Making Connections: OREB% to ORTG
In the days of yesteryear, basketball was simple. Awkward, bespectacled white guys would run, occasionally dribble, shoot without jumping, and slam dunks were witchcraft. It was a simpler time. The great rebounders were measured solely by the number of rebounds they scooped up. But as the game evolved, so did our analysis. First we split rebounds into the “offensive” and “defensive” persuasions. Then we started to apply some critical thinking; who is a better rebounder – the guy who just grabs the most rebounds, or the guy that grabs the highest percentage of missed shots? Dean Oliver broke basketball down into his “Four Factors” to help us better understand how offenses (and defenses) perform.
Is this the next step?
Of all the functions necessary for taking the next step, this research, I suppose, is lifting our heel a few millimeters off the ground; the first step towards the next step, in a manner of speaking.
We currently evaluate the best offensive rebounding team by calculating which team grabs the highest percentage of their own missed shots. This is pretty intuitive, but there’s a subtle problem with it – it treats the offensive rebound as the end point. It doesn’t say anything about how those offensive rebounds influence overall offensive output. And isn’t that the point? Yes, offensive rebounds are nice, but isn’t what really matters how well teams are able to translate those offensive rebounds into more points? After all, the point of offense is to score, right?
So we started tinkering. What would it look like if we started charting Offensive Rebound Rate and Offensive Rating, for every team, for every game, and seeing how the two correlate? There have already been several studies done on how much each of the Four Factors explain overall performance (like this one done by Evan Zamir back in 2010), but they’ve only looked at the league as a whole. We were curious as to how these relationships change on a team-by-team basis.
For example – let’s say there is a team that does a good job of rebounding their own misses, but in general, just aren’t very good at shooting in the first place (like, say Memphis – 2nd in ORB%, 28th in eFG%). What is more likely – that their ORB% correlates strongly with their Offensive Rating (because they need those 2nd-shot opportunities to sustain their offense), or that it doesn’t correlate at all (because regardless of how many offensive rebounds they get, they still aren’t any good at actually putting the ball in the basket)? Or, conversely, a team that doesn’t rebound well but shoots the ball exceptionally (cough, San Antonio, cough)? These are the types of questions we wanted to answer.
So, Ian went to work on the hard part – scraping data from box scores for every game of the entire 2013 season. Then we charted every game to give us X-Y graphs, like you see below.
Then we put all the information together on one (hopefully) understandable graph.
As it turns out, it seems like poor offensive teams see a higher correlation between Offensive Rebound Rate and Offensive Rating than good offensive teams. Obviously, there are a few exceptions (Brooklyn and Oklahoma City have high correlations, Memphis and Detroit have low ones), but for the most part, the left side of the chart is decidedly red (low Offensive Rating) and the right side of the chart is decidedly blue (high Offensive Rating).
So what do we take away from this? It’s possible that defenses would be better served crashing the defensive glass against poor offensive teams, and focusing energy elsewhere against good offensive teams. Or perhaps poor offensive teams should sell out more to rebound a higher number of their missed shots.
There is certainly more research to be done on the subject, and we’re going to be expanding this research into other areas as well – correlating the four factors against offensive and defensive efficiencies, pace, and even each other. So enjoy the data, and keep your eye out for more.