Pomeroy on steroids?

[Cheriehoop]

The problem is that that particular stat seems pretty difficult to measure. It can't be easily pulled out of publicly available box scores. I looked through play by play data the other day when you mentioned this and there wasn't an apparent way to tell the difference between half court and transition offense.

Wonder if time of a possession when scoring could be used. For example, if a team scores within say 10 seconds of a possession, it's a transition basket? I think it can be broken down by using play by play writeups.

 

[moqui]

The problem is that that particular stat seems pretty difficult to measure. It can't be easily pulled out of publicly available box scores. I looked through play by play data the other day when you mentioned this and there wasn't an apparent way to tell the difference between half court and transition offense.

I don't think it is that hard, it is just tedious (and a definition problem) . . . Pomeroy told me he used to do it for teams that hired him as a consultant (he has given that gig up) and that he might do it later in the season for the top 10 teams or so. My assumption is that it would not be a perfect measure, but you could define a transition baskets as any score that came after a live ball turnover and less than X seconds had expired or any scores that occurred after a defensive rebound and X seconds had expired. The crucial factor will be defining X - maybe the famous "7 seconds or less." I'm sure it wouldn't be a perfect measure, but it would give an approximation.

 

[BlackSquirrels]

I think your best bet to determine how teams rank in halfcourt and transition offense is to contact college basketball writers who mention Synergy Sports statistics in their work. I'd be surprised if Synergy does not track those numbers.

 

[Knicks411]

I think the biggest problem might be parsing out the transition possessions that don't turn into buckets.

 

[albanycuse]

"steroids" would imply that he is unusually muscular, you probably mean drugs where he makes no sense

 

[moqui]

I think the biggest problem might be parsing out the transition possessions that don't turn into buckets.

the play by plays list missed shots as well, so once you define X you should be able to parse out most of them.

The ones that will fall through the cracks are when a team gets a defensive rebound, tries to break, and turns the ball over. You wouldn't be able to tell from the play by play whether they were trying to break or not when the turnover occurred

 

[Knicks411]

Definitely Synergy is the way to go here.

 

[BuffOrange]

Btw, just noticed he has a blogpost about Wisconsin today. Haven't read it yet.

If you are going to trash the entire system based on the biggest outlier over the last six years, I suspect you had no desire to use the system in the first place.

That about sums it up.

 

[SWC75]

Sometimes you have to give your results the "look test". One year Mike Aldretti was the "Total Average" champion of the Major leagues over players like Don Mattingly and Eric Davis because he was a part-time platoon player having a career year. Total Baseball had a "Total Player Ranking" that listed Joe DiMaggio as the 36th best player of all time. Bobby Grich was #28.

And, yes, Mookie Jones' career "Net Points" per 40 minutes is 14.0 while Gerry McNamara's is 13.4.

No system is perfect. If the results don't come out the way you'd think they would you then try to figure out why and that's where the truth really comes out. But if most of the results look right, the system is probably pretty good.

 

[shandeezy7]

I think I need to go dust off my college statistics book so I can follow this conversation.

 

[Cusefannotindc]

I don't think it is that hard, it is just tedious (and a definition problem) . . . Pomeroy told me he used to do it for teams that hired him as a consultant (he has given that gig up) and that he might do it later in the season for the top 10 teams or so. My assumption is that it would not be a perfect measure, but you could define a transition baskets as any score that came after a live ball turnover and less than X seconds had expired or any scores that occurred after a defensive rebound and X seconds had expired. The crucial factor will be defining X - maybe the famous "7 seconds or less." I'm sure it wouldn't be a perfect measure, but it would give an approximation.

I am sure I could find it easier if I were getting paid to do it as well.

 

[Knicks411]

Sometimes you have to give your results the "look test". One year Mike Aldretti was the "Total Average" champion of the Major leagues over players like Don Mattingly and Eric Davis because he was a part-time platoon player having a career year. Total Baseball had a "Total Player Ranking" that listed Joe DiMaggio as the 36th best player of all time. Bobby Grich was #28.

And, yes, Mookie Jones' career "Net Points" per 40 minutes is 14.0 while Gerry McNamara's is 13.4.

No system is perfect. If the results don't come out the way you'd think they would you then try to figure out why and that's where the truth really comes out. But if most of the results look right, the system is probably pretty good.

I think Bill James has a line something to the effect that you don't want a ranking system to be totally by the consensus, because then you aren't adding anything. But you don't want it to be totally off the wall, because then it's probably wrong. No system is perfect, you definitely want one that challenges conventional wisdom while at the same time not looking totally bat-s*** crazy.

 

[moqui]

I think your best bet to determine how teams rank in halfcourt and transition offense is to contact college basketball writers who mention Synergy Sports statistics in their work. I'd be surprised if Synergy does not track those numbers.

Pomeroy sold his consulting biz to Synergy, and he said that they do, in fact, track transition and half court offense. I don't believe that they sell those numbers to the general public, however.

 

[Bornorange]

"steroids" would imply that he is unusually muscular, you probably mean drugs where he makes no sense

Finally somebody who understands my mixed up metaphor, now if my daughter would get home she could explain this math stuff everyone else is talking about. I still say Pomeroy must have steroid rage.

 

[Cusefannotindc]

the play by plays list missed shots as well, so once you define X you should be able to parse out most of them.

The ones that will fall through the cracks are when a team gets a defensive rebound, tries to break, and turns the ball over. You wouldn't be able to tell from the play by play whether they were trying to break or not when the turnover occurred

Off of the Tulane and Seton Hall play by play, I made any possession beginning from a turnover or defensive rebound that ended with a shot or foul on the other team with 10 seconds.

Tulane game
Transition - 1.07 points per possession
Half court - 1.45 ppp

Seton Hall game
Transition - 1.04 ppp
Half court - 1.23 ppp

 

[moqui]

Off of the Tulane and Seton Hall play by play, I made any possession beginning from a turnover or defensive rebound that ended with a shot or foul on the other team with 10 seconds.

Tulane game
Transition - 1.07 points per possession
Half court - 1.45 ppp

Seton Hall game
Transition - 1.04 ppp
Half court - 1.23 ppp

very interesting . . . I think 10 seconds is probably too long; probably even 7 seconds is too long. Did you chart all turnovers, or did you only do live ball turnovers? Please post the numbers from other games if you do them.