The Reclamation of Connor Atchley's Reputation

As many of you probably know by now, Michael Lewis wrote yet another epic article for the New York TImes Magazine this weekend.  The author of Moneyball, Liar's Poker, The Blind Side, and the NYT Magazine article on Mike Leach that started the whole pirate thing in the first place, Lewis has a gift for seeking out instances in which someone in sports is exploiting inefficiencies in old ways of doing things and then translating what those people are doing for the layperson.  The article published this weekend was about Shane Battier, Darryl Morey (the GM of the Houston Rockets), and the coming SABRmetric explosion in basketball statistics.  You should absolutely read it if you haven't already.

The thesis is that, just as in baseball before Bill James' theories gained traction in front offices, traditional cumulative statistics do a poor job at evaluating the contributions individual players make to the team, the Houston Rockets front office is at the forefront of new efficiency statistics that make an attempt to more accurately measure what a player does for a team, Shane Battier is one of the most efficient players the Rockets have ever seen and makes his team a lot better simply by being on the court, but Shane Battier puts up terrible traditional statistics so he is undervalued by other teams and thus is a steal for the Rockets.  Also made explicit in the article is that Shane Battier is, at best, a marginal NBA athlete with poor ball handling skills and a jumpshot that requires him to be open to make.  Sound like Connor Atchley to you?

It sounds sort of crazy, I know, but Connor Atchley is one of the most efficient players on the Texas team this year and last year was one of the most efficient players in the entire NCAA.  Note, of course, that "efficient" doens't mean "best."  But a player who is efficient without putting up great traditional numbers is often overlooked and maligned for not doing anything on the court when in fact he is hard at work making the team better in difficult-to-measure ways. I believe Connor is in this category.  Hear me out after the jump.

The comments that have been floating around BON about Connor this year are astoundingly short-sighted.  He's not shooting the ball well this year, that much is certain, and his points, rebounds and minutes are all down.  But sometimes, it's not traditional "box score" statistics like points, rebounds and assists that define how good a player is.  This is especially profound in basketball because, as Darryl Morey points out to Michael Lewis in the article, individual basketball statistics can be accumulated to the detriment of the team (whereas baseball is a sport in which individual success more or less translates into team success).  Luckily, over the last few years, Ken Pomeroy has been leading the charge of creating/implementing efficiency statistics and applying them to college basketball.

If you've been paying attention to BON's basketball coverage the last few years, you should know that we favor KenPom's rate statistics in terms of team analytics simply because while traditional statistics often belie the true nature of the game, reducing it to its most base formulation of "who can make more shots, get more rebounds, and dish more assists than the other guy," these KenPom statistics make an attempt to get deeper into the true nature of the game--the movements, the flash-outs, and in short, the little things that make it possible for certain players to score a lot of points, get a lot of rebounds or dish out a lot assists.  In basketball, it's a lot easier to develop team efficiency statistics than individual ones, simply because what a team does is a whole entity that is affected by only one other variable, the other team.  However, what an individual does is only part of that whole entity, meaning that there's an added layer of variables to account for (i.e. how that player operates within that team).  As I said earlier, baseball doesn't have this problem because it is an individual sport in almost every capacity other than defense.  It's difficult in basketball, however, to isolate and examine what, exactly, an individual is doing to help a basketball team, especially when he's not putting up traditional "box score" statistics.

But KenPom and some others have developed some metrics for doing just that.  And by those metrics, Connor Atchley is an excellent basketball player even when his shots aren't falling.  And when they are, like last year, he is elite in terms of efficiency.  Let's look at some of those statistics.  I explain some of the stats below; for explanations of others please see this and/or this.

In sum, what this tells us is that even though the offense doesn't really go through Connor Atchley (only taking 14.8% of the team's shots that occur while he's on the court this year and 16.5% last year), he has a strongly positive effect on the offense while he is on the court, as is Offensive Rating (ORtg) ranks him quite highly this year and was astronomical last year.  ORtg is the efficiency statistic meant to measure what effect a certain player has on the offense.  It's effectively a modified +/- statistic that measures how well an offense does with the player on the floor versus with him off of it, and then figures out how much of that plus or minus is attributable to that individual player.  This year, Atchley is behind only two players on the team, Dexter Pittman and AJ Abrams (for the record, Clint Chapman is last on the team with an abysmal 67.1 ORtg).  Last year, Atchley was behind only 11 other players in the entire NCAA, among those who played at least 40% of their team's minutes.

How did a guy who averaged 9.5 ppg, 5.3 rbg and .8 apg in 2008 end up as the 12th most efficient offensive player in the entire NCAA?  Numerous ways.  His effective field goal percentage and true shooting percentage were off the charts, he never turned the ball over, he got a fair amount of offensive rebounds, and he made everyone around him better.  That is, the team scored more points when he was in the game than when he wasn't, it got more rebounds when he was in the game as opposed to when he wasn't, etc.  In effect, while other players were running the show, Connor was behind the scenes making things happen.  It's the little things that don't show up in the box score.  

And that's just on offense.  KenPom doesn't have an individual defensive efficiency statistic yet, but I have to imagine that Connor would be highly rated by that one as well, especially considering that in 2008 he blocked a full 8.5% of all shots that were taken while he was on the floor, good for 49th in the entire NCAA (and it's only down to 7.9% this year, a #64 ranking).

Note also the offensive and defensive rebounding percentages (OR% and DR%, respectively) for Atchley.  These percentages show the numbers of rebounds he pulled down out of the number of possible rebounds that he could have pulled down in his minutes on the floor.  KenPom says that 10% is a good OR% while 20% is a good DR% and that offensive rebounds are generally regarded as being more the result of individual effort than defensive rebounds.  Note that Connor is a lot closer to the 10% threshold for "good" OR% than he is to the 20% threshold for "good" DR% (at least in 2008), and that while his OR% has gone down in 2009, his DR% has stayed the same.  This tells me that Connor gets his own offensive rebounds fairly well, but his lack of overall defensive rebounds isn't indicative of his lack of worth on glass because even when he's clearly not playing as well as he did last year, his defensive boards are still more or less the same.  And while I don't have the specific stats to show it, I would be willing to bet that a large part of his high ORtg is attributable to the fact that the team gets more defensive rebounds with him in the game than without.  Boxing out your man and allowing Damion James to jump high for the rebound is probably more effective than trying to outjump James for the box score statistic.

And that's the crux of all of this.  Connor does the little things that make teams better that they would be without him, even though those contributions don't show up in the box score.  Last year, we had a love affair with "Good Connor" because in certain games, his contributions did begin to show up in the box score, mostly because of his ridiculous shooting percentages.  This year, he's not making any shots and thus many of you have deemed him "useless."  That's absolutely not true.  There's more to basketball than scoring points, getting rebounds, and handing out assists.  There's the rotation on defense that Connor is so good at, flashing an arm out in front of a cutting offensive player to deny him the ball, weakside help defense, cutting off the high pick and roll without getting too far out of position, and always keeping his blocked shots in bounds.  On offense, it's the subtle moving screen he sets at the top of the key and the flash out for the 3 that defenses still respect, boxing out the weak-side to allow someone else to go get a rebound, keeping the spacing in tact in a way that our other big men just can't or don't.  He makes the team better when he's on the floor, even if he's not hitting shots.  And if he is, look out world, you've got an elite player on your team.

I've been noticing this about Connor for a long time, but one thing hit it home yesterday.  An Aggie drove to the basket and Connor came over to block the shot from behind.  He didn't block it, but ended up under the basket when the missed layup came down right towards an A&M player.  Connor could have tried to grab the rebound, but he was out of position and would have easily lost the ball to the Aggie for an easy putback.  Instead, he saw Abrams at the top of the key and patted the ball back out to him.  If AJ hadn't lost the ball in a scrum with another Aggie, it would have been a fast break, a 4-point turnaround.  Some got mad at Connor for seemingly not hustling to get the ball, but he actually made a good play that helped the team and didn't show up as a rebound in his stats.

Now, I'm not saying that Connor Atchley is necessarily a great basketball player, or that he's anywhere near as good as Shane Battier, even in the realm of these efficiency metrics.  He's often out of position rebounding and does not have the strength to muscle bigger players.  He can't create his own shot, and this year has had plenty of trouble making the ones that were created for him.  But he is a very good basketball player and, if he's making his shots, he's a potentially great college basketball player.  But because he doesn't put up a lot of traditional box score stats, he is undervalued, both by fans and potentially by his own team.  And this will afford the possibility for a smart NBA team to pick him up as an undrafted free agent and start to mold his game here and there, while allowing him to do what he does best--make his basketball team better.  In the meantime, he's doing just that for this Texas basketball team, whether you realize it or not.