Building Teams Using Quantative Analsyis in the NBA: The Last Rites

What? Chemistry and style are important?

LOL - been saying that for a long long time.......and had many arguments with the stat junkies as to why the stats are not transferable etc.

However, fear not Rocket fans, Morey uses stats as only a portion of his analysis, he also relies on the good old fashioned eyeball test.

A smart man realizes the limitations of his data, and fills in his weaknesses with others around him more capable in his shortcomings.

Morey is a smart man.

DD

 

Other than on this board, there is a generally negative bias against stats in basketball, so it would be very strange if there did not continue to be opportunities from further stats-based analysis. It's now 30 years after Bill James got into baseball, they are just now starting to understand defense.

But I agree that (current) stats are not the whole story. The chemistry issues Cuban brings up are partly personality, and partly play-style. You can use stats for the play-style aspects, based on players preferred specializations. Personality is harder, but the Rockets have done a pretty good job of that. Chemistry on our team seems solid.

 

If stats are not transferable, why does Morey bother to use statistical analysis at all in team building?

There's a confusion here. You are arguing one thing: statheads think stats are perfectly transferable and don't change with team context. But most statheads actually think something different: stats are important to look at and should be considered as part of the decision process, but not to the exclusion of traditional scouting.

You are willing to accept that stats can be useful to look at so long as it does not disagree with your opinions. Well, that is NOT how Morey or any serious team builder should use them. Might as well just throw them out completely in that case.

 

Quantitative stats can still give you a rough guess as to the absolute talent of a player irrespective of a system. Regardless of how much usage a guy gets, they still will not be able to maximize their PER beyond a certain level if they just aren't efficient. This holds true no matter what team or coach an NBA player has.

Case in point: you're not going to ever convince me that Brooks can one day put up a 20PER because his career curve suggests that 17-18 is his max, and that's even with about as much usage as you could reasonably give a guy last year. That same rationale held for Trevor Ariza.

Kevin Martin, however, has a great shot to return to the 21-22 range because he's done it before, and in his case I can convince myself that injuries and chemistry were the reason for his struggles last year.

I think advanced stats are great for telling you how a player truly performed offensively in the past. Chemistry and injuries will always play a role just as they do in any sport, but the variance is not so large as it is in, say, football with a RB as a function of his O-line.

 

I think you definitely can use stats for team building, but you have to be able to project how those stats my change once a player is on your team. As an example, you can't argue that stats didn't play a part in Morey's trade for Lowry. No scout would ever confidently say that trading a starting PG on a playoff team for a 3rd string PG for a non-playoff team was going to be a big win.

Questions I think Morey asks after he realizes that a player not yet on his team has some skills:
Will Coach be able to utilize that skill?
Will utilization of said skill diminish returns from other players?
Will I be able to get him at a price below value?
Is it likely that his perceived value will go up after his skills are on display?


Just as another example, a team could be in need of a 3pt shooter and find a bargain in a player that isn't known for his 3 pt shooting. If said player is currently asked to create his own shot and often has to throw up a buzzer beater. If there are few plays that allow the player to shoot from the corner where he's most comfortable. If the player does getting many reps in practice from the corner, since it's never asked of him.