2016 Recruiting Class Analysis, Alternative to Stars

Submitted by NOLA Blue on February 4th, 2016 at 2:34 PM

This is a story about the impressive and competitive depth of Michigan’s newest recruiting class, including position groups we won nationally, and how they stack up to an overall record of 12-3-2 for the recruiting year.

Every year there is a nationwide homage paid to the players who have fought hard to obtain 5-star status, as well as those who came close to making it to the top 0.01% but missed out on that final star and are left in the lowly top 0.05% of graduating high-school football players.  I do not wish to denigrate their hard work, and extensive time spent travelling and taking on all comers.  These dudes are beasts (5-stars, 4-stars, and 3-stars alike.)

An alternative view of recruiting success for a given team would be to analyze the zero-sum addition of players by position.  Every position has its own top-24 or 48 or 96 – with the value of a top 96 WR being different than the value of a top 96 Kicker based upon the number of WRs and Ks being competed for on an annual basis by the top schools.  With whom are we playing this zero-sum game?  I would suggest that we are playing against the top 24 most-prestigious Universities in the football nation.  (You may select your own top 24 teams by prestige – we’ll probably all wind up with similar lists.)

The next assumption is that each team is attempting to recruit a perfectly balanced class each year.  Granted, this is an imperfect assumption – impacted heavily by injuries to position groups on each team, strength of position groups on each team, recruiting busts from prior classes, shifting strength of position groups coming out of high-school, coaching offense/defense strategies, etc.  However, across the 24 highest-profile teams there should end up being some standardization to these granular imperfections on the curve.  For each team seeking no TE, there is probably a team seeking four TE; for those teams seeking 6 WR, there is a balance of teams seeking 2WR.

I presume the perfect class, as averaged across the spectrum of the Top-24 teams, to include:


4WR, 2TE, 2OT, 2 OG, 0.75 C, 1 QB, 1.5 RB, 0.75FB, 0.5 APB


2 CB, 2 S, 2 ILB, 2OLB, 1 SDE, 2 DT, 1 WDE

Special Teams:

0.5 K/P, Some number of LS

(Mix and match your own numbers at each position, multiply the total numbers by 4 to acquire the number of scholarships utilized on the complete team over 4 years)

Taking the above number of players at each position every year would yield a team of 108 scholarships over a 4-year period  – far above the 85 allowed.  It is impressive to consider the purposeful impact of coaching schemes on Offense/Defense, reaction to attrition vs strength of the position over time, and necessary impact of dual-threat players/flexible linemen on these numbers.


With 24 teams, seeking the above average number of players at each position, the zero-sum game becomes:  24 teams competing for the top 96 WR.  24 teams competing for the top 48 TE.  24 teams competing for the top 48 OT and 48 OG.  18 of the 24 teams competing for the top 18 centers.  Repeat this down the line utilizing mathgebra.

What does this mean?  Stars be damned, every prestigious team is competing for overlapping contingents of the top 96 WR.  Getting a recruit from the top of each 24-person strata (i.e. #1 and #25 from a group of 48 being competed for) is a win over getting several recruits from a lower strata (i.e. #25 and #48.)  Moreover, taking above your lot from any one position is limiting the lot available for another team from the Top-24 (let alone when a pesky non-power 5 team snags a top-rated recruit.)  Think about the year that Hoke signed 5 of the top 24 LBs (among 96 total being competed for) - that left 19 upper-echelon LBs to be split among the other 23 prestigious teams… the laws of space and time dictates that multiple prestigious schools missed out on a top-24 LB.  Hoke won big, someone had to lose... and Michigan was eventually blessed with great LB play.

Thus, another look at Michigan’s current class, utilizing the composite position rank of each player published by 247, and in a format simulating an approximate line-up on the field:

K - #1, Nordin

WR - #19, Crawford

WR - #54, Hawkins

TE - #3, Asiasi

OT - #5, Bredeson

OG - #4, Onwenu

C – None

OG - #32, Spanellis

OT – None

TE - #15 Eubanks, #45 McKeon

WR/Ath - #9 Mitchell

Slot WR - #64 McDoom, #93 Johnson (ranked as WRs, still made top-96 which is impressive for slot-types.)


QB - #6 Peters

RB - #4 Walker

FB - #1, Davis

APB - #9 Evans


CB - #8, Long

S - #23, Hudson

S - #49, Metellus, #65 Gil

CB - #12, Hill  (and #9ATH, Mitchell)

OLB – None

ILB - #12, Bush, Jr

ILB - #17, Mbem-Bosse

OLB – None

SDE - #15, Kemp  (and #1DT Gary)

DT - #1, Gary

DT - #58, Dwumfour  (and #4OG, Onwenu)

WDE - #23 Johnson, #45 Uche


Without applying advanced statistics to determine deviation from a random spread of the top sought recruits by position (because I don’t know how…) it appears via eye-ball (a time-honored non-statistical test) that Michigan won big at Kicker, FB, QB, TE, OG, CB, and ILB.  Further, Michigan won at WR, RB, APB, SDE and DT, while holding serve at S and WDE.  Losses were at OT (although Bredeson on his own is a big win,) Center and OLB.

Therefore, our 2016 recruiting record was:  12 – 3 – 2.  Seven of those wins were blow-outs, one of those losses was a nail-biter (dammit, Hamilton/Swenson.)

Geaux Blue!


turd ferguson

February 4th, 2016 at 5:26 PM ^

Thanks for doing this.  It's an interesting way of thinking about things.

This is tangential, but I'm having my own "stars be damned" moment.  I think the five-star system the recruiting services use gives annoyingly little information.  

No one gets one star.  Hardly anyone at a top school gets two stars.  And hardly anyone gets five stars (~25 guys total).  That means that the vast, vast majority of prospects for top programs get one of two ratings: 4* or 3*.  There's a line in there to distinguish the 4* prospects from the 3* prospects, but (a) that line is arbitrary and (b) it hides the enormous rankings disparity within those groups.  For example, 247sports had 4* players ranked #26-#338 in the country and 3* players ranked #339-#2100.  That's a huge difference between high 4* and low 4* guys and another huge difference between high 3* and low 3* guys.

It'd be nice if these services actually used the full range of five stars, or maybe used a scale that's more nuanced than the 5-star system and more intuitive than their individual goofy ratings systems.


February 4th, 2016 at 9:12 PM ^

I think you're on to something, Turd.  As you stated, just utilizing all the stars would give us a better feel of separation among schools (more specifically, schools we care about.). That would be killer.

How about:  Assume average of 24 players to be selected by each of the top 50 schools - 1200 players total.  Pre-set the ratio of each position group to the total.  Create equal strata within the position groups and give them 5-stars, 4-stars, 3-stars, and 2-stars.  Anyone outside the top 1200 gets a 1-star.  For a continued sense of dominant NFL men among boys, give a gold star to the top 25 players each year, regardless of their position and ratio of that position.

If you're a fan of what you consider a top 50 team, and your team has a few one stars on it... you lost a few position battles.  If your team has three 5 stars in a given position... they killed it!