Is it time to revisit the topic of playing time?

[Peoriabird]

Earlier in the Year, there was a rather spirited discussion on who should start/get more playing time between Zaza/Ivan & Deshawn Stevenson. At the time I was using +/- as one of my metrics to compare the impact each player had while on the floor and was told that the sample size was too small and that this discussion should be revisited after an adequate sample size was obtained. Well the time has come and here are the stats!

http://www.82games.com/1213/1213ATL1.HTM

As you can see, eventhough Deshawn Stevenson generally plays with the team's better players i.e. the starters, his overall impact is negative. On the other hand, Zaza generally does not play with the teams's better players, his overall impact is positive.

I am interested in seeing how the YKW apoplogist will attempt to explain how this happened.

[gurpilo]

Look, for all of you blaming Jeff, he is the best player on this metric!!! Not a surprise for me, I wonder what this statistic would look like if we definately would give him the keys of the car...

[MaceCase]

You must have been hibernating in your cave when this came out:

http://blogs.ajc.com/hawks/2013/01/11/atlanta-hawks-korver-second-on-hawks-best-in-plus-minus-rating/?cxntfid=blogs_hawks

Korver is part of a Hawks’ five-man combination, along with Al Horford, DeShawn Stevenson, Josh Smith and Teague, that ranks 11^th in the NBA with a plus-49. The Hawks also have the 15^th ranked five-man combination with Lou Williams, Korver, Teague, Smith and Horford at a plus-44 rating.

Pay mind to the Win% in your own link too.

[Peoriabird]

You must have been hibernating in your cave when this came out:

http://blogs.ajc.com/hawks/2013/01/11/atlanta-hawks-korver-second-on-hawks-best-in-plus-minus-rating/?cxntfid=blogs_hawks

Pay mind to the Win% in your own link too.

Again is that line up successful because of Stevenson or despite Stevenson? Also if I'm not mistaken, Win% also is a function of who is on the floor with you right? Or are you still making the argument that YKW is a better player than both Zaza and Ivan Johnson?

[MaceCase]

If it's a matter of is it "because or despite" why aren't any other lineups approaching it by plus/minus then if replacing Deshawn with "productive player X" is all it takes? And Win% and +/- are both dependent on who is on the court with you, against you and replacement value of who comes in for you so again, you sound like a scorned lover. There is simply no such thing as individual +/- and further down that argument it should be noted that if one starts and the others come off the bench then you have to take into account the quality of the opponents they face as well. This is, of course, entirely too logical for someone arguing from a clearly bitter base.

Edited January 12, 2013 by MaceCase

[thecampster]

If it's a matter of is it "because or despite" why aren't any other lineups approaching it by plus/minus then if replacing Deshawn with "productive player X" is all it takes? And Win% and +/- are both dependent on who is on the court with you, against you and replacement value of who comes in for you so again, you sound like a scorned lover. There is simply no such thing as individual +/- and further down that argument it should be noted that if one starts and the others come off the bench then you have to take into account the quality of the opponents they face as well. This is, of course, entirely too logical for someone arguing from a clearly bitter base.

Mace is your metric total or per game....thanks for clarifying.

[MaceCase]

Mace is your metric total or per game....thanks for clarifying.

Anything focusing on +/- has to be accumulative to have any reasonable meaning.

[thecampster]

Anything focusing on +/- has to be accumulative to have any reasonable meaning.

Quite the opposite

If a player plays 30 minutes a game and posts a plus minus average of +4.3 per game and a player plays 10 minutes per game and posts a +3.0 plus minus average, the per 36 (measure of efficiency) for player a is +5.16. But for player B it is +10.8. Clearly player B's +/- for his time on the floor is better. But translated out to a full season with cumulative numbers, player A has a total of +352.6, while player B has a total of 246 making player A look far more impactful than player B which is not the case.

I want per 36 +/- over a stretch of at least 20 games and closer to a full season for it to be a useful metric of efficiency. It is the best way to use the stat.

Win% as well is a bit skewed as it favors starters over reserves.

[MaceCase]

Camp, +/- is an accumulative stat. Even the measure you are trying to create will still just be taking the Net after X amount of games and then trying to divy it up and apply it on a per minute basis. This would still be a monumental waste of time because all it would lead you is to an incorrect assumption that someone like Mario West has just as much if not greater impact than Josh Smith because per minute your stat is telling you they are both plus 4. Now would you agree with this considering that Josh is facing multiple opponents while Mario is coming in on end of quarter situations? This is why it's an extremely flawed stat because it suffers from far too many variables while telling you little about an individual. Developers of Adjusted PlusMinus even require 3 YEARS worth of statistics to reach any meaningful conclusions on a particular player's impact . This is why Peoria is a fool, yes, a FOOL for trying to make any definitie claims by the metric when aspects of itself contradict his very own argument. I'm more than happy enough to play along though and point out his folly.

[Peoriabird]

This is why Peoria is a fool, yes, a FOOL for trying to make any definitie claims by the metric when aspects of itself contradict his very own argument.

At least I learned something as a result of this thread. Thanks!

[TheNorthCydeRises]

Earlier in the Year, there was a rather spirited discussion on who should start/get more playing time between Zaza/Ivan & Deshawn Stevenson. At the time I was using +/- as one of my metrics to compare the impact each player had while on the floor and was told that the sample size was too small and that this discussion should be revisited after an adequate sample size was obtained. Well the time has come and here are the stats!

http://www.82games.com/1213/1213ATL1.HTM

As you can see, eventhough Deshawn Stevenson generally plays with the team's better players i.e. the starters, his overall impact is negative. On the other hand, Zaza generally does not play with the teams's better players, his overall impact is positive.

I am interested in seeing how the YKW apoplogist will attempt to explain how this happened.

It happens because when the Hawks go small, it's not necessarily a good thing. Look at this link that lists the 5-man lineups that Stevenson is a part of, and tell me what you see.

http://www.82games.com/1213/12ATL5.HTM#5man

Specifically notice which position Stevenson is playing, and his corresponding +/- number.

And you're not a "fool". You simply needed to dig more deeper.

Edited January 12, 2013 by northcyde