Offense is back and it’s better than ever. Through the first nine days of the season, teams were averaging 104.6 points per 100 possessions. Tuesday night was a shining example of our fun new world. All of the top games featured lots of points and most of them were quite exciting:
If the season ended on Tuesday, this would have been the second-highest offensive rating in history, short of last season’s 105.2. But the season didn’t end Tuesday and the real hoopheads know that teams tend to get better offensively as the season progresses. Comparing the first nine days of the season to the first nine days of past years, this season’s start is easily the best, edging out the 2014 season which was at 103.8.
I am sure there are still some complainers out there, but it’s hard to really have a problem with the state of the game. This season will likely set records for free throw percentage, two-point percentage, and three-point attempt rate. What’s not to like?
In addition, non-steal turnover rate is at an all-time low. That’s our simplest proxy for the number of offensive fouls being called so officials are sticking to the change to the charge rule adopted last season. It seems that players have fully adjusted and aren’t as willing to every try for the charge. Joe Few even tried to challenge a highlight-reel dunk from V.J. Edgecombe, with much potential for personal embarrassment that was ultimately realized. If that’s not evidence of our new culture, I don’t know what is.
So just how good will offense get as the season progresses? Let’s try to predict the full-season offensive rating based on the data we already have and see how badly the previous record will get smashed.
For starters, here’s a graph of the last 28 seasons, showing the relationship between offensive rating after nine days, and offensive rating over the entire season:
The blue line is the linear trendline based on all of the data and as a first estimate we can use this to get a forecast of where the full-season value will end up. If you eyeball where this season’s value of 104.6 would intersect the line, it looks like we get something like…just a little less than 105.0. Um, that actually wouldn’t even beat last season’s value, let alone crush it. Is this real life or is it just make believe? Could this season possibly not set a record?
It’s sort of real. The gray line on the plot is where the points would end up if there was no change from the early season values to the final values. The seasons that start with better offense tend to end up closer to that gray line. Which means they don’t have as much of an increase through the rest of the season.
Now, many of the seasons with worse offense occurred a while ago and it’s hard to disentangle seasonal effects here. For instance, the dot from 2014 - the best offensive start before this season - was fueled by an early emphasis on cleaning up contact. As officials stopped calling fouls as that season progressed, that artificial boost in offense went away, effectively counteracting the improvement in shooting and turnovers over the course of the season. There was barely any improvement in offensive rating that year.
But excessive fouls are not a factor this season. In a normal year, a lot of the improvement in offense over the course of the season is due to shooting and turnovers. And given that shooting and turnovers are at their best numbers ever to this point in the season, there may not be that much improvement in offensive rating coming. In addition, we know that the change in the charge rules for secondary defenders has fundamentally improved offenses as well and that’s a change that’s already baked into the early season numbers.
Sure enough, after an impressive start to the season where teams combined for an opening-night record offensive rating of 104.4, the cumulative average hasn’t budged over the past week and a half. In fact, after a sluggish Wednesday, the season average after day 10 of the season is down to 104.3, lower than opening night:
And yes, it’s a small sample and all that. But over the previous 28 seasons, there was an average increase in offensive rating of one point per 100 possessions over the first ten days. So teams across the nation are already a bit behind schedule in their effort to crush last season. I’m starting to have my doubts!
I’d still bet that this season finishes at a record level, although it’s far from a given. A slightly more sophisticated model, assuming a non-linear relationship between early offensive rating and season offensive rating, and using some turnover and shooting information, indicates that it’s more likely than not that this season’s offense beats last season’s record.
Furthermore, last season still had a decent increase despite starting out hot. I mean, there’s really no reason to explain why offense would get worse in coming weeks, other than officials deciding to call the charge rule from the 2010 rule book1. And that didn’t occur last year despite so many of my twitter mentions telling me that’s exactly what would happen.
It deserves emphasizing that men’s college basketball is the best it’s ever been from an aesthetics standpoint. I’m here for the extinction of games where the winning team has fewer than 50 points. There were a record-low seven such games last year and last night we had this season’s first2.
The surge of higher-scoring games is especially fun in an environment where there aren’t many fouls called. The story of this season may still be about record-setting offense. Or it could be that offenses that were more prepared at the start of the season than ever before.
This piece from Eamonn Brennan announcing the implementation of the charge circle for the 2011 season really hits. Even then it felt ten years overdue.