Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by rossdavidh on Saturday November 25, 2017 @08:08PM
(#15778913)
The hot hand "fallacy" has actually been under a bit of a revisionist assault lately. An example (from baseball): https://fivethirtyeight.com/features/baseballs-hot-hand-is-r...
TLDR: there is a hot hand (in baseball pitching, anyway), but you have to be very, very good at statistics to see that, or else just a big baseball fan. If you have just pretty good statistics, and you don't have a baseball fanatic's intuition, you can (erroneously) conclude that there is no (pitcher's) hot hand (in baseball).
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by fossuser on Saturday November 25, 2017 @08:48PM
(#15779048)
I’ve always suspected this fallacy in terms of sports was a bad example - since humans doing certain motor/coordination actions and getting feedback on that success seemed likely to stack.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by cirgue on Saturday November 25, 2017 @10:05PM
(#15779245)
Another point of view is that this makes sports a great example: we believe the 'hot-hand' phenomenon intuitively, but the statistics tell a complicated and seemingly contradictory story.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by magpi3 on Saturday November 25, 2017 @05:44PM
(#15778401)
Interesting paper, but as a former basketball but, I disagree with the premise that free throws have anything to do with a "hot hand."
To me a hot hand is only something that can exist in the flow of the game. It is not about making shots. It is about making contested shots and quick decisions. A player with a hot hand is one that is feeling especially focused and confident for a specific period of time. Ideally a player tries to hold onto their focus and confidence for the entire game, but unless they have a Buddha-like nature that is just not possible.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by MisterTea on Sunday November 26, 2017 @01:13AM
(#15779666)
I completely agree. Reminds me of my hot streaks when playing FPS games, specifically when I played TF2. Normally I'm a mediocre engineer, pyro or solder. But something about playing the spy gave me a rush at times where I could rack up well over a hundred kills on the map. And there were a few times I topped 200 in a map. But once I lost focus, I crashed hard and I would stop playing for the day.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by glaugh on Saturday November 25, 2017 @06:35PM
(#15778590)
The original paper only covered free throws as one of three examples of the hot hand fallacy. The other examples were of shooting in the normal run of play.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by currymj on Saturday November 25, 2017 @07:52PM
(#15778856)
more recent statistical analyses have found evidence for a hot hand effect, though.
i also don't necessarily like calling it a fallacy, as i've definitely run into some people who misinterpret that word, basically assuming that there can never be autocorrelation in data, and anyone who thinks there is must be falling prey to a logical fallacy.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by thecupisblue on Sunday November 26, 2017 @06:05AM
(#15780359)
I kind of disagree.
In the flow of the game, it's hard for hot hand to exist and what you describe is `the zone`. But in a free throws game, there surely is a hot hand. I always get the first few hit/miss, but when I finally feel calibrated with my stance and my hand `whip` and get a hit I like, then the next one is probably a hit again and after that I can streak for at least 4-5 shots.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by harry8 on Saturday November 25, 2017 @10:10PM
(#15779255)
Means nothing. Get a thousand people to toss a coin 20 times in a row. Repeat 82 times each. Someone probably throws 20 heads in a row. Monte Carlo sim it.
That model I've just described does not disprove the hot hand. It does show that an occasion of a good shooting run also is not evidence that it exists.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by Bromskloss on Saturday November 25, 2017 @09:30PM
(#15779152)
By the way, is there anywhere one can bet on the outcome of a game while the game is being played? It would be awesome to see the odds change in response to something that happened in the game that second, just like how a stock market reacts to news about the business.
Re:Simpson's Paradox and the Hot Hand in Basketball (1995)
by bayonetz on Saturday November 25, 2017 @10:04PM
(#15779244)
Yeah. Any online sports betting site. Football for example, you can bet realtime on things like halftime score, final score, etc. as the game progresses. Even cooler is you can bet on the next play. Like "next play will result in a first down" or "next play they will run the ball".
I wonder if it holds in other contexts.
To me a hot hand is only something that can exist in the flow of the game. It is not about making shots. It is about making contested shots and quick decisions. A player with a hot hand is one that is feeling especially focused and confident for a specific period of time. Ideally a player tries to hold onto their focus and confidence for the entire game, but unless they have a Buddha-like nature that is just not possible.
http://node101.psych.cornell.edu/sec/pubPeople/tdg1/Gilo.Val...
i also don't necessarily like calling it a fallacy, as i've definitely run into some people who misinterpret that word, basically assuming that there can never be autocorrelation in data, and anyone who thinks there is must be falling prey to a logical fallacy.
In the flow of the game, it's hard for hot hand to exist and what you describe is `the zone`. But in a free throws game, there surely is a hot hand. I always get the first few hit/miss, but when I finally feel calibrated with my stance and my hand `whip` and get a hit I like, then the next one is probably a hit again and after that I can streak for at least 4-5 shots.
Yeah, like Kay Thompson dropping 37 points in 1 quarter and not missing any shoots. Even himself seems confused about what’s going on.
https://www.youtube.com/watch?v=daP4JysFcUg
That model I've just described does not disprove the hot hand. It does show that an occasion of a good shooting run also is not evidence that it exists.
It's a really interesting problem to analyse.