Similarity scores -- semi-complicated, multifaceted mathematical devices used to calculate one player's similarity to his peers -- can be truly useful instruments. Other times, they're just fun to look at. For example, see Barry Bonds's "similar players by age" list on Baseball Reference. Like all Bondsian stats, it's mind-boggling stuff. You can spend an hour staring at that page before your jaw manages to writhe itself up off the floor.
But this is not a post about Barry Bonds. (Thank goodness.) It is a post about former Richmond forward-slash-current NBA prospect Justin Harper, and in particular the hilarious strangeness of Harper's similarity scores. Harper is a 6-foot-9 forward with an uncanny mix of athleticism and shooting touch. ESPN Insider Chad Ford ranks Harper as the No. 30 draft prospect; he's a likely selection in the lower half of the first-round.
So: Tall, athletic, shoots the lights out ... is Harper the next Rashard Lewis? What about, say, Mike Dunleavy? Hedo Turkoglu? Bueller? Bueller?
Turns out, as Big Apple Buckets discovered, Harper's collegiate similarity scores include precisely none of these players. Frankly, they include a bunch of players you may never have thought of again were it not for the blog post in the first place. Harper's 10 most similar players are:
Kyle Weems, Missouri State, 2011 (1.71)
Daniel Kickert, Saint Mary’s, 2005 (2.17)
Levi Dyer, Illinois State, 2007 (2.42)
Nate Funk, Creighton, 2005 (2.50)
Craig Craft, Southern Miss, 2007 (2.56)
Dale Lamberth, Missouri State, 2007 (2.57)
Dale Lamberth, Missouri State, 2008 (2.60)
Dale Lamberth, Missouri State, 2006 (2.73)
Clint Cuffle, Evansville, 2003 (2.81)
Anthony James, Northern Iowa, 2011 (2.84)
Holy Missouri Valley Conference, Batman! Only one player on the list -- former Saint Mary's stalwart Daniel Kickert -- didn't play his collegiate hoops in the MVC. That's either some sort of coincidence, or something in Harper's game (size with shooting, limited touches but efficient usage, et al.) is a lasting point of order among the MVC's better scorers.
In actuality, as Big Apple Buckets explains, all of these similarities are pretty far off. There are a handful of guards in here; Harper's a forward. Really, the only thing these similarity scores tell us is that Harper is a really, really unique player. Perhaps a better picture would emerge if Harper spent a year in college as Richmond's only go-to scorer, but he never had to face that challenge.
Either way, it's an intriguing little bit of data. If I'm an NBA general manager, I've got my eye on Harper as a late first round steal. (Hey, Bulls: Lookin' at you.) But I'm also sure I've got as good a handle on Harper's unique skill set as possible before I assume he'll fit into my team's style. Something to keep an eye on, at least.