In this week's Sports Illustrated, I wrote a feature about the extreme basketball journey of two brothers from Durham, N.C.: Torian Graham, a top-100 recruit from the Class of 2012 who went to five high schools, decommitted from the same college twice and is now at a juco in Florida; and Tyree, a Class of 2008 recruit who fell off the map a bit while playing for four high schools, has since been at five colleges, and is looking for his sixth. I became fixated on the case of the Brothers Graham while doing a study on the commitment and transfer habits of the past seven classes of top-100 recruits* -- every high school (or prep school) they attended, every college commitment and decommitment they made, and every college they attended.
While the Grahams' nomadism demanded an investigation, the study's big-picture data revealed plenty about transfer culture. These are the 10 things you need to know from SI's tracking of the behavior of 700 elite players:
1. A one-high-school career is no longer the norm.
From the recruiting class of 2007 (the first one in the study), 28 percent of the players attended multiple high schools**. That number grew to 52 percent for the Class of 2013, mostly due to the rise in popularity of basketball-powerhouse prep schools such as Findlay (Henderson, Nev.), Huntington (W.V) and Montverde Academy (Fla.), which often groom top prospects for 1-2 years before college.
The chart at right shows the percentage of multiple-high-school players from each class in the study.
2. Those who transfer in high school, transfer more often in college.
From the 2007-11 classes (2012 and '13 haven't had enough time to transfer yet), 26.8 percent of the single-high-school players went on to transfer in college, while 37.8 percent of the multiple-high-school players did.
Splitting the multiple-HS group into subcategories^ shows even stronger correlations: 33.8 percent of the two-high-school prospects transferred college, while 38.0 percent of the three-high-schoolers did, and 60.9 percent of the four-or-more high-schoolers did.
3. A third of top-100 players don't finish college at the same place they first enrolled.
To be more specific: Of players in the study who spent at least two seasons in college (thus eliminating the one-and-doners), 34.3 percent didn't end up where they started.
The chart at right breaks it down by class and shows the percentage to be rising each year. Expect the 2011 rate to keep growing, since that class still has two years left to transfer.
4. That 34.3 percent transfer rate shouldn't be considered abnormal ...
... because a recent study by the National Student Clearinghouse Research Center showed that 32.6 percent of all full-time college students transfer. Are college students as a whole transferring too much for their own good? Maybe. But are basketball players transferring far more than non-athletes? That's a definite no.
5. College decommitments have been on the decline.
Are prospects becoming less fickle? In the class of 2013, just 11.8 percent of the players made multiple college commitments, and that rate was 12.4 percent in 2012. Those are both well below the early years of the study, when it was common for 15-19 percent of a class to decommit at some point in the recruiting process.
6. The calendar could be playing a role in the decline in decommitments.
Using the first (not final) commitment dates of each player, I plotted out how each class went from 0 to 100 commitments^^ -- and the chart below shows that the class of 2013 had the most recruits wait until their senior year to make an initial pledge. Is it a coincidence that 2013 also had the lowest rate of decommitment, at 11.8 percent?
7. Extra-early commitments rarely work out.
I plan on writing more about this in another story, but the track record of players who commit to a school three or more years in advance is not pretty. There were 29 such players in the study, and 15 of them went on to decommit. Of the 19 who went on to play at least two seasons of college, nine transferred.
8. Coaching-change culture should take most of the blame.
We shouldn't be surprised that half of the extra-early committers go on to decommit. Why? During the course of this study (from 2007-on), the average annual rate of turnover in the Division I coaching ranks was 13.5 percent, which outpaces the average annual transfer rate for men's basketball players during the same stretch (10.8, according to the NCAA).
So, let's say a player commits to a school during his freshman year of high school, and goes on to play four years of college ball. In that eight-year span, it's likely that there will be more coaching changes than there are teams in D-I. (Maybe we should be calling that an epidemic.)
9. The record for earliest top 100 commitment belongs to ...
Ryan Boatright, a point guard from the Class of 2011 who committed to USC's Tim Floyd after attending a camp on June 19, 2007, in the summer after Boatright's eighth-grade year. Boatright would later decommit from USC (which had NCAA trouble), decommit from West Virginia (which recruited another point guard after he made his pledge), and go to UConn (where he could have left after the school's APR disaster, but did not).
10. The Brothers Graham's combined resume of high school transfers, decommitments and college transfers cannot be topped, but there were a few players who beat them in individual categories.
The record for most high school stops by a top-100 player is six -- a tie between Michael Beasley (2007), LaRon Dendy (2007), Nate Miles (2008) and Norvel Pelle (a 2011 recruit who has yet to arrive in college).
And finally, one outlier worth monitoring is Jelan Kendrick (2010), who's already made college stops at Memphis, Ole Miss, Indian Hills (juco) and UNLV. He's veering in on Tyree Graham territory, and still has two years of eligibility remaining.
Asterisks and miscellany
* The study uses top 100s from the Recruiting Services Consensus Index, which combines the rankings of reputable recruiting websites. Some players required reclassification from their final spots in the RSCI -- for example, someone who appeared in the Class of 2008, but made a late decision to take a prep year and thus needed to be moved to the Class of 2009. So while SI's database includes the same names from the RSCI, some of them have been moved ahead a class for accuracy.
** A couple of rules on counting high schools:
When tabulating high school transfers, SI did not count European or African high schools or academies attended prior to arriving in the U.S. -- the reason being that in most cases it was necessary for these international players to spend time in a U.S. high school to get qualified for the NCAA.
On the occasions that players made multiple stops at one school, we counted each stint -- so, someone who transferred from a high school, to a prep school, and then back to their original high school counted as a three-HS player.
^ The sample sizes, after subcategorization, go from 291 (one HS) to 136 (two) to 50 (three) to 23 (four-plus). The final sample is admittedly small, but it's difficult to dismiss that more than half of those players made multiple college stops.
^^ Credit goes to Drew Cannon, now of the Boston Celtics, for the idea for this chart. He did some excellent work mapping commitment dates for Basketball Prospectus last summer.
Additional research on this study was provided by Chris Johnson.