Yeah, I mean but it has to be measured using actual play by play data, though, right? Ie when the ESPN game tracker says player x attempted 3 point jump shot at 16:41.
Because every possession ends with a FGA, a TO, or FTs and possessions are equal
3 Likes
That seems like a reasonable assumption. But I havenāt run it down on their website glossaries
Got it. I wasnāt sure if they made any adjustment for pace of opponents played like there is with efficiencies.
Iām curious (curious enough to type out a post, not to actually look at the data) if power conference teams tend to play faster because they have better athletes
Basically the answer is yes, PBP data. Not sure how he accounts for ORs. It seems like circa 2013 he wasnāt able to clean that out.
Since Iām not reading that, my guess is the data just got better. ORs are not counted as extra possessions. they are subtracted from FGA. So, assuming the timing is precise (which seems like a leap across all D1), you just need a time stamp on makes, DRs, TOs, and FTs and a big ass excel spreadsheet to calculate.
If you had good enough data (to your point, questionable in all of D1), you could measure something like first shot possession length, until any rebound (offensive or defensive). But given that heās still linking an explanation from 2013, I wonder if heās just still counting the full possession.
Also worth noting that these components arenāt adjusted, but overall pace is adjusted.
1 Like
Hmmmā¦delving into the realm of a debate about almost completely useless informationā¦
A) thereās no need to adjust average possession length for pace because it is what it is
B) still not reading it
C) ORs are subtracted from FGA. Even though the shot clock resets, an OR is not counted as an extra possession (even though it kinda should be) in a pace calculation.
D) fine maybe Iāll read it when I walk my dog later. Looked like a lot of words.
He adjusts the overall pace for competition. So if a super fast team plays Bennett-era Virginia he finds a way to normalize that
Irrelevant for this calculation, from what I can tell. Pace measures possessions. These O/D component** stats donāt. Itās basically just a measure of how long you have the ball before the other team has the ball. So circa 2013 at least an OR would just make it seem like a longer possession.
** so theyāre not really components, to be clear
I expect when most people talk about pace, the metric they really want is āhow long from taking possession to first taking a shot (or drawing a foul)ā. That ignores possessions ending in turnovers, it ignores everything that happens after an OR, and of course it ignores defensive possessions. So itās a completely different metric! But itās still probably what people are casually thinking about when āpaceā comes up.
3 Likes
Also, to be clear, this is called a discussion. If we were debating, youād know, because youād be wrong
4 Likes
Exactly! And for all I know, thatās what Ken is now doing, but if so, he hasnāt updated the explanation
Yes I was just coming back to confess my sin on this matter
1 Like
I was wrong. So itās a debate. So youāre wrong.
1 Like
Consider yourself absolved. Itās tricky terminology - Iām not being consistent with it.
Pace is adjusted (in Kenpom)
Average possession length isnāt
Itās such a different metric that it would be really confusing to change it and yet keep calling it the same thing, so I hope he didnāt. The only good way to introduce it would be as a new metric (ātime to first shotā).
Seems more like something Torvik would do tbh. He likes to fiddle with new toys.
1 Like
How was the .475 derived? Iāve also seen .44 and .4 elsewhere.
What do you do with a TO followed immediately by another TO? Or ball goes OOB off opposing player? New possession probably, but opposing team basically never had real possession. Also happens so infrequently that it probably isnāt worth parsing the data for that.
Yup, saw that. But also saw this. https://kenpom.com/blog/the-possession/
The glossary description, which uses .475 is from 2012, and the other one is from 2004, so I guess he settled on .475
I just clicked the first google link. I donāt think the value of one FTA is going to change much in any argument about pace but you guys do you guys.