Shooting Streaks: Do they exist?
- Thread starter Siro
- Start date
fishonjazz
Well-Known Member
Contributor
2018 Award Winner
2019 Award Winner
20-21 Award Winner
2022 Award Winner
2023 Award Winner
2024 Award Winner
2025 Award Winner
This
I don't think I'm NAOS, but I agree that you get the last word on what does and doesn't exist.
Beer
Well-Known Member
Siro, sorry I not red letters good. Not sure where NAOS came from.
Maybe. But you would not be the first person fooled by randomness
Ellipse
Well-Known Member
That was from a study on whether people believed in the hot hand hypotheses; however, here is a quote from your reference: "Recent studies using modern statistical analysis show there is evidence for the "hot hand" in some sporting activities."
The study finds that previous studies debunking the "hot hand theory" were wrong. Here is the abstract from the study:
Joshua Miller
Bocconi University - Department of Decision Sciences; IGIER - Innocenzo Gasparini Institute for Economic Research
Adam Sanjurjo
Universidad de Alicante - Departamento de Fundamentos del Análisis Económico
Date Written: November 15, 2016
Abstract
We prove that a subtle but substantial bias exists in a standard measure of the conditional dependence of present outcomes on streaks of past outcomes in sequential data. The magnitude of this novel form of selection bias generally decreases as the sequence gets longer, but increases in streak length, and remains substantial for a range of sequence lengths often used in empirical work. The bias has important implications for the literature that investigates incorrect beliefs in sequential decision making---most notably the Hot Hand Fallacy and the Gambler's Fallacy. Upon correcting for the bias, the conclusions of prominent studies in the hot hand fallacy literature are reversed. The bias also provides a novel structural explanation for how belief in the law of small numbers can persist in the face of experience.
D
Deleted member 5000
Guest
EPIC
What did you use to go grab the data to populate your tables, is there an opensoure database out there people can query?
What did you use to go grab the data to populate your tables, is there an opensoure database out there people can query?
GVC
Well-Known Member
Haven't looked through the thread, but stats.nba.com has an API. Grabbing all sorts of NBA data is incredibly easy.
I can probably dig up some old python code for anyone who's interested.
checkbookdoc
Well-Known Member
I am a little unclear on what is meant by the hot hand hypothesis. Some of my questions are listed below:
1) Does the hot hand have to do with a streak, or with an abnormal shooting performance in the game up to that point?
2) Does the hot hand hypothesis apply only to makes or also to misses?
3) Does the hypothesis predict a gradual increase in hotness, or does the hypothesis only apply once a certain level is attained? Relatedly, if it is gradual, is the increase expected to be linear?
4) Does the hypothesis differentiate between the cause of the hotness?
I think the existence of the hot hand, and the appropriate way to identify it, depends on the answer to these questions.
1) Does the hot hand have to do with a streak, or with an abnormal shooting performance in the game up to that point?
For example: Assume player X is typically a 35% three point shooter. Now assume that through 3 quarters of a game, player X has hit every-other their 3 point shot taken (i.e., they are shooting 50% without ever having a streak of makes). Would player X have a hot hand?
I would argue that player X is hot, meaning that a streak is not necessary for the hot hand. However, the data analysis I have seen has always focused on streaks.
I would argue that player X is hot, meaning that a streak is not necessary for the hot hand. However, the data analysis I have seen has always focused on streaks.
2) Does the hot hand hypothesis apply only to makes or also to misses?
In other words, does it also predict a cold hand? If it also predicts a cold hand, then it really is just a momentum prediction.
3) Does the hypothesis predict a gradual increase in hotness, or does the hypothesis only apply once a certain level is attained? Relatedly, if it is gradual, is the increase expected to be linear?
4) Does the hypothesis differentiate between the cause of the hotness?
For example, is it behavioral (e.g., the player is extra focused, or has an increased belief in themselves), is it based on physical conditions (e.g., sleep/rest), is it based on the actions of defenses?
I think the existence of the hot hand, and the appropriate way to identify it, depends on the answer to these questions.
sahlensguy
Well-Known Member
Even shooters with a 'hot hand" miss more often than not.
checkbookdoc
Well-Known Member
It depends on what you mean by hot hand (see my post above).
sahlensguy
Well-Known Member
We will have to wait for the op to clarify.
My hypothesis was that hitting any number of shots in a row does not improve your chances on the following shot. You'll still hit your percentages on any shot in a large sample regardless of whether you hit or missed the previous few shots. That's how I'm defining it.
If the hot hand is real, then after hitting, say, 5 shots, the 6th should go in at a higher rate than your normal fg%. This should show up regardless of any complexities because we're averaging out all the times you've hit that many shots, ideally.
If the hot hand is real, then after hitting, say, 5 shots, the 6th should go in at a higher rate than your normal fg%. This should show up regardless of any complexities because we're averaging out all the times you've hit that many shots, ideally.
But it's shown most players take harder shots and become more closely guarded once they hit multiple shots in a row. I think those "complexities" would be significant. That's why I like the idea of doing the study off the 3 point contest, because there arent as many variables.
sahlensguy
Well-Known Member
What is one's "normal fg%"? A players' career % ? Their latest seasonal % ? Their latest monthly or weekly % ?
All players have peaks and valleys. Hot hands usually don't happen in the valleys, so why guage the probability of the chances of a hot shooters' 'next shot' going in, based o %s that include those valleys and or early career?
I understand. It's a good point, like some others made. I also love the idea of studying the three point contest and such.
But I'm not sure I'm sold on that explanation. There is no point where any number of made shots improves the next one. You'd still see improvement after 2 or 3 shots, before defenses hone in. Or maybe not. I'm not claiming that the hot hand has been conclusively disproven. Just that there are good reasons to doubt it.
But if we're measuring streaks, what does it matter if they happen in a peak or a trough? Also, I'm not sure more constrains on which FG% matters. As long as you're being consistent, the fg% after hitting several shots should be better than after missing a bunch of shots.