Dec 3, 2014; Brooklyn, NY, USA; Brooklyn Nets center Brook Lopez (11) reacts after making a shot late in the fourth quarter against the San Antonio Spurs at the Barclays Center. The Nets defeated the Spurs 95-93. Mandatory Credit: Adam Hunger-USA TODAY Sports
From a standpoint of understanding the game of basketball, I’m almost fully in the camp of learning as many things as possible about the game at an atomic level before moving on to the big picture stuff like, well, rankings.[1. And one number metrics. The internal Nylon Calculus Civil War over the desirability of focusing on APM/RAPM/RPM-style analysis rages on.] One problem with that approach is that you’re left with a whole bunch of little things to keep track of, and understanding that not everyone has read and re-read everything I’ve written on the topic it might be useful to collate some of the more interesting stuff[2. Your mileage may vary regarding exactly how interesting it is, but since we’re all basketball nerds here if you’ve made it to the footnotes, I’m going with it.] in one spot. Without further ado, here’s something of a best of 2014/year in review.
1. Rim Protection. Good measures of individual defensive ability and impact are few and far between. Steals, blocks and defensive rebounds typically tell us something about how well a player defends, but not everything, and they can be misleading as players who collect large numbers of these counting stats at the possible expense of other aspects of defense[3. Corey Brewer’s steals, JaVale McGee’s blocks, and Kevin Love’s defensive rebounds to name three notable culprits] might show up deceptively well in certain metrics. In part, the difficulty is that evaluating defense is in someway trying to prove a negative — how can you show the offense didn’t score because of a given defender or set of defenders’ contributions? The SportVU data starts to give at least some insight. While the general “shot defense” data available in the detailed shot logs or player dashboards is problematic for the purposes of assigning credit or blame[4. Probably a longer post for another time, but there is simply too much going on with the “all-court” closest defender data for it to be used as a meaningful apples-to-apples comparison of players.], the somewhat more clearly defined and detailed rim protection data allows for reasonable estimates of various players’ (especially big men) ability to defend shots at the rim. While the results of this information are largely inferential, in that it’s still not possible to look at most individual shot attempts and determine the degree to which the defender affected the make or miss, over a decent enough sample, it’s possible to start to differentiate between how various big men have performed in this key defensive function.
Though a detailed explanation is available here, this metric provides an estimate of how many more or fewer points opponents would have scored at the rim if the defender in question was replaced by a hypothetical “league average” rim defending big. While not perfect by any means,[5. shooting fouls, charges drawn and the effect on offensive rebounding are not considered, while team scheme and positional differences in the roles of power forwards and centers are observable as well.] it does provide some insight into which players are best performing the valuable task. Through the New Year, here are the top 15 bigs in terms of per game rim protection value:
It’s crucial to remember this value represents only one facet of defense. So even if these stats accurately represent Brook Lopez’s rim deterrent value[6. Lopez and Ibaka represent the two hardest cases for the metric to deal with, as Lopez contests a very large number of shots, but does so at a basically league average rate, while Ibaka contests a roughly average proportion of shots at the rim, but does so at an elite rate. Most right thinking people would prefer Ibaka, but that’s probably not a judgment based strictly on rim protection itself. Rather Ibaka’s all-court defensive presence combined with his uniqueness as an excellent rim defender from the power forward slot.], his lack of rebounding and poor mobility defending the pick-and-roll could easily give those gains back and make him a defensive liability.
2. A More Complete Picture of Offensive Involvement, (With a helping of a better measure of turnover propensity). Rather than go into a detailed soliloquy, I’ll just direct you here with the question of what would “Usage” look like if we allowed possessions to be measured cooperatively, rather than assigning all credit/blame to the shooter? True Usage, is a measure which captures a players’ overall offensive involvement, incorporating both shooting and playmaking. An additional bonus by looking at both shooting and passing, it’s possible to get a better sense for which player’s are most turnover prone than traditional Turnover % which tends to reward the chuckers over the playmakers. Here’s a comparison of some of the top PGs in terms of their True Usage (expressed as a percentage of plays in which they are involved by either shooting, setting up a scoring attempt or turning the ball over) broken into scoring and playmaking components. True TOV% represents turnovers as a percentage of plays involved[7. NBA average for all players is right around 10% while it’s around 9% for point guards only.]:
Speaking of point guards, the SportVU data also enables a more detailed look at point guards’ style of play to provide some content to back up such terms as “game manager, coach on the floor, combo guard rather than true point” and so on.
3. Playmaking is important. For reasons which should be obvious, the 2014 NBA Finals brought a certain fetish for ball movement into vogue. However, this doesn’t mean passing just for the sake of passing does much for an offense — there is little if any correlation between number of passes (either overall or per possession) and offensive efficiency. That said, effective passing, which might otherwise be known as playmaking, is important. In 2013-14, assisted shot attempts resulted in an Effective Field Goal Percentage 15% higher than unassisted shots. While the exact gap in productivity between assisted and unassisted shots varies by team, every team in the league is substantially better on assisted chances:
Note the Warriors having the 2nd highest percentage of shots assisted. They were 24th in this category last year and though their shooting has improved to some degree both with and without potential assists, their offensive improvement stems largely from their increased fluidity on the offensive end after finishing dead last in the NBA in passes per game in 2013/14.[8. Again, passing for it’s own sake isn’t the end all be all, but the Warriors have and even last year had a surfeit of talented playmakers who were largely wasted in Mark Jackson’s mid-90s era schemes.]
4. Corner Threes are Better because they are more open, not really because they are a foot or two closer.
5. Expected Effective Field Goal Percentage (XeFG%). Much more on this to come, but by combing through the detailed shot logs, some insight can be gained into which players and teams are getting open looks versus those that have either been good enough or lucky enough to hit more contested attempts, and vice versa on the defensive end[9. Luck is not a factor to be discounted over the relatively short span of a single game.]. This method of examining shooting confirms that Kobe really is taking a lot of terrible shots (5th lowest XeFG% among players with at least 200 attempts) and is also not making very many. Similarly, the struggles of Andrew Wiggins, recently maligned for the poor statistical start to his career, can be put in a little bit of context with the fact that he has the 18th most difficult mix of shots despite the knock on him coming into the league that he had yet to develop the ability to get his own shot. If one is an optimist (as I am), this would be evidence that his struggles are tied to the fact that Minnesota’s plague of injuries have resulted in the teenager being asked to do far more than he’s ready to do competently.
There is so much interesting stuff to dig into here such as this list of the players who have been the best shotmakers through the end of 2014. Not a bad list of shooters and finishers at all (with the inclusion of a few early season unsustainable supernovas, step forward Rasual Butler):
Please click through, peruse, enjoy and definitely get in touch either in comments or on twitter with questions, comments or suggestions for future areas of inquiry.