Nov 21, 2014; Washington, DC, USA; Cleveland Cavaliers forward LeBron James (23) dribbles the ball as Washington Wizards forward Paul Pierce (34) defends in the fourth quarter at Verizon Center. The Wizards won 91-78. Mandatory Credit: Geoff Burke-USA TODAY Sports
During last Friday’s Cavs-Wizards ESPN broadcast, there was a strange observation from one of the commentators. Perhaps it was Jon Barry with his usual flair for the wildly inaccurate. Perhaps it was someone else. But whoever it was, it got me thinking. The comment: “NBA offense is down so far this year.”
Now, if you recall, I boldly wrote last month about how ever-evolving NBA offenses could set a record in efficiency this season. The external reasons seemed fairly logical. Paul George’s injury terrorized the defensive juggernaut Indiana Pacers. Also in the East, the creation of Cleveland’s new pairing of LeBron James and Kevin Love seemed to be on the verge of record-breaking offensive production. Teams might be trying to replicate San Antonio’s Finals offensive masterpiece. And three-point shots are on the rise.
But a more currently applicable issue was then addressed briefly at the end: Does offense improve as the season goes on? This was left as a somewhat open-ended question. What factors could be affecting this? Does it make sense that offenses would start sloppy and sluggish after a long offseason? Is the math of offense truly winning out over defense over time, as Grantland’s Zach Lowe mentioned in his great piece on bourgeoning defensive analytics?
Fortunately, with the help of Nylon Calculus stat wizard Darryl Blackport, we have some better answers today than I did last month. Darryl, of course, has been gathering fantastic actual possession data for this website. And he was able to create cumulative NBA Offensive Rating day for every day during a season [1. Necessary shoutout to Randall Cooper (@laughingcavs), who has created two similar charts (here and here) on Twitter. Cooper is a great follow and was a big inspiration for this post.]. Here are the results:
First, you’ll likely notice the season-by-season history. I’ve covered this before: Offense took a gigantic drop during the 2011-12 lockout-shortened season (only 124 league game days, as opposed to the usual 170). The red line was 2010-11. League-wide offense rating was 104.49 at the end of the year. The orange line was 2011-12, not even close to that mark.
Since that lockout-shortened season, offense has been on the rise. But notably, it’s not just that offense is on the rise, but how it is on the rise. The final actual possession NBA ORTG for 2011-12 was 101.83. League-wide has been below that number each of the past two full seasons until about the first week of December. Gradually, offense got better and better. The final mark for 2012-13 was 103.11 and the final mark for 2013-14 was 104.01 [2. Note: These are different numbers from Basketball-Reference’s estimates that were cited in my article last month. Blackport’s numbers are way more accurate and will be used here at Nylon Calculus when available.].
This season? League-wide offense is already at 103.29 through games played on Nov. 24. That’s higher than at any point after the first week of 2012-13. The pace is just off the trend set in 2010-11. Here’s a quick look at cumulative ORTG at day 28 of the past five seasons:
What makes predictions difficult is that the in-season increase (although always positive) has been inconsistent. There are some mild ebbs and flows. In 2010-11, there really wasn’t a gigantic increase as ORTG hovered around a fairly steady range by Thanksgiving. last season, ORTG jumped tremendously after actually being below the 2012-13 mark at day 28.
So for now, we know that NBA offense is on a very impressive early pace. It is just below the mark set in 2010-11. To say that offense is down is inaccurate because we have seen concrete data the past few years that offense increases, but the rate of increases have been inconsistent.
What made the in-season increase in 2013-14 so substantial? Let’s revisit a chart that I shared in that article last month, via NBA.com/stats data this time around. Note some of the biggest monthly increases in last year’s offensive data:
According to NBA.com/stats, Turnover rate was 16.0 percent through Nov. 30. It was then a pretty consistent 14.9 percent for the rest of the season. Three-point attempt rate was 25.5 percent through Feb. 28. It was then a much higher 26.9 percent for the rest of the season. Those are the two biggest things one might spot immediately.
Turnover rate is only at 15.1 percent and we haven’t yet reached the very end of November yet. Three-point rate is at 26.5 percent already this season, thanks to Houston’s incredible 44.4 percent rate, by far lapping the rest of the field. Will both of those numbers decrease at the same rate as we saw during the course of last season? If so, then league-wide ORTG could easily shoot over 106 by the end of the year. If the potential isn’t there for those to drastically change as much, perhaps league-wide ORTG will just hover below 2010-11 for the rest of the year.
Either way, we’ve seen solid evidence this year that offense is finally at pre-lockout levels. It might be only a matter of time before the NBA breaks the record offensive season tied again in 2008-09. If offense can continue to increase rapidly in season, then it could happen sooner than we think.