Re: [APBR_analysis] Re: the Ruben Patterson effect
- I bring it up b/c there's a couple cases in the chart where this is clearly incorrect -- Earl Boykins playing nearly half his minutes at off guard, for instance, or Zoran Planinic having 0 minutes at point - with obvious ramifications for the value of their P-Diff numbers.
last season denver as a team played 19730 minutes. thats 19730/5 = 3946 minutes per position (C, PF, SF, SG, PG). andre miller played 2838 minutes. assuming for now he was the PG when he was in the game someone else had to play 3946-2838 = 1108 minutes at PG when miller wasn't in the game. earl boykins played 1849 minutes, so if he played those 1108 minutes at PG when miller was out of the game then at a minimum he was on the floor with miller at the same time for 1849-1108 = 741 minutes (odds are they were on the floor together a little more) and technically speaking when they were on the floor together one was the PG and the other the SG...
now you can think all you want that when boykins and miller were on the floor together that the 5'5" boykins was the PG and the 6'2" miller was the SG because of their heights, but when you look at the stats miller was clearly the "truer" PG between the two - he had the higher touches/min (1.6 to 1.5), shot the ball less often per touch, and passed the ball more often per touch (not by much tho) compared to boykins, so from that vantage point he could easily be called the PG and boykins the SG...
now because of their height difference odds are miller guarded the better or perhaps taller player when he and boykins were on the floor together at the same time, but even if that was true if when the two were playing together miller still brought the ball up the floor more often that boykins did to me that means he was the true PG...
my point here is not that either miller or boykins was actually the PG and the other the SG when they played together, but just to show how truly difficult it is - without having watched all the video - to designate one player or another as having played a particular position. how about assigning def FGM and def FGA to defenders when a team plays zone? again, not an easy thing to do, but something that needs to be done to get meaningful data, data that until we watch all the video is still much much more valuable than no data at all...
I don't mean this as a knock on Roland & Co. -- putting that together was a monumental task and they deserve our thanks for it -- but it warrants mentioning that we have to get the matchups part right, or our resulting data isn't reflecting reality.
agreed - but this is where we all can help. for a system of collecting useful/meaningful data from play-by-play to work some assumptions have to be made, and in all honesty without watching all of the video from every single game i just don't see how much more improved the effort by roland and his staff at 82games.com (or anyone undetaking this kind of endeavor and then making the data available to the public for free, DanVal, whatever) could be - not that roland won't find more ways to do just that, improve the data. his data reflects reality far more than no data - and we had no data that was anything like his data just a couple of years ago....
but having said that i think any information we can pass along to roland to help his effort would benefit us all - its the least we can do to return the favor of the information he has made freely available. and he has certainly listened to us and made changes/enhancements to his methods based on our input. just look at how our APBR_analysis discussions have changed over the past 1-2 years - meaningful discussions now include PER differential, on court versus of court stats, etc. and not just per game or per minute stuff...
- -----BEGIN PGP SIGNED MESSAGE-----
On Wed, 4 Aug 2004 smckibbi@... wrote:
> Great idea. Perhaps the criteria for deciding which swingman is guarding who
> would be the team's overall def points per 100 possessions when the swingman
> is on the court. The one with the lower pts/100 poss is guarding the better
> offensive player.
No, this is wrong. It's wrong because it favours bench players. Guys like
Bowen and Artest (and especially Bowen, because he's older) usually don't
guard the opposing bench's best player; they take a rest when the other
side's coach brings his bench players onto the floor.
You get a guy like Peterson on the Raptors, he's usually on the floor when
the other side's bench is, and usually not when their starters are. Now,
he's a good defender, but his stats are going to get enhanced under this
Here's a Modest Proposal: Track the overall offensive production per
minute or possession of the opposing team's players that a particular
player defends. Compare that to their offensive production per minute or
possession while he's on the floor. A negative result means a good
defender, a positive result means a bad one, zero means average.
GnuPG public key available from
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Made with pgp4pine 1.76
-----END PGP SIGNATURE-----