Recruiting: What affect does Geographic Origin have on signing school?

Submitted by StephenRKass on December 22nd, 2012 at 1:41 PM

The thread below on the visit of recruit Jay-nard Bostwick has made me curious about something. In the comments, MGoUser "Don" says of Jay-nard,

He's the classic elite-level DL from the south that everybody up north in A2 convinces themselves we have a legit shot for, only to be disappointed again.

The assumption, which makes some sense, is that Michigan does a dreadful job of pulling any decent recruits out of SEC country.

Here's my question for the board:  I'd like to know what the correlation is between the geographic origin of a recruit and the school where they end up attending.

Common sense says that a recruit is typically going to stay close to home. After all, most of us like to stay in our comfort zone. Players would like to be somewhere that their family members can easily watch them, somewhere they can hop in a car and be home within a few hours. All you have to do is look at the Michigan Roster to see that more than 2 of every 5 players comes from the State of Michigan. If you add in Ohio and Illinois as neighboring states, you see that 2 out of every 3 guys on the roster are from those 3 states alone. (Interestingly, there is not a single recruit on our roster coming from the state of Indiana. If that isn't an indictment of HS football in Indiana, I don't know what is.)

There are other things at work in where a recruit ends up. Possible factors:

  • Did a recruit originally live in Michigan? For instance, David Dawson was in Texas for some personal family reason, but really was always a Michigan native.
  • Are there family ties? While Troy Woolfolk came from Texas, his father Butch played for Michigan.
  • Someone who just grew up as a fan of Michigan. These tend to mostly be in-state anyway, or to have lived in Michigan, but I suppose there are a few Michigan fans out there somewhere else in the world who have no georgraphic or family ties to Michigan.
  • A wealth of talent in a state. Florida is the best example of this, but Texas, Ohio, and Pennsylvania also are relevant. There may be so many good players in a state that it is impossible for the flagship state schools to make room for all of them.
  • Lack of a close Top 10 team. It is possible that historically, players from Utah, Kentucky, Colorado, Minnesota, and Arizona have looked further afield because they were dissatisfied with the prospects of team success at the big state university in their state.
  • A true opportunity at Michigan that doesn't exist elsewhere. This is extremely rare, but sometimes, there is a coach who gives a player an opportunity they won't get anywhere else. The classic example of this was Denard playing quarterback at Michigan for Rich Rodriguez. We never would have been blessed with Denard at Michigan except for RR. IIRC, no one in the SEC was inclined to let Denard QB.
  • Failure of the local school to properly assess talent and recruit it. While RR gets credit for Denard, he also gets blame in my book for missing big time on Johnathan Hankins, the DT leaving Ohio early for the NFL.
  • The allure of opportunity at another big time program, along with something wrong with the fit at Michigan. I wonder if RoJo fell in this category, spurning UM for USC. I know he was heavily recruited out of Muskegon, but maybe that was when Carr and staff were starting to deteriorate in recruiting prowess.

I may be asking for something impossible, but I'm curious if someone out there has analyzed past data for say, the top 50 teams, looking at this correlation or lack thereof. Or maybe someone like the Mathlete, or Ace, or some of the recruiting geeks, can look at this. I don't know statistics, but one thing I do know is that you can't correlate a single data point to the whole. However, you could probably see trends that give a sense of what has happened with recruits choosing to either stay home, or to go elsewhere.

Addendum:  The basketball team would seemingly have more diversity geographically, because of only 15 on the roster, compared to 113 for the football team. However, not only isn't there more diversity, there is far less. On the basketball roster, fully 87% (13 out of 15 listed on the current roster) come from Michigan or a neighboring state:

  • Michigan - Morgan, Horford, Person, Akunne
  • Ohio - Burke, LeVert
  • Illinois - Vogrich, Bartelstein, Bielfeldt
  • Indiana - Robinson, Albrecht, McGary
  • Ontario - Stauskas. Ok, neighboring country. But for all practical purposes, Ontario can be classed a neighboring state. Toronto is closer than Peoria.

Only Hardaway and McLimans come from somewhere else. (Florida and New York, respectively.)



December 22nd, 2012 at 1:52 PM ^

I'm not really sure what you're asking.

There are numerous determining factors in almost every kid's recruitment.

Most kids end up staying close to home.  Some don't.

Hardware Sushi

December 22nd, 2012 at 1:53 PM ^

Part of the Indiana problem, for us, is that a lot of the talent either goes to Catholic schools and feed to Notre Dame or go to Indianapolis' Harding or Lawrence North, which are sort of like a Cass Tech type program, schools that have some suspiciously anti Michigan coaching staffs. Granted, for all the help we get from Wilcher, it probably all evens out.

It's not like Indiana lacks for high end talent - they put out the #1 QB last year, #3 overall player this year, top 5 DT, etc etc etc. I think Sheldon Day is on our team if his coaches weren't OSU fans.


December 22nd, 2012 at 2:20 PM ^

Andy Staples did an article about this a few years ago (LINK - early 2009, but I think the trends in the data are fairly representative even now) - the Google Maps link in the right-hand  column used 2004-2008 data and can be researched using positions, conferences and schools as well. If you click on the recruit's name in the listing by whichever map you select, it will tell you how far their hometown is from the college they chose. 

What would be interesting to see is the model that this article talks about, a predictive model for which schools top recruits might choose. There was an article in a 2005 issue of "The Journal Of Sports Economics" that I would now like to find. One of the authors, quoted below:

"In general, DuMond says, the top recruits are looking for "a place that is in a BCS conference with a big stadium that is close enough that they can be seen by family and friends."


December 22nd, 2012 at 2:54 PM ^

Great article, good read, answered my questions. Yes, geography matters the most. This makes it harder for Michigan, but not impossible, because of the possibility of pulling recruits from neighboring states. We would clearly struggle more if the state of Illinois had top 25 teams. I also wonder what effect fallout at Penn State will have on future Michigan recruiting.

david from wyoming

December 22nd, 2012 at 2:03 PM ^

Even with a really high quality data set (which I don't think is possible), you are looking for such a relatively small effect and I don't see how someone with a good stats understanding is going to find anything. Recruiting is such a complex process that the effect of the location on a map is going to drowned out in all the noise and other processes going on.

To put it in story form... Say there is a high school that encourages recruits to stay home to go to the local college. Then, that high school coach is replaced with a coach that encourages students to go his alma mater. After a while, the college coach makes enough of an impression on the high school coach that the high school starts to send students to the local college. This is a very limited example, but it highlights what you would need to understand before you could look for larger trends (ie, southern high school coaches being more likely than northern high school coaches to tell recruits to stay close to home).

There is no way (without spending months on the data set that I don't think is possible to create anyway) to control for all the possible effects to the point where you can see what is really going on.




December 22nd, 2012 at 2:11 PM ^

What it seems like you're really looking for is how often do we actually land 4-5 star recruits outside of the midwest compared to how many we offer.  Why don't you just look at our offer sheets from the past 10 years of 4-5 star recruits from outside of the midwest and look at how many of them we actuall landed.  It may be useful to break it up by region as well since they will likely vary based on the coaching staff's background and history in the region.  For example Meyer can probably recruit well in FL & Hoke probably can in CA.  Can either in TX or PA, idk?

eamus_caeruli (not verified)

December 22nd, 2012 at 2:47 PM ^

I quite reading as soon as you singled out one arbitrary comment by Don. That is flat out wrong, and a horrible way to make an argument about recruiting or garner discussion. You can do better than that with statistics or aggregating opinions from people who follow the topic not by singling out one person on a MGOboard. Terrible!


December 22nd, 2012 at 3:59 PM ^

Since the arbitrary comment was mine, I'll elaborate. I've been following Michigan recruiting pretty closely since the early 1990s, and my memory is that in several of those years we've offered scholarships to 4- or 5-star defensive linemen from the deep south who supposedly are seriously considering us alongside their normal offers from LSU, Alabama, Georgia, FLA, etc etc. One that prominently comes to mind was Marquise Hill in 2001, who supposedly we were solidly in the running for, with some saying we were the favorite. Naturally Michigan recruiting fans from all over the country were giddy with anticipation of signing day, and then their hopes were dashed when Hill signed with LSU, and basically said he was never really considering Michigan seriously. Going back to 2002, we've signed just one DL from the deep south, Richard Ash, who was a 3-star according to Scout. The closest we've come are James McKinney from Kentucky, who was injured and left school after a year or two, and Ondre Pipkins from Missouri. We've gotten DLs from the west or from CA, but it's really hard to get elite-level DLs out of the south for some reason. That's not a knock on Michigan; even in this era of easy plane travel, most players want to play in relative proximity to their home. However, we have been able to pull a few elite-level players at other positions out of the south, so from my perspective it's been harder when it comes to DL.

When it comes to offering top-level kids in the midwest and Pennsylvania, we can compete with any program in the country, regardless of position. Only a fool would bet against Brady Hoke when he's working his own territory. Conversely, when we offer kids from the south or the far west who don't have some sort of pre-existing tie to UM or the state of Michigan, I'm always going to be a bit skeptical until signing day.

Autocracy Now

December 23rd, 2012 at 7:49 AM ^

That is an interesting trend you've identified with DL recruits over the years, but I can't really think of any compelling theory to explain it. I doubt DL players in the South fundamentally view Michigan differently than guys at other positions. I would guess this is a somewhat random effect or simply has to do with how tireless (or not) we were in pursuing them.

The generally low success rate for recruiting elite talent out of the South is what it is and I am sure the coaching staff takes that into consideration when they are deciding how to target their efforts.


December 23rd, 2012 at 9:34 PM ^

1. How far do recruits stray from their hometown?

This is probably not that hard to figure out, but you don't really want a "correlation" but just an average with standard deviation. Sort recruits by star ranking over a set period of time then calculate the distance to their college of choice. There will be skewing of the data towards closer proximity as some kids will move closer to the school at which they want to enroll. My guess is the standard deviation will be higher for lower star and higher star recruits (because the low star ones will go anywhere they can get in and the high star ones can go anywhere).

2. How do northern schools (or Michigan in particular) fare in regards to higher ranked recruits from southern states?

Again, this is pretty easy to answer by looking at the offer list and seeing if they got any of those recruits.

Ultimately, though, this information may not help very much in the decision-making process of any particular recruit as "past performance is not a predictor of future performance". But it would tell you, potentially, that the odds are slim.

Pulled P

December 22nd, 2012 at 4:19 PM ^

You'd have to consult the robot from Foundation. The robot(Chetter Hummin, I looked it up) has more data than anybody, not just recruiting-wise but every relevant gene, emotion, personality trait and sociological influence that is involved in selecting a BCS football scholarship. But even then, I think you'd have a hard time understanding his findings.

EDIT: OK, after an hour or so with no response, I'm officially disheartened. #Mgoblog's nerd/geek o'meter #Crestfallen


December 22nd, 2012 at 9:14 PM ^

My 15 yr. old son is a big Asimov fan (I'll take credit for that, thank you.)  But if you get a chance, please ask Chetter about the impact of weather and cultural similarity/difference as well as geography.  

I would argue A2 is culturally more similar to Portland, Oregon than it is to Kentucky despite being much further away.  But back to the weather... it's a hard sell pulling kids from down south when they visit in winter months. Then there's the honeys...


December 23rd, 2012 at 12:54 PM ^

Second Foundation==Ohio State. Cheats to win.

Foundation==B1G. Swallowing up territory across the galaxy, one Sayshell Sector at a time.

Galactic Empire==pre-BCS order and general pageantry, tradition, dignity etc. associated with intercollegiate athletics. Now has been stripped and sold for scrap to the highest bidder. Thanks, Trantor.

Autocracy Now

December 23rd, 2012 at 6:25 PM ^

Isaac Asimov himself could have done this calculation in minutes, and presented the findings in elegant prose.

I totally appreciate this post, as I am huge fan and Asimov wrote a letter to the children of Troy, MI (hometown) congratulating them on the magic of the new public library constructed there in '71, which was almost closed down this year.  His words: "It is a space ship that will take you to the farthest reaches of the Universe, a time machine that will take you to the far past and the far future, a teacher that knows more than any human being, a friend that will amuse you and console you---and most of all, a gateway, to a better and happier and more useful life."


December 22nd, 2012 at 3:24 PM ^

distance is obviously a factor as more of a tie-breaker if the other factors are equal.

Typically hear the #1 factor for many kids is preparation for NFL and #2 is liking/respect for the coaches.


December 22nd, 2012 at 4:04 PM ^

why would you think basketball would have a more geographically diverse class? the only thing a smaller sample size guarantees you is more variation from whatever the actual population mean is.


December 22nd, 2012 at 4:29 PM ^

SEC Boosters have an easier time giving a kid an offer they can't refuse... or at least a shorter drive for Cam Newton's father to submit an invoice.

Call me jaded when discussing 4* or 5* recruiting in SEC country. 

Zone Left

December 22nd, 2012 at 5:23 PM ^

Basically, school choice comes down to the following equation:

5.1 * Distance in parsecs + .22 * Parent preference + .74 * Relationship with coaches + 4.6 * Coed hotness + / - .03 * Academics + .49 * Weather + 1.3 * Locale = School score

Highest score wins. Every single player in the country uses this equation and files the results with the NCAA.


December 22nd, 2012 at 8:17 PM ^

Wait, you're staying that uniformz is not the #1 recruiting factor?!?!?!

Seriously, whenever someone brings up recruiting as a benefit of uniformz I wanna pimp slap them through my computer screen


December 23rd, 2012 at 11:51 AM ^

Uniforms are important in recruiting. In modern urban hip culture having the throwback uniform is cool. With Michigan changing uniforms every now nd then we are ensuring future recruiting success.


December 23rd, 2012 at 5:26 PM ^

I think we need to give these kids and their families a little more credit.

Sure, the uniforms may be attention grabbing. However, geographic location, school fit, coaches, nfl potential, playing time, friends, weather all are much more influential factors than uniforms.

As someone who works with 18 year olds, trust me, they do actually think about this decision carefully.

Steve in PA

December 23rd, 2012 at 6:08 PM ^

2 years ago I was working on models to predict where kids would sign.  I was putting kids into the spreadsheet when they got down to a final 3 and posting on here when I thought kids would wind up.


The largest factor (42% by memory) was the distance of the school from the player's home.  Next highest factor was the record of the recruiting coach and the last I had pinned down was the previous 4 year record of the school.

I got to about 80% accuracy, but didn't have time/energy to pursue it any longer.


edit: the formula was different for 4 and 5 stars with 4 stars being more willing to travel.

King Douche Ornery

December 23rd, 2012 at 10:55 PM ^

How would you factor in which team needs a particular recruit? For instance, when UM went spread with Rodriguez, the recruiting philosophy completely changed, and some kids who might have wanted to go to Michigan were not recruited by Michigan.