WSJ: Write-up of academic study of the "Hot Hand"

[Trapped in California]

If you are a hardcore basketball fan and/or stats geek (or behavioral econometrics), this is candy. You may need to have a Wall Street Journal subscription.

http://www.wsj.com/articles/the-hot-hand-debate-gets-flipped-on-its-head-1443465711

[Panther100]

Re: WSJ: Write-up of academic study of the "Hot Hand"

If you are a hardcore basketball fan and/or stats geek (or behavioral econometrics), this is candy. You may need to have a Wall Street Journal subscription.


http://www.wsj.com/articles/the-hot-...ead-1443465711

The article is right. A game setting is difficult to test the hot hand theory. I subscribe to Investor's Business Daily.

[UNIdad]

Re: WSJ: Write-up of academic study of the "Hot Hand"

Is this similar to the "Let's Make A Deal" principle?

If you pick Door #1 and they eliminate Door #2, you should change your pick to Door #3. It might seem like you still have 1/2 odds, but you don't. You still have 2/3 odds of being wrong because you still have only the 1/3 odds with your original pick. If you change your pick, you are taking advantage of the new 1/2 odds of being right.

Not sure I explained that right, but it is interesting.

[sivert]

Re: WSJ: Write-up of academic study of the "Hot Hand"

In a way, yes. Mathematically, it is like a more complicated version of the Monty-Hall paradox.

And it has nothing to do with a "hot hand" or sports. And it's about as ground-breaking as me publishing that the sky is blue.

Imagine that I make some cookies and I make sure that every cookie has 12 chocolate chips.
How many will each quarter cookie average?
3, but it might be: (0,5,3,4) or (0,0,6,6) or (1,4,5,2) and so on.

Now- what if I say, "Do the cookies have a 'hot-hand'?"
So I take some of these cookies and I pick the left two quarter cookies and count the chips.
Only in cases where each of the two left quarters have at least 2 chips, I count the chips in the right two quarters. Guess, what? The cookies have a "cold hand"! Duh! Of course they do. If you eliminate the few-chip-on-the-left cookies (where the right quarters have tons of chips), the moderate-chip-on-the-left cookies and the many-chip-on-the-left cookies will have fewer than average chips on the right.

The math in this paper has nothing to do with sports. Here's the universe in which it would: every player shoots his average (within one shot) in every single game he plays. So if you knew that Bo was going to shoot 40%, and he has made 5 shots out of 13 tries (38%), and there's 20 seconds on the clock, tell him to shoot! He will make it. But after that he has made 6 of 14 (43%). He must not shoot again. The next one will rim out.

For random processes, there is no hot hand. For players, with focus, conditioning, mental issues, sure. This paper, though, applies to neither.

[UNIdad]

Re: WSJ: Write-up of academic study of the "Hot Hand"

In a way, yes. Mathematically, it is like a more complicated version of the Monty-Hall paradox.

And it has nothing to do with a "hot hand" or sports. And it's about as ground-breaking as me publishing that the sky is blue.

Imagine that I make some cookies and I make sure that every cookie has 12 chocolate chips.
How many will each quarter cookie average?
3, but it might be: (0,5,3,4) or (0,0,6,6) or (1,4,5,2) and so on.

Now- what if I say, "Do the cookies have a 'hot-hand'?"
So I take some of these cookies and I pick the left two quarter cookies and count the chips.
Only in cases where each of the two left quarters have at least 2 chips, I count the chips in the right two quarters. Guess, what? The cookies have a "cold hand"! Duh! Of course they do. If you eliminate the few-chip-on-the-left cookies (where the right quarters have tons of chips), the moderate-chip-on-the-left cookies and the many-chip-on-the-left cookies will have fewer than average chips on the right.

The math in this paper has nothing to do with sports. Here's the universe in which it would: every player shoots his average (within one shot) in every single game he plays. So if you knew that Bo was going to shoot 40%, and he has made 5 shots out of 13 tries (38%), and there's 20 seconds on the clock, tell him to shoot! He will make it. But after that he has made 6 of 14 (43%). He must not shoot again. The next one will rim out.

For random processes, there is no hot hand. For players, with focus, conditioning, mental issues, sure. This paper, though, applies to neither.

Here is the extent of my math. Cookies don't make their own chocolate chips!