APBRmetrics

Just wondering if anyone had some good results on this. If offensive efficiency generally trends upward with pace, the offensive efficiencies of faster pace teams may be a slight overestimates of offensive ability in comparison to the offensive efficiencies of slower paced teams. This is because faster pace teams then have an advantage that is generated by pace, and pace is a variable that is basically a constant for any two teams playing a game against each other. A new statistic could be generated that adjusts the offensive efficiency of a team based on the pace that they employ and how offensive efficiency varies with pace in general.

Just wondering if anyone has done considerable work on this issue.

Thanks.

As far as I know, any effects of faster-paced teams having offensive "bonuses" are nullified by higher turnover rate.

But realistically, how would we "adjust" a team's offensive efficiency by pace? Wouldn't their offensive efficiency already be reflected accurately? Or are you saying that we would somehow use math to drop good fast teams (Miami) efficiency against a slow team (like...the Pacers)?

EDIT: In the most recent 5 years in the NBA, Team Pace and ORTG have an R^2 of 0.02.

The offensive efficiencies as they are defined are reflected accurately at any pace. I was saying that we could create an adjusted offensive efficiency that considers the impact of pace on efficiency to more accurately measure offensive effectiveness. However, this would obviously be difficult, even if there is a clear trend in the plot of pace verses offensive efficiency. This is because there is no way to separate causation and correlation. Offensive efficiency might increase with pace, but maybe teams that are better offensively simply choose to play at a faster pace in general. This is different than the faster pace actually being a cause of the greater offensive efficiency.

OK, but offensive efficiency really *doesn't* increase with pace.

Well that is what I wanted results on because I had never investigated it. I'll take your word for it though. Maybe my idea that offensive efficiency might increase with pace was influenced too much by all those Sun teams. I also noticed that Denver and Houston are two of the faster paced teams and also two of the more efficient offensive teams.

Yeah I should have posted a graph but that's what I was referring to about the R^2 value. I think it is true that extremely good offensive teams might like to run the ball, but it's not significant enough to be explanatory of Pace *in the long run*.

But I think you're on to something (which would be much easier to see on the surface in college which has 10x as many teams) - finding which teams tend to play better or worse when they play faster or slower. Definitely some correlation on a per-game basis?

Kind of sounds like I'm contradicting myself

bbstats wrote:But I think you're on to something (which would be much easier to see on the surface in college which has 10x as many teams) - finding which teams tend to play better or worse when they play faster or slower. Definitely some correlation on a per-game basis?

Ken Pomeroy has been doing this for years now.

I think if you graph team average age against pace then you'll see a relationship, and since the aging curve exists I would actually expect the best teams to sit around the middle of pace space.

I did a very small study with this and found basically there's a weak correlation of offensive efficiency and pace going back a few seasons.

http://ascreamingcomesacrossthecourt.bl ... ation.html

It's not on the level of analysis as you usually see on this site, but the graphs are at least nifty.