Comments on: Ben Gordon and the ‘hot hand’ http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/ Your Go-To Source For Pistons Coverage Wed, 22 May 2013 02:58:01 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-61132 Max Mon, 26 Mar 2012 17:57:07 +0000 http://www.pistonpowered.com/?p=9527#comment-61132 I've enjoyed this debate as well.  One thing about players "deciding" to shoot again after they hit a shot because as Abbot says they are over confident---he is totally ignoring the idea that the coach became over confident and called for the player to shoot the ball.   He is also totally ignoring the notion that a player's teammates can think a player is hot and start deferring to him.  Either case can lead to teammates forcing the ball to a player in a bad position from which he has to force up a shot. I’ve enjoyed this debate as well.  One thing about players “deciding” to shoot again after they hit a shot because as Abbot says they are over confident—he is totally ignoring the idea that the coach became over confident and called for the player to shoot the ball.   He is also totally ignoring the notion that a player’s teammates can think a player is hot and start deferring to him.  Either case can lead to teammates forcing the ball to a player in a bad position from which he has to force up a shot.

]]> By: tarsier http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60998 tarsier Sun, 25 Mar 2012 22:40:36 +0000 http://www.pistonpowered.com/?p=9527#comment-60998 But I have enjoyed this back and forth with you. Particularly because of your limited use of anecdotal evidence. You continually make good points. But it is just very difficult to convince me stats are not the answer unless there is either a limited sample size or extenuating factors that could actually compromise the results instead of merely toning them down. I guess the defender trying harder could fit that description. But if being on fire can be entirely counter-acted by the additional effort it inspires in an opponent, it may be a real thing, but it's not a very useful thing. But I have enjoyed this back and forth with you. Particularly because of your limited use of anecdotal evidence. You continually make good points. But it is just very difficult to convince me stats are not the answer unless there is either a limited sample size or extenuating factors that could actually compromise the results instead of merely toning them down.

I guess the defender trying harder could fit that description. But if being on fire can be entirely counter-acted by the additional effort it inspires in an opponent, it may be a real thing, but it’s not a very useful thing.

]]> By: tarsier http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60996 tarsier Sun, 25 Mar 2012 22:34:58 +0000 http://www.pistonpowered.com/?p=9527#comment-60996 Yeah, studies vary. And I do hate the definition this one uses. I don't remember where I have before read one, but it basically compared how well a guy shot after hitting his last shot, his last two, his last three, etc. And the point was that it wasn't rising at all. Now all the things you say are true. And when I watch a game, I definitely sometimes get suckered into the notion that someone is really "feeling it." But I guarantee you, you would feel the same way if you were playing poker and won a couple hands in a row. Hence why it is called gambler's fallacy. The point is that feeling like you are hot doesn't make it so. Given things like increased confidence and whatnot, I would totally believe that people did get "hot hands" if a study showed it reasonably well. And it is totally possible that one will at some point. Honestly, I think this issued has not been analyzed nearly enough by people who are paid to be stats junkies. But by and large, I do not trust humans' feelings or perceptions when it comes to matters of probability. Any anthropologist, mathematician, or logician will tell you that is one of our weakest areas. People instinctively think pretty much only categorically in impossible, possible but unlikely, 50/50ish, likely, or definite. That's why people do things like buying lottery tickets and are more afraid of planes than cars or lightning than fast food. We aren't good at dealing with odds. So why would I expect us to be good at perceiving the likelihood of hitting a basketball shot? No, I'll concede that these studies tend to be weak and could therefore easily be wrong. But I still think that the best way to get an answer then is a better statistical study. It still won't be perfect, but at least it has limited bias and relies on a dearth of facts rather than a handful of anecdotal instances. Yeah, studies vary. And I do hate the definition this one uses. I don’t remember where I have before read one, but it basically compared how well a guy shot after hitting his last shot, his last two, his last three, etc. And the point was that it wasn’t rising at all. Now all the things you say are true. And when I watch a game, I definitely sometimes get suckered into the notion that someone is really “feeling it.” But I guarantee you, you would feel the same way if you were playing poker and won a couple hands in a row. Hence why it is called gambler’s fallacy.

The point is that feeling like you are hot doesn’t make it so. Given things like increased confidence and whatnot, I would totally believe that people did get “hot hands” if a study showed it reasonably well. And it is totally possible that one will at some point. Honestly, I think this issued has not been analyzed nearly enough by people who are paid to be stats junkies.

But by and large, I do not trust humans’ feelings or perceptions when it comes to matters of probability. Any anthropologist, mathematician, or logician will tell you that is one of our weakest areas. People instinctively think pretty much only categorically in impossible, possible but unlikely, 50/50ish, likely, or definite. That’s why people do things like buying lottery tickets and are more afraid of planes than cars or lightning than fast food. We aren’t good at dealing with odds.

So why would I expect us to be good at perceiving the likelihood of hitting a basketball shot? No, I’ll concede that these studies tend to be weak and could therefore easily be wrong. But I still think that the best way to get an answer then is a better statistical study. It still won’t be perfect, but at least it has limited bias and relies on a dearth of facts rather than a handful of anecdotal instances.

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60984 Max Sun, 25 Mar 2012 21:37:43 +0000 http://www.pistonpowered.com/?p=9527#comment-60984 BTW: The article link in this thread's post regarding the the study on hot hands seems to be defining it in an incredibly stupid way and is asking questions about what happens after 1-1s instead of the 4-4s we've been using to hypothesize. BTW: The article link in this thread’s post regarding the the study on hot hands seems to be defining it in an incredibly stupid way and is asking questions about what happens after 1-1s instead of the 4-4s we’ve been using to hypothesize.

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60971 Max Sun, 25 Mar 2012 19:50:38 +0000 http://www.pistonpowered.com/?p=9527#comment-60971 Also, I think all of your arguments would apply equally to the idea that players don't ever have "cold hands" when it is clear to me that players go through slumps.   The same logic you use would apply though since the stats assume the resumption of a player's normal shooting pct.  This is why I repeatedly cite how stats fail to account for context.  The stats are blind to when the hot hand or cold hand will go in the other direction but because the data is created by humans performing tasks for which true consistency is impossible, the base line average pct always resumes unless the player is truly getting much better or worse in the moment.   My interpretation of the data saying the hot hand doesn't exist being wrong comes down to thinking "on fire" is very poorly defined and that stats ignore context while assuming a resumption or normality whenever there are temporary outliers.  The trouble is that the data doesn't say anything meaningful about the periods between a player playing as he usually does other than that he is most likely to resume playing as he normal does during the next sample size.  It may be true, but it is not revealing and it does not disprove the "hot hand". Also, I think all of your arguments would apply equally to the idea that players don’t ever have “cold hands” when it is clear to me that players go through slumps.   The same logic you use would apply though since the stats assume the resumption of a player’s normal shooting pct.  This is why I repeatedly cite how stats fail to account for context.  The stats are blind to when the hot hand or cold hand will go in the other direction but because the data is created by humans performing tasks for which true consistency is impossible, the base line average pct always resumes unless the player is truly getting much better or worse in the moment.   My interpretation of the data saying the hot hand doesn’t exist being wrong comes down to thinking “on fire” is very poorly defined and that stats ignore context while assuming a resumption or normality whenever there are temporary outliers.  The trouble is that the data doesn’t say anything meaningful about the periods between a player playing as he usually does other than that he is most likely to resume playing as he normal does during the next sample size.  It may be true, but it is not revealing and it does not disprove the “hot hand”.

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60968 Max Sun, 25 Mar 2012 19:31:45 +0000 http://www.pistonpowered.com/?p=9527#comment-60968 You don't think Isiah was on fire when he scored 25 in the third quarter of a finals game while limping and hopping on one leg?   You don't think he had more adrenalin than was usual for him?  You don't think he turned his pain into focus and energy? Anyway, another fault of the data is that it does not allow for the idea that when a player hits a shot and then hits another shot and so on, the collective defense will pay increasing attention to him and his individual defender will try harder. As for the establishment of difference; I thought I addressed that when I said that the only absolute was that a any player will eventually miss if he keeps shooting and I would add that every period of being on fire comes to an end.  As raw data on field goal pct fails to provide context, it ignores everything that happens between shots as in trips to the bench and 2-40 minute stretches between shots. My overriding point is that most of the time when a player goes 4-4 it is circumstantial and not due to their being on fire.   Being on fire should apply to games like Kobe putting up 82 or LBJ scoring 25 straight points in the 4th quarter and overtime and not 4 times a hustling player got a layup due to other players receiving attention and even not when they just go 4-4 without forcing a shot in the normal flow whether the shots are 3 pointers or a layup. BTW: @Tarsier....no offense, but haven't you ever done a much better job for the girl in bed than usual?  Weren't you going through a peak experience where you were "on fire"?   The trouble with pure stats is that it doesn't account for conscientiousness and mood which do influence how well one performs. You don’t think Isiah was on fire when he scored 25 in the third quarter of a finals game while limping and hopping on one leg?   You don’t think he had more adrenalin than was usual for him?  You don’t think he turned his pain into focus and energy?
Anyway, another fault of the data is that it does not allow for the idea that when a player hits a shot and then hits another shot and so on, the collective defense will pay increasing attention to him and his individual defender will try harder.
As for the establishment of difference; I thought I addressed that when I said that the only absolute was that a any player will eventually miss if he keeps shooting and I would add that every period of being on fire comes to an end.  As raw data on field goal pct fails to provide context, it ignores everything that happens between shots as in trips to the bench and 2-40 minute stretches between shots.
My overriding point is that most of the time when a player goes 4-4 it is circumstantial and not due to their being on fire.   Being on fire should apply to games like Kobe putting up 82 or LBJ scoring 25 straight points in the 4th quarter and overtime and not 4 times a hustling player got a layup due to other players receiving attention and even not when they just go 4-4 without forcing a shot in the normal flow whether the shots are 3 pointers or a layup.
BTW: @Tarsier….no offense, but haven’t you ever done a much better job for the girl in bed than usual?  Weren’t you going through a peak experience where you were “on fire”?   The trouble with pure stats is that it doesn’t account for conscientiousness and mood which do influence how well one performs.

]]> By: tarsier http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60934 tarsier Sun, 25 Mar 2012 13:16:33 +0000 http://www.pistonpowered.com/?p=9527#comment-60934 No, because the entire point was that even if it mislabeled them 80% of the time (the very most I could see it doing), it would still pick up a difference if there was one. How did you not pick up that that was what I was saying. The point was that unless the definition does not at all or very barely correlates with being "on fire", it doesn't matter that there is a ton of error because the sample size is large. In other words, incorrectly labeling a guy 80 percent of the time would not invalidate it. That was the entire point of the comment. And yet what you took out of it was "perhaps 80% of the time it mislabels so it is completely untrustworthy." It would be like if I said even though Lou Williams scores the most, Iguodala is the 76ers most valuable player and you took it to mean because Lou Williams scores the most, he is the 76ers most valuable player. And no, I don't doubt in the slightest that adrenaline increases performance. But I bet every player on the floor has a good bit of it coursing through their veins, And probably no more in those "on fire" than those not. As for my question about establishing a difference, my point was that studies have shown no difference in shooting percentage after hitting the last several shots. You are saying many of those could be explained away by the dude isn't actually feeling it or faced a double team or something. But those extenuating circumstances also exist for the people not labeled as "on fire." So as long as the labeled group is a somewhat higher percentage of guys legitimately "on fire" as you call it (I still maintain the term is a nonentity), then that group should be hitting a higher percentage. All these other factors will only reduce how much higher that percentage is. No, because the entire point was that even if it mislabeled them 80% of the time (the very most I could see it doing), it would still pick up a difference if there was one. How did you not pick up that that was what I was saying. The point was that unless the definition does not at all or very barely correlates with being “on fire”, it doesn’t matter that there is a ton of error because the sample size is large. In other words, incorrectly labeling a guy 80 percent of the time would not invalidate it. That was the entire point of the comment.

And yet what you took out of it was “perhaps 80% of the time it mislabels so it is completely untrustworthy.” It would be like if I said even though Lou Williams scores the most, Iguodala is the 76ers most valuable player and you took it to mean because Lou Williams scores the most, he is the 76ers most valuable player.

And no, I don’t doubt in the slightest that adrenaline increases performance. But I bet every player on the floor has a good bit of it coursing through their veins, And probably no more in those “on fire” than those not.

As for my question about establishing a difference, my point was that studies have shown no difference in shooting percentage after hitting the last several shots. You are saying many of those could be explained away by the dude isn’t actually feeling it or faced a double team or something. But those extenuating circumstances also exist for the people not labeled as “on fire.” So as long as the labeled group is a somewhat higher percentage of guys legitimately “on fire” as you call it (I still maintain the term is a nonentity), then that group should be hitting a higher percentage. All these other factors will only reduce how much higher that percentage is.

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60899 Max Sun, 25 Mar 2012 04:21:47 +0000 http://www.pistonpowered.com/?p=9527#comment-60899 I do think what you wrote "if the definition picks up on guys who are on fire 20 percent of the time" helps make my point.  It would create a situation where the stats were incorrectly labeling a guy as on fire 80 percent of the time and thus invalidate their findings. I do think what you wrote “if the definition picks up on guys who are on fire 20 percent of the time” helps make my point.  It would create a situation where the stats were incorrectly labeling a guy as on fire 80 percent of the time and thus invalidate their findings.

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60898 Max Sun, 25 Mar 2012 04:16:01 +0000 http://www.pistonpowered.com/?p=9527#comment-60898 @Tarsier....I don't understand the question you are asking me about establishing a difference.... @Tarsier….I don’t understand the question you are asking me about establishing a difference….

]]> By: Max http://www.pistonpowered.com/2012/03/ben-gordon-and-the-hot-hand/comment-page-1/#comment-60897 Max Sun, 25 Mar 2012 04:12:56 +0000 http://www.pistonpowered.com/?p=9527#comment-60897 Do you seriously believe that an increase in adrenalin does not enhance performance? Do you seriously believe that an increase in adrenalin does not enhance performance?

]]>