Average Recruiting Class of Playoff Teams

Submitted by MGoStrength on December 6th, 2017 at 10:54 AM

TVH posted this on Twitter earlier.  I have to assume he's using the ESPN rankings.  Here are the average rank of the recruiting classes for the last 4 years of each playoff team.


Okla - 14.75

Clemson - 8.5

Georgia - 6.75

Bama - 1.25

Average of all 4 Teams - 7.81

(UM - 17.5)

(OSU - 5)


The bad news is UM is still well behind the current playoff teams average.  The good news is that Harbaugh's only two full classes to date average #6.  2015 and 2014 really dragged down our average, which means the vast majority of our talent is in the freshman and sophomore classes. 


Conclusions:  Going by recruiting alone, if UM continues to recruit at the pace Harbaugh has in his only two full recruiting classes, one could guess that's good enough to get into the playoffs.  The bad news is that might not be good enough to win the conference as OSU is still not only ahead of us, but also ahead of the average playoff team at #5.



December 6th, 2017 at 11:02 AM ^

The most talented teams usually start with players possessing a significant amount of talent. When the fully mechanized Harbaugh program has been fired up, (all his recruits) the teams will be in good shape.


December 6th, 2017 at 12:58 PM ^

Didn't Harbaugh just have a record setting number guys drafted , all of which he inherited ?

I don't follow the "cupboard was bare" argument. I'm willing to concede she needs "his" guys for "his" program style though.

I just don't understand why we are going into year 4 without a proven QB and yet again talking about pulling another transfer QB.

How about sticking with a QB and focusing on an offensive line to protect him while he develops ? I really thought Speight would be the guy to lead us to championships. Pretty rotten that a cheap shot robbed him of that opportunity.


December 6th, 2017 at 1:19 PM ^

He was hired late in the recruiting year so his first class rushed.  I'd be happy with Peters, McCaffrey, Milton and Doyle next year. I saw enough of Peters to say he has the tools.  He had a good start to his career. 

I think Peters was going to win the job from Speight and Harbaugh's experienced QB decided to transfer. I don't think coach was looking for a 5th year grad transfer, but has an opportunity to make the team better with a transfer from a school going on probation.  Not sure why you wouldn't take that kid. 


December 6th, 2017 at 2:03 PM ^

The qb cupboard was worse than bare. Peters is harbaugh's first proper recruit, and he'll be a redshirt sophomore. He is servicable. If Patterson can come in and do what cam Newton did at Auburn, give us one solid year while peters gets better, I don't see what's not to like.


December 6th, 2017 at 2:55 PM ^

I'd be happy if he could come in and do what Ruddock did. Doesn't need to win games by himself but needs to be good enough not to lose games. That's about the level I saw Peters at when he was playing against the better competition.

Very curious to see what Dylan can do, hope he doesn't miss a chance to showcase it with an upper classman transferring in


December 6th, 2017 at 11:04 AM ^

But the sad fact remains if we had been able to get just average QB play in our four losses we definitely beat MSU & OSU and Wisconsin would be a toss-up.  Still think we lose at Penn State but again, with just AVERAGE QB play we're 10-2 or 11-1 right now and feeling a whole lot better about ourselves.

Sorry but as a Cleveland Browns fan I can tell you from brutal first hand experience that when you dont have a QB you dont have a football team that can win.  And for all our OL & WR struggles this year the fact remains it was one position and one position alone that cost us two games and possible three.

II personally could care less where we're ranked.  Solve the QB problem and everything else looks a lot better.  


December 6th, 2017 at 12:24 PM ^

Practically everybody I work with and friends with here are either OSU alums or if not, they're big fans of the Buckeyes.  And just about all of them know I went to Michigan and that I went to the game a couple of weeks ago.  What's interesting to me is that to a man they ALL remarked about how terrible our QB was this year and last and they are very surprised Harbaugh hasnt "fixed that'.  They are more surprised than anything that Harbaugh hasnt fielded a QB against them that they fear (Rudock did  jack-squat in that game) and laugh at our struggles at that postiion.

I cant even really bitch about the officiating without looking like an idiot cause they're right. 


December 6th, 2017 at 6:30 PM ^

the past two years.  Just say

1) Speight was injured last year, and despite that you were outplayed and gifted a game by officiating

2) O'korn was the 3rd option and we were 1 or 2 plays / calls away from beating them.


Let them pound their chests, as they have owned this rivalry the past 15 years.  But deep down they can feel it, they know it, they can smell it.  The reckoning is coming, the talent gap has closed, and Harbaugh will have a QB in place that can beat them down soon enough.


December 7th, 2017 at 2:02 AM ^

in that game, actually.

He was 19/32 for 263 yards at 8.2 ypa with a TD and no INTs.  First drive he marched us to the OSU 36 before we had Jabrill throw an incomplete pass (smh at that late season Jabrill package) then punted on 4 and 5 from their 36...for a touchback.  Again, smh.  Harbaugh needs to get more aggressive in The Game.  That's the one area in which Hoke was better: more aggressively going for it on fourth, especially in games that we needed it.

He got two first downs on the next drive and then led a 72 yard FG drive (ending at the 6) and a 92 yard TD drive to be down only 14-10 at halftime (amazing we only had 4 drives in the first half).

The wheels came off for the defense in the second half, but Rudock still led a 57 yard FG in one of his two drives in the half before getting hurt.

The defense was awful that game and the running game abysmal (21 carries for 65 yards from the RBs), but Rudock singlehandedly kept us in it for a half.


December 6th, 2017 at 11:06 AM ^

Gonna guess those rankings don't account for recruits who have since left the team.

We ended up with what amounts to half a recruiting class over a two year period when you factor in attrition.


December 6th, 2017 at 11:06 AM ^

I am pretty sure a mathematical average with a sample of 4 is a potentially severely deceptive statistic that tells us very little about the underlying variance in the metrics, unless (as in the case of Alabama) there is no variance. For instance, Clemson at 8.5 could mean that it had classes ranked 32, 1, 1, 1 over the last 4 years, that their senior starters all came from a decidedly mediocre class and yet they made it to the CFP anyway. Oklahoma, on the other hand, could have classes ranked 1, 30, 8, and 20 over the same period, meaning that their senior starters all came from an incredible class, while their junior contributors all sucked on paper at least. 

In sum, statistics are dumb and easily abused.






December 6th, 2017 at 3:31 PM ^

It's mostly due to the dilemma of qualitative versus quantitative.

For instance think of the ranking of a H.S Quarterback (247, ESPN, Scout, etc). 

Under what criteria are they ranking them?

"How good they will be in college"? (Terelle Pryor and Denard would be high here)

"How likely they are to go to the NFL"? (High here, but not as QBs!)

"How good they are compared to a H.S. QB"? (who the hell would know that?)

"How much potential they have"? (Very much depends on the coach)

"How good their physical tools are at the position"? (Rumored to be Saban's primary criteria)

Often once you start to quantify something, as you go deeper and deeper (keeping in the foobtall ranking domain of course), you often end up transitioning to a largely quantitative description (which isn't a bad thing necessarily).

This is also essentially a great description what a top coach/staff is doing, they're internally (and subsconsciously) quantifying things (or analogously) and forming a qualitative "feel' for why they want/don't want certain recruits (why a #16 ranked at "__"-position would be more valuable to their own specific team over a #8 ranked at that same position).

Also once an obscure player commits, it's almost guaranteed that they raise in the rankings and sometimes even gain stars. That should be the first tell, of the limitations of the "experts" at ESPN and 247 etc. The "experts" suffer no consequences from their poor prognostications, while the coaches paycheck depends on it. Skin in the game matters. (Hence why Hilary and Trump--and all the rest of them for that matter--really DGAF about you or your "life". Just recently that guy McCain voted to take away single-payer healthcare despite HIMSELF HAVING single-payer healthcare which HE VOTED for, FOR HIMSELF, lol /rant---sorry!)

Of course you can see here that ranking can get tricky, and when you really try to hone in and zoom in, you're often splitting hairs trying to parse a #2 QB-PP versus a #8 QB-PP.

Essentially this is the fundamental error people make when presenting statistics corresponding to the more "human" endeavors: they're reporting a mere sliver of reality, from a very specific viewpoint, using everyday common language (problematic), even if they mean well, and attempting to project/predict it into the future (see: The Monte Carlo Paradox). So it's very difficult, and this is not even to mentioning people acting in bad faith/perverse incentives (read: journalism)

I think there is a general statistic that probability drops immensely once you leave the Top-20 Average-recruiting class. 

It reminds of one of my favorite quotes from a pretty great statistician, "All models are wrong, but some are useful."


December 6th, 2017 at 12:07 PM ^

would also be a better number too, as I am guessing OSU could be the number two in talent according to this.  

Seeing as you should probably way seniors and juniors higher than freshman and sophmores, you'd probably have to add in a weighted multiplier (like freshmen * 1.1 and seniors * 0.9).  This would be arbitrary, which is why statistics can lie.And as others have pointed out, you need to recalculculate after attrition, so that could have a huge affect. 

Fun information to have anyways. 


December 6th, 2017 at 11:10 AM ^

Part of me wonders what the ten-year historic look for recruiting class rankings for each of these might be since all of these programs have been through a certain amount of change in the last decade, just to see where the dips and spikes are. Well, I think we're well aware of ours and maybe OSU's, but for the others. 


December 6th, 2017 at 11:32 AM ^

I think they do it more through aggressive roster management* (employing as many medical redshirts as the rest of the SEC combined, pressuring transfers when guys can't crack the starting lineup, etc.) a history of guys moving onto the NFL and recent success (guys want to play on winners) than paying players. 

*money quote:

FBS teams are limited to 85 scholarships at a time, and managing those is a years-long process. In the past few seasons, Bama’s brought on larger classes: 29, 23, 24, 24, and 26.



December 6th, 2017 at 12:06 PM ^

no...they pay players.


So does OSU to no one's surprise.

Roy SquareTree

December 6th, 2017 at 11:26 AM ^

I am not a member so it should work. I think they use their own composite rankings of recruiting and include tranfers and things like that to come up with a score for each team. 

Edit: Can't figure out why the link doesn't work. It's under the NCAAF tab on the homepage of 247


December 6th, 2017 at 11:38 AM ^

This may have been done at some point, but I think something that gets missed here is figuring in attrition to these class rankings.  When you are a total NFL factory and you have a ton of guys leaving at 3 or 4 years and guys transferring in even less time if they don't get PT, your classes have a "shorter life" and need to be replenished faster - meaning more top ranked classes are needed.  It becomes almost required to continue recruiting at that pace once you start - otherwise you end up with huge holes with players leaving early.

I really think this obscures the connection between recruiting rankings and results .  Teams  who recruit "well" but not off the charts (say in the 10-20 range 4 year average) probably get more guys to stay 4 or 5 years and tend to compete well with the teams that are recruiting top 5.  By that I mean teams full of RS Senior/RS Junior/True Senior 3 and 4 star guys can and do compete well with teams full of second/third year 4 and 5 star guys. 

A whole different point to consider as well that recruiting info is monetized in the form of subscriptions to recruiting websites, and you have an obvious reason for bias towards bigger programs.  Giving more (and more importantly, good) news to big fan bases who have more fans willing to pay, and you make more money.  I don't think this makes a huge impact on the rankings, but clearly lower level players that commit to big programs are far more likely to get a second look for a ratings bump than a similarly ranked player that commits to a G5 program or a less successful P5 program. 

There also seems to be a bias that recruiting sites don't like to admit they are wrong about a rating they have maintained for a long time.  As a result, it seems like players who fall late don't fall far enough, and players that rise late don't rise enough.  Since I have a full time job I'm not going to spend a month trying to prove that theory, but you can find a lot of examples of this just in players M has been involved with.  Perhaps it's anecdotal, perhaps not.