Pretty, pretty cool.
http://espn.go.com/blog/playbook/dollar ... ba-gambler
Voulgaris Article
Re: Voulgaris Article
Very interesting.
I thought this quote kinda incapsulates things for many here:
"He feels now that for all the momentum of the quant revolution in the NBA, there may be a glass ceiling for its true practitioners. “There’s a real disconnect between the basketball people, the business people and the -- for lack of a better word -- stats nerds. The stats nerds have no chance of ever becoming general managers. They’re just being used as a resource to mine.” At the end of the contract’s term, in the summer of 2010, Voulgaris decided to end his NBA flirtation and go back to being a gambler."
Re: Voulgaris Article
We must band together and fervently deny being nerds.
Re: Voulgaris Article
Interesting, yet rather strange. Reads to me like a very public job interview, despite the text.
Regarding the remarks on crushed stat nerd aspirations, however, it seems to me that there is a bit more to it than that (ironically paralleling the decreasing return on gambling investments expected to be seen over time). The ambitious stat nerd is doomed, less because of some culture clash, but because what he has to offer is either of rapidly diminishing value (given the sclerotic nature of the player labor market) or because his services are being efficiently replaced by technology.
Someday soon someone will publish, for free, the "final" +/- reckoning of player values, incorporating aging curves and all other relevant bells and whistles (Jeremias is dead, long live Jeremias). At that point, the market for player transactions is basically set. Never again will a future Mike D'Antoni refuse to give up an Amare Stoudemire for a Kevin Garnett. (And even if there remains such a Mike D'Antoni, there won't be 29 of them.) And so on and so forth. Maybe there will be a bit of room left for special stat sauce when it comes to drafting, but probably not much. And if the David Stern cartel is smart, it will socialize the expense and provision of all such information.
So what comparative advantage might then be left for the stat nerd GM? Line-up and match-up optimization? Is there much of value that remains (now and in future) that a check to Mark Cuban for his software cannot deliver?
The window of opportunity is closing, but it seems to me that a famous gambler/ex-gambler/re-gambler could yet climb through.
Regarding the remarks on crushed stat nerd aspirations, however, it seems to me that there is a bit more to it than that (ironically paralleling the decreasing return on gambling investments expected to be seen over time). The ambitious stat nerd is doomed, less because of some culture clash, but because what he has to offer is either of rapidly diminishing value (given the sclerotic nature of the player labor market) or because his services are being efficiently replaced by technology.
Someday soon someone will publish, for free, the "final" +/- reckoning of player values, incorporating aging curves and all other relevant bells and whistles (Jeremias is dead, long live Jeremias). At that point, the market for player transactions is basically set. Never again will a future Mike D'Antoni refuse to give up an Amare Stoudemire for a Kevin Garnett. (And even if there remains such a Mike D'Antoni, there won't be 29 of them.) And so on and so forth. Maybe there will be a bit of room left for special stat sauce when it comes to drafting, but probably not much. And if the David Stern cartel is smart, it will socialize the expense and provision of all such information.
So what comparative advantage might then be left for the stat nerd GM? Line-up and match-up optimization? Is there much of value that remains (now and in future) that a check to Mark Cuban for his software cannot deliver?
The window of opportunity is closing, but it seems to me that a famous gambler/ex-gambler/re-gambler could yet climb through.
Re: Voulgaris Article
You see, I don't think plus-minus with an ageing curve, a strong spm prior, and allowance for in-season weighting solves basketball. I think a statistic of that sort goes a long way to doing so, but I think there is a huge, underused area of analysis that is also of important: that of player characterisation and player-type interaction.
The best recent example of this is the sky-high predictions that were coming out of the stats community for the Heat when they acquired LeBron and Bosh on top of Wade. Assumptions were made about the additive nature of plus-minus statistics which simply do not hold at these extremes. The fact was that Miami employed three potent outside scorers, three franchise players, and the result was really good, but not all that amazing like some expected. Not greater than Wade and Shaq, yet does anybody doubt that LeBron + Bosh is generally better than Shaq alone? I believe it was because the combination of dominant centre plus dominant outside scorer made them far more versatile than simply filling every gap up front with a very talented scorer. Some research has been done on this, but more research and characterisation could help not only player trades, but minute allocation and lineup design.
The best recent example of this is the sky-high predictions that were coming out of the stats community for the Heat when they acquired LeBron and Bosh on top of Wade. Assumptions were made about the additive nature of plus-minus statistics which simply do not hold at these extremes. The fact was that Miami employed three potent outside scorers, three franchise players, and the result was really good, but not all that amazing like some expected. Not greater than Wade and Shaq, yet does anybody doubt that LeBron + Bosh is generally better than Shaq alone? I believe it was because the combination of dominant centre plus dominant outside scorer made them far more versatile than simply filling every gap up front with a very talented scorer. Some research has been done on this, but more research and characterisation could help not only player trades, but minute allocation and lineup design.
Re: Voulgaris Article
The point is not that such an approach "solves basketball". It doesn't. The issue is whether what remains of the basketball solution would provide meaningful, actionable information. My sense is that there is simply too much junk in the CBA for this to be the case.v-zero wrote:You see, I don't think plus-minus with an ageing curve, a strong spm prior, and allowance for in-season weighting solves basketball.
Maybe, but my argument about this would be the same: between what will become freely available and the constraints of the CBA, there will be little latitude to profit from any such proprietary knowledge.v-zero wrote:I think a statistic of that sort goes a long way to doing so, but I think there is a huge, underused area of analysis that is also of important: that of player characterisation and player-type interaction.
I recall having been on the more optimistic side of the offensive potential of the 2010-11 Heat, not believing in strict additivity of +/- statistics, but that they would have done a bit better job predicting performance. After the fact, it isn't clear to me that the shortcoming owed entirely to the naivite of the approach. Regardless, the consequences of piling on of super-star talent isn't the typical NBA management concern.v-zero wrote:The best recent example of this is the sky-high predictions that were coming out of the stats community for the Heat when they acquired LeBron and Bosh on top of Wade. Assumptions were made about the additive nature of plus-minus statistics which simply do not hold at these extremes. The fact was that Miami employed three potent outside scorers, three franchise players, and the result was really good, but not all that amazing like some expected. Not greater than Wade and Shaq, yet does anybody doubt that LeBron + Bosh is generally better than Shaq alone? I believe it was because the combination of dominant centre plus dominant outside scorer made them far more versatile than simply filling every gap up front with a very talented scorer. Some research has been done on this, but more research and characterisation could help not only player trades, but minute allocation and lineup design.
Re: Voulgaris Article
I feel that view is pessimistic, I think it's more likely we get one team taking a non-superstar-roster to the title through such work and then the equilibrium/problems with CBA will actually occur. Until player and lineup valuation can be combined I believe there will be players out there that are situationally undervalued and could be exploited.schtevie wrote:Maybe, but my argument about this would be the same: between what will become freely available and the constraints of the CBA, there will be little latitude to profit from any such proprietary knowledge.
-
AcrossTheCourt
- Posts: 237
- Joined: Sat Feb 16, 2013 11:56 am
Re: Voulgaris Article
He had one of my favorite examples of player evaluation: Jason Collins.
He can't rebound, he doesn't protect the rim, he's unathletic and he absolutely cannot score. How is he useful? He's one of the best post defenders in modern league history, but that's only against huge centers. His value is completely dependent on the opponent. There is no single number rating for Collins. Versus Howard he's an asset. Versus Nick Collison he is worthless.
He can't rebound, he doesn't protect the rim, he's unathletic and he absolutely cannot score. How is he useful? He's one of the best post defenders in modern league history, but that's only against huge centers. His value is completely dependent on the opponent. There is no single number rating for Collins. Versus Howard he's an asset. Versus Nick Collison he is worthless.
Re: Voulgaris Article
Right, in addition to looking at player types, and the interactions between player types on a team, the other big thing to look at is interactions between player types when opposing each other. Collins able to defend against big post players. Gary Payton had trouble defending against the really quick scoring PGs ... Kevin Johnson, Tony Parker, etc. Granted everybody has trouble stopping a really quick point guard, but whereas Payton was a better than average defender against most point guards, players like KJ and Parker made him look like an ordinary defender.
And this type of analysis might extend to specific 1-on-1 matchups. Sabrmetricans warn against looking at individual batter-vs-pitcher statistics due to small sample size. But I wonder if 20-30 minutes of 1-on-1 matchups, 4 times a year, might give us enough data after two or three years do draw useful conclusions. Doug Moe liked to sic Bill Hanzlik on Ralph Sampson. The Celtics were thought by the fans to have acquired Dennis Johnson to slow down the Boston Strangler, Andrew Toney (this seems to have been an urban legend however).
Defense on the perimeter, defense inside, against quick players, against slow players, against good outside shooters, against high flying dunkers, against players moving without the ball, against players with killer crossovers, etc. etc. It's a multi-dimensional skill, and defenders with different skill sets will have different success depending on who they're matched up against.
And this type of analysis might extend to specific 1-on-1 matchups. Sabrmetricans warn against looking at individual batter-vs-pitcher statistics due to small sample size. But I wonder if 20-30 minutes of 1-on-1 matchups, 4 times a year, might give us enough data after two or three years do draw useful conclusions. Doug Moe liked to sic Bill Hanzlik on Ralph Sampson. The Celtics were thought by the fans to have acquired Dennis Johnson to slow down the Boston Strangler, Andrew Toney (this seems to have been an urban legend however).
Defense on the perimeter, defense inside, against quick players, against slow players, against good outside shooters, against high flying dunkers, against players moving without the ball, against players with killer crossovers, etc. etc. It's a multi-dimensional skill, and defenders with different skill sets will have different success depending on who they're matched up against.