Football Recruiting: Relax

This is maybe a bit wonky, but I think does a decent job rebutting an argument you'll be reading a trillion times between now and September.

Tomorrow we panic, but today we relax. Let's slow down, and halt the senseless violence of another Ten Year War.

One Reason to Relax

Even in the bevvy of incoming 'New Ten Year War' articles, it's a chore to find ones that actually make a potentially compelling argument. This is one of the better attempts I've seen for predicting a coming B1G TW0 reckoning, in that it's actually bothering to try and advance a nuanced, statistical, hypothesis. I like that it tries to head off some of my past recruiting counter-arguments, that it compares to past data across the conference, and also appreciate that he's willing to caveat his argument.

OK, so first you'll have to scroll through some stuff that will make your eyes roll, have you mutter "c'mon.", and have you make sighing noises, but press onward, about a third of the way down he presents an interesting argument.

His question and hypothesis is this:

The natural question that arises is this: are Michigan and Ohio State dominating recruiting in the Midwest to a greater degree than they have in the past? If the Bucks and Wolverines are recruiting at the same level relative to the rest of the conference that they have over the past decade, then we have reason to be suspicious of claims that we are entering a period of hegemony for those two programs. On the other hand, if Michigan and Ohio State are monopolizing talent to a new degree, then we may be seeing something here.

His methodology is this:

To answer the question, we should take a look at the ratios of Michigan/Ohio State blue chip recruits (defined as recruits who got four- or five-star ratings; for ease of reference, we'll use Phil Steele's term: VHT [short for "Very Highly Touted]) to those of the rest of the conference over time. Rivals' database goes back to 2002, so we can use that as our measuring stick. We will include Nebraska for the entire time period even though they are a recent addition to the league, because that provides the best snapshot as to the state of the conference going forward.

His conclusion and data, here:

Here are the ratios:

Year	Michigan VHTs	Ohio State VHTs	Rest of B1G VHTs	Ratio
2013	17	14	23	1.34
2012	12	16	22	1.27
2011	6	10	27	0.59
2010	6	8	33	0.29
2009	14	17	41	0.75
2008	17	13	25	1.20
2007	7	12	39	0.48
2006	11	10	40	0.53
2005	11	10	36	0.58
2004	13	10	20	1.15
2003	13	7	24	0.83
2002	12	17	31	0.93

In short, the answer is that the disparity between the classes of Michigan and Ohio State on the one hand and the rest of the Big Ten on the other are wider than they have been in the last twelve years. This year's class and last year's have the two highest ratios for the period covered by the Rivals database. In fact, there were only two years in the prior 10 in which the ratio was greater than one, i.e. years like the last two in which Michigan and Ohio State had more combined VHTs than the rest of the conference combined.

The problem with his analysis, to not pull punches, is this: class sizes.

I'm going to start with a potentially contentious statement: if you want more four and five star players than normal, you want a lot of your scholarships to be open.

-Piece of data #1:

In 2007, our conference's own Urban Meyer signed a monstrous, monstrous, class of recruits while at Florida landing 20 four or five star players out of 27 total commits. The next year in 2008, he had another giant haul, with 16 out of 22 players being of four or five stars, just barely worse than his previous class in terms of both class average, and the percentage of his class that were fours or fives. But still, you know, still a few percentage points worse, and he slipped down the rankings. Now in 2009, 75% of his recruiting class was a four or five star, his best number yet, and his class average was 3.94, also his highest yet! That's a number one class right there! But it wasn't, it wasn't even in the top ten, coming in at number 11. His problem was, he only had 16 scholarships to work with, leaving him with 'just' 12 four or five star players. Alabama, the nation's number one class found room for 18 such players in their 27 player haul.

In short, when it comes to four and five star players, it's easier to go 3 out of 4 on a 28 man class than it is to go 100% on a 21 man class.

-Piece of data number #2

In 2005, Nebraska, then of the Big 12, signed a titanic 30 man class, that netted them 13 four or five star players. It was a top five recruiting class nationwide, that bested every program in the Big Ten. In the other 11 years between this class and the class of 2002, Nebraska broke double digits in four or five star players just twice, once with 11 such players with another huge 27 man class in 2007 and then with 10 such players in a class of only 19 in 2011. The closest they got to getting back into the top 5 nationally again was a 13th place finish, in 2007.

In short, you don't even have to have that high of a hit rate. if you can sign thirty guys you can get to double digits of four and five star players with a lot of strike outs or lower rated pick-ups along the way.

I could investigate this relationship deeper, but shit, I ain't got that kind of time. For now: Bigger classes mean more good recruits will want to come to your school as a general rule. I got some data and a persuasive argument, come at me bros. With that haphazard assertion under my belt,

LET US LOOK AT BIG TEN RECRUITING CLASS SIZES:

So super quick, you're going to catch that these last two classes, these two case studies where everyone is like, "OSU AND UM ARE LEAVING THE B1G IN THE DUST" happen to coincide simultaneously with Michigan and Ohio State's largest recruiting classes and the rest of the Big Ten's smallest recruiting classes. I think this is a key find.

Now, maybe Hoke and Meyer have attained such a level of power of the Big Ten's recruiting game that they have literally siphoned recruiting scholarships away from other teams through some sort of insidious life steal mechanic or...

A. Those two wealthy, traditionally successful, schools have each had about 50 scholarship slots open up in the past two years, something that happens, once, maybe twice over a dozen year span to the average Big Ten program and,

B. At the same time, the B1G's other programs have hit the dips of their scholarship numbers, (particularly the other 'contenders', Michigan State, Penn State, Wisconsin, Iowa, Northwestern and Nebraska have ALL average under 20 recruits a class the last two seasons), which leads to a situation where...

C. In addition to traditional advantages, for the past two recruiting classes OSU and UM have been able to leverage 'playing time' and 'early opportunity' and actually mean it, while in addition to their traditional disadvantages, the other contenders can't do that as easily, and when...

D. Four and five star players are limited commodities, and when a four or five star signs somewhere he is by definition not signing somewhere else (I see you "Committed, but still taking visits" kid) so when 27 scholarship Michigan takes one, they're out of 18 scholarship Michigan State's pool leading to...

E. Essentially a perfect combination of factors for this sort of recruiting blowout, even when it is super-duper likely we see large reversion to the recruiting mean as the class sizes flip around over the next year or two.

BUT, BUT, BUT-

Given this factor, I don't even think this is the most impressive stretch of UM-OSU recruiting in the last 12 years, contrary to that article's claim (and contrary to the claim of like, every other article on this subject).

Alright, here's where I have to be careful or I'm going to get LAUGHED AT ON THE INTERNET, FUCK. *takes deep breath*

The key is, you have to keep adjusting by this class size disparity between OSU and UM and the other ten schools. If OSU and UM's class sizes are bigger, their numbers get normalized downwards, if they're smaller, they get normalized upwards.

Ok, here we go:

Note: I'm pretty sure his ratio in 2010 is wrong. Rather try and guess at his dataset, I'd just say that I have the 'old ratio' value calculated at .42, and not .29. IDK.

By that adjusted ratio, after considering class sizes you see a couple really big drops and one really big leap over the course of 12 years. As you'd guess, the drops hit heavy in the last two seasons, while the big adjustment upwards is in 2003.

This leads to a new challenger in our 'biggest disparity in Big Ten recruiting' discussion. In 2003 and 2004 we see that OSU and UM we still able to pull in a high number of high value targets (43 players over the two years were VHT) but they were working on much tighter scholarship numbers (about 20 each between the two of them) against a much more competitive landscape as opponents each had on average over 22 scholarships of their own. If effect, in 2003 in particular, UM and OSU were basically able to match other teams in raw VHTs, like they do in their best years, with like, 5 fewer slots than their average competitor.

The article goes on to insinuate that other Big Ten school's need to spend more on recruiting resources to keep up (probably) and that they're being outworked (maybe). But you can run a correlation on the ratio between class sizes and the ratio this article ends up with and get an inverse correlation of .46 (meaning, that as the class size ratio moves down (UM and OSU's classes get bigger compared to the others), the VHT ratio moves up (UM and OSU get proportionally more of these players than the other schools). -.46 isn't a correlation to go all-in on, but it's large enough to you know, have to consider it seriously in addition to other explanations. What looks to some like killer recruiters operating B1G TW0 battle stations could just be structural happenstance, two men taking over the right programs at the right time.

These two schools efforts have been undoubtedly impressive, and I'm sure it will be a total pain to deal with some of the players they've brought in. I'm just saying, that with MSU's 'max! do not exceed or else!' cap seeming to be somewhere around 36-40 kids the past two years, if MSU had tried to sign 48-52 kids over that span, the NCAA would've burnt East Lansing to the ground with a space laser.

Now OSU and UM are each essentially facing a math equation that looks a little like 85 = X + Y + 50, whereas MSU's equation looks more like 85 = X + Y + 36. It is now, theoretically, going to be MSU's chance over the next two seasons to finally be able to sell the success of the past three years to above-average sized classes of recruits, and as a bonus, new scholarship crunches at OSU and UM as they find room for those 50 kids should mean less competition on the recruiting trail in that time period as well. It seems to be paying off so far in 2014, though we will all probably be on Harris watch til signing day.

Both programs will almost surely continue to recruit very well. Notre Dame will also undoubtedly be a thorn in the side of MSU's recruiting efforts. But these massive recruiting hauls by Michigan and Ohio State should chill considerably over the next two years, just by function of scholarship limits alone. It's also going to be really hard for them to keep up with top five classes, again, as a function of numbers, without serious attrition. And well, if they find a way to keep it up somehow, MSU should still be OK anyways. I think.
--

Football recruiting: Laugh

Football recruiting: Panic