Of course this applies to many, many schools, but Luke Winn's piece and statistics on decommits is imo, a fascinating study in an area we all hate: the top 100 decommitment.  There are plenty of stats, and Winn and his research crew do a decent job of drawing conclusions from much of the data.  I will not critique all of it, as I imagine many may try to dissect it down for our internal decommit review.

The one interesting stat that is very relevant, and one we should keep in mind, is the top 100 player data for multiple HS players.  The risks for decommit from colleges increases greatly for the multiple HS players.  

"Top-100 players who attended multiple high schools went on to decommit from colleges at nearly twice the rate of their peers."

Read more: http://sportsillustrated.cnn.com/2011/writers/luke_winn/08/01/commitment.trends/index.html#ixzz1UAmUo7aa