Sunday, March 29, 2009

Protect Ya Rim


Say you're watching a game, the Lakers and the Rockets. You see Kobe Bryant with the ball at the wing, with Shane Battier in front of him. Bryant makes a jab step, gets Battier off-balance, and then quickly begins a drive towards the baseline before Battier can recover. Bryant lowers his shoulder as he turns the corner and has his body in front of Battier's. It looks like he has his man beat, and is about to throw down a highlight dunk. And then, for no apparent reason, instead of attacking the rim, Bryant pulls up from about 12 feet out for a jumper, with Battier behind him. 

What happened? Why didn't he dunk? You decide to rewind the tape and watch the play again. This time, though, instead of ball-watching, you pay attention to the battle of the bigs down low. You notice that as soon as Bryant has a step on Battier, Yao Ming slides over in the paint and places his body between Bryant and the basket. The television announcers often talk about the importance of having a center who can block shots, or even "alter" shots, but in this case Yao isn't altering a shot, he's discouraging the shot attempt. 

Some of the most important moments in halfcourt defense are significant not because of what does happen (like a block) but because of what doesn't happen (a shot attempt from close to the basket). I was reminded of this recently when reading the stuff published over at basketballgeek. First, there was the study of the relationship between shot location, shooting percentages, and efficiency which, in addition to other things, showed that good defenses not only decrease their opponents' field goal percentage from inside, but also minimize their opponents' shot opportunities from inside. With that in mind, I thought it might be interesting to take a look at how centers around the league do in protecting the painted area.

The method I chose to go through this data is based on the way Football Outsiders looks at cornerbacks. Instead of just counting interceptions, as is traditionally done, they count the number of times a corner was targeted (ie, the number of times the quarterback threw to the receiver a corner was covering), and how successful those throws were when they happened -- accounting for incompletes and short completions as well as interceptions. That gives you an idea of how effective a corner is, filtering out corners who are taking a lot of risks to make big plays and rewarding those who are blanketing receivers so completely that they never even have a chance at an interception.

Assuming that the center is supposed to be the last line of defense at the rim, I've looked at the 2007-2008 regular season and counted the number of times opponents took shots at the rim while a center was on the court, how often those shots went in, and how often they were blocked. This all comes from the very useful play-by-play data made available at basketballgeek. The definition I'm using of "low paint" is the same as what Ryan used at that site, which is shots in the paint within 6 feet of the basket. Hence, some of these numbers will be slightly different from what 82games reports as "close" shots. Here's the spreadsheet:




You can also take this link to see the same spreadsheet:


You'll notice a couple of columns at the far right that might need a little extra explanation. Opponent quality matters -- it's a different proposition to stop Lebron James from scoring in the paint than it is to stop Tyronn Lue -- so I calculated how much above or below average each center held each of his opponents. For instance, individual opponents on average shot 6.8% below their season averages in the paint when Dikembe Mutumbo was in the game. Similarly, "% of Opponent shots in low paint compared to average" compares how many of an opponent's shots were taken in the low paint against each center versus that opponent's proportion of shots taken down there for the whole season.

(As a matter of curiousity, I calculated how each center in the league performed against Lebron James. For what it's worth, last year, Joel Pryzbilla had the most success against Lebron in the paint, holding him to 4-12 shooting. Shaquille O'Neal was at the other end of the spectrum, allowing Lebron to go off for 11-12. For reference, Lebron averaged 66% from the low paint on the season. Tim Duncan, Marcus Camby, and Jeff Foster had the most success at keeping Lebron out of the paint altogether, as the only centers holding him to taking less than 30% of his attempts in the paint). 

Notice in the numbers that centers who defend the paint well don't always register a lot of blocked shots (see: Jason Collins), while some centers get a lot of blocks but seem to do a better job of blocking and altering shots than they do of "discouraging" them (see, for instance, DJ Mbenga or Samuel Dalembert). 

These numbers aren't meant to be an end-all to measuring low-paint defense, but (I hope) they do tell a story, and I hope you'll find them useful. Keep in mind that the number of shots a center allows in the low paint will be related to the talent of his perimter defenders, coaching, and other factors. For a more complete story of how players affect opponent shot distribution and FG%, please see this post, which uses adjusted plus/minus methodology to measure these things. My goal here was to present some easily countable and understandable numbers that can be used in conjunction with other numbers to tell a complete story. 

No comments:

Post a Comment