


The five teams rated highest on most services

The bottom half of all the great classes this year are virtually in distinguishable from each other, except for Tennessee. While the Volunteers have put together a really nice class, this chart helps expose the formulas used by all the major services for team rankings. Tennessee is rated no lower than sixth overall by any of the four major services and although they have a very good class, you can see the separation between the great classes and theirs. Getting a giant class isn’t about being better at recruiting, it’s about having a fluky roster situation. Almost all coaches are going to recruit to their 85 (or more) roster spots so having more commitments is vastly overrated.

Ohio State has the weakest top end of all the four serious contenders but the middle third of their class is as good as anyone’s. Texas A&M’s class shows a big drop after the marquee headliners. LSU is strong throughout but Alabama, once again, clearly has the class from top to bottom. If you take any spot along the line of 25, the Alabama point is rarely behind any other team and no one is as consistently strong as they are.

Michigan’s Hall of Highly Touted

In the past two classes, Brady Hoke has inked eleven players that made the first or second team for Michigan’s Hall of Highly Touted.* After two loaded classes, this year’s smaller class was also lighter on top rated talent. Drake Harris cracked the second team as a wide receiver while all-everything signee Jabril Peppers was a no-brainer first team defensive back.

Peppers scored a 96.5 out of a possible 99 (unanimous #1 rated recruit) which makes him the highest rated recruit at Michigan in the internet era of recruiting. When you expand the field beyond Michigan to the whole Big Ten, Peppers comes in at #2 behind Terrelle Pryor (97.9) for highest rated Big Ten signee over the last 12 classes.

*The top players based on composite recruitindg rankings



February 5th, 2014 at 2:01 PM ^

The best rating system I have seen by far.  Normalizing to a standard 25 man class was an excellent tweak to what was already leagues ahead of the competition.  Ratings that still lean on class size to the detriment of quality are ridiculous.  A gold medal to the Mathlete on the eve of the Olympics.  Thank you.


February 5th, 2014 at 3:39 PM ^

Looks great.  It still shocks me that Alabama has as many spots open as UM but pulled in 9 more kids.  So when you combine that math with these types of hauls, no wonder they are so good.


February 5th, 2014 at 3:43 PM ^

I was trying to compare the charts - it appears that Michigan and Tenneessee are about even using the Nth best recruit approach, with Michigan clearly having the best recruit - is that right?

Swayze Howell Sheen

February 5th, 2014 at 4:20 PM ^

though the normalization can lead to problems of course.

One might deal with the normalization issue by also including a summary of the entire roster (thus summing recruiting rankings over the years).

Nice work!



February 5th, 2014 at 4:41 PM ^

You should compute some sort of Area Under the Curve metric for these plots.  A nice way to summarize and compare.  You could probably even put a p-value on differences!


February 5th, 2014 at 5:16 PM ^

ESPNU has been showing how many ESPN Top 300 recruits each school has signed.  Michigan has signed an incredible 9 Top 300 recruits, more than schools such as Alabama, Ohio State, etc.  2014 is a very strong class for Michigan.

The Geek

February 5th, 2014 at 6:33 PM ^

Excellent work, per usual.

Assuming MM was not factored into Sparty's numbers as of Tuesday night? Come on FSU!

Seriously though, this is awesome. I was surprised at how good this class looks in the first graph. What a great day!


February 5th, 2014 at 8:43 PM ^

I think you should have stuck with the chart without normalizing, or at least shown then both.

I love the work and think it's the best evaluation out there but this tweak is a (minor) step back.

While average rank is certainly far better than the algorithms heavily based on quantity, I think it's still worth acknowledging that bigger classes are more productive.  In a vacum, having a class of 25 guys averaging 3.9 stars is better than a class of 18 averaging the same - there is no question Alabama is well-served by oversigning and cutting players.  Getting five 5-stars in a class of 25 is better than getting four 5-stars in a class of 20...and so on.  Normalizing at the top end skews the comparison of classes. At the bottom end, it doesn't matter if you land five 3-stars or two - that's just a function of attrition, demand for kickers and other irrata.  I'd rather just ignore the tail and focus on the left half of the graph...but now that it's normalized that's harder.



February 5th, 2014 at 11:33 PM ^

set aside the numbers, would you swap this class for any class in the B1G?  I wouldn't even consider it.  OSU has better numbers.  don't care.  

numbers aside, few in the country are comparable.