A Quick Look at Offensive Rebounding with SportVU


Nov 9, 2014; Portland, OR, USA; Denver Nuggets forward Kenneth Faried (35) grabs a rebound over Portland Trail Blazers center Robin Lopez (42) at Moda Center at the Rose Quarter. Mandatory Credit: Jaime Valdez-USA TODAY Sports

In the first week of the season, several new player tracking stats were released on NBA.com. Among them were logs for all shots and rebounds. Since the SportVU cameras were installed in all arenas at the beginning of last season, when these new stats were released they included data from last season. The shot logs give fans a chance to look at things that have never been available to the public such as time remaining on the shot clock, distance to the nearest defender, time of touch and number of dribbles before taking a shot. The rebound logs include the rebound distance, whether the rebound was contested, the shot distance and who took the shot. I decided take these two data sets from last season and link them together to see which shot corresponds to each rebound to dig deeper into where offensive rebounds come from. I was more interested in rebounds off of jump shots than shots near the basket so I only looked at shots taken from 15+ feet from the basket.

The first think I looked at was the offensive rebounding percentage based on the time left on the shot clock. The results were pretty interesting and can be seen in the graph below.

There appear to be a lot of offensive rebounds off of shots from early in the shot clock. Presumably these are shots in transition or off of offensive rebounds where the defense probably is a bit out of position making it more likely for an offensive rebound. More interesting to me is the upward trend in the last 10 seconds of the shot clock. I would have thought that when you are shooting at the end of the shot clock it’s more likely that the defense has done a good job defending and is in good position to box out and get the rebound but at first glance this doesn’t appear to be the case.

The other thing I looked at was the offensive rebounding percentage based on the distance between the closest defender and the shooter. These results are more along the lines of what I expected and are in the graph below.

One thing to note when looking at this graph is once the distance between the shooter and nearest defender gets over 10 feet the sample size starts to get pretty small. As you can see when the shooter is closely guarded the offense is less likely to get the rebound. This makes sense because a wide open shot is often off of a defensive breakdown which leaves the defense scrambling and out of position, making it less likely that they are in a position to box out and secure the rebound.

This is just a quick look at some league-wide trends for last season. All this data is so new and there is a lot of room for deeper analysis especially as more data comes in from this season.