"Hot hand" research

[bostrem00]

Do you also yell "Kobe" after each shot?

 

[besserheimerphat]

Doesn't an increase of 2.5% in shooting % by a player just mean they would make 1 more out of every ~40 shots?

It's not even that much. Again, according to the article, it's a 1.5% CHANGE in shooting percentage. The average NBA FG% is 45%. 1.2% of 45% is 0.54%. So an average NBA player that becomes hot doesn't go from 45% to 46.2%, he goes from 45% to 45.54%. That's 1 extra basket out of 186 shots! From the paper (not the article):

Looking at column (3), we finally see a positive and significant coefficient on heat. Though significant, we acknowledge that the effect size is modest. A player who makes one more of his past four shots sees his shooting percentage increase by 0.54 percentage points. Given that the average NBA player has a field goal percentage of about 45%, this represents about a 1.2% improvement. In the same vein, if a player makes two more of his past four shots (perhaps more indicative of what it truly means to be "hot"), we see a 2.4% improvement.

Based on these results, concluding that the Hot Hand exists is contingent on defining "hot" correctly. A player who makes three out of his past four layups is not hot, but a player who makes three out of his past four three-point attempts is. In other words, being hot is not about the absolute number of shots a player has previously made, but rather is about how much he has outperformed, conditional on the types of shots he has taken.

 

[VeloClone]

Bingo.

Player A has FG% = 35%. Player B has FG% = 40%. Player A is "hot" and Player B is "cold," so their adjusted FG% is 36% and 39%, respectively. You should still try to get the ball to Player B because he has the greater chance of making the shot. And the paper even says that "hot" Player A is more likely to take a higher difficulty shot which he has less chance of making.

I wouldn't say that I'm "cavalier" about it, but to me it's the correlation versus causation arguement. All of the hot hand studies have focused on correlation. But I want to see something that indicates causation other than "I played basketball and I just know it's true." Maybe it's just confidence and focus allowing muscle memory to take over. Well, there are methods to measure confidence and focus (like the brainwave research mentioned earlier). Put some electrodes on Kobe and LeBron and 3 other NBA players and have them play knockout - see if their brain activity is different when they hit a few shots in a row vs. missing a few in a row. Or wire up some 3 point shooters and have a 3 point contest. That's when I'll start to really pay attention. Figure out what it is that makes a player feel like they have a hot hand.

So in short Ejim and Niang should have remained in the game the other night. :spinny:

 

[twocoach]

Pretty interesting read IMHO. I know a person who was a star for Wisconsin a few years back that shed some light onto this subject. I guess I had no idea, but each player on the court is being followed VERY closely and would be taken out/put in based on info gathered from the shape the player was in & crossed referenced to the amount of time they were playing, physical effort and enviroment. Basically one big math equation. Once a player had reached what they thought was their peak of effective time they were sat for a specific amount of time no matter what. His senior year though they began waiving the stats if you were on fire at the time. Maybe most people already knew this info, but I was amazed to see how meticulous and calculated everything was.

Or you could do like Calipari and leave it to the players to decide when to come out...

 

[TurbulentEddie]

There is likely enough data to confirm that shots 3 to 5 feet from the hoop are great shots, maybe Niang's game will work in the NBA?

If you want to know more about shot efficiency, I would definitely recommend Kirk Goldsberry's work over at Grantland. Here's a primer:
Extra Points «

(Spoiler: 3 point shots, especially corner 3's, are almost as efficient as point blank shots, which are the most efficient shot.)

 

[andymhallman]

If you want to know more about shot efficiency, I would definitely recommend Kirk Goldsberry's work over at Grantland. Here's a primer:
Extra Points «

(Spoiler: 3 point shots, especially corner 3's, are almost as efficient as point blank shots, which are the most efficient shot.)

Hmm...I wonder how that could be. Perhaps the angle of the miss favors the offensive team more than misses from other angles?

And does it take into account the fact that shooting a corner three leaves your defense vulnerable to a fast break?

 

[TurbulentEddie]

Hmm...I wonder how that could be. Perhaps the angle of the miss favors the offensive team more than misses from other angles?

And does it take into account the fact that shooting a corner three leaves your defense vulnerable to a fast break?

It's in terms of points per shot, not per possession. But if you want to know where corner three rebounds usually end up, Goldsberry's got you covered there too: Where Do Rebounds Go? | CourtVision

 

[cyclonewino]

It's in terms of points per shot, not per possession. But if you want to know where corner three rebounds usually end up, Goldsberry's got you covered there too: Where Do Rebounds Go? | CourtVision

Very good article, another from the same site:

The Year in Shooting: A Visual Review | CourtVision

CFH's negative point of view on mid-range jumpers is based on the evidence, both shot efficiency and rebounding.

 

[CYcoFan]

I believe that the study the Original Post brought up should be viewed as a working model that is more a theory then it is fact or law. I believe this is what is causing some to see only a small correlation rather then causation.

I think the research done by those Harvard researchers needs to be expanded upon. They only use data drawn from NBA games. The "funnel" of players is so narrow at the NBA level that you have the elite of the elite at the NBA level. How would this translate to high major D 1 ball? Does an NBA player with that much talent and experience recognize when he is "feeling it" on a given night and does that lead to his hot hand? Is this something that all basketball players could experience? Would the "Hot Hand Effect" have a greater impact or less significant impact at the high major D 1 level?

I think we have realize that in the NBA a great player can impact an individual game more consistently then in any other sport. It is why Billy Beane says his "Money Ball" approach wouldn't work in the NBA. The NBA sees far fewer upsets in its playoffs then almost any other sport. Again how would this apply or translate from NBA to D 1 ball. I think it would amazing if that same team of researchers was allowed to use the same camera and tracking system for 3 or 4 of the High Major D 1 Conf's and see what the results were. (Maybe Big 12, Big 10, ACC and Pac 12)

 

[TurbulentEddie]

It is why Billy Beane says his "Money Ball" approach wouldn't work in the NBA.

Daryl Morey disagrees!