What's new
By:

Shooting Streaks: Do they exist?

Siro said:
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.
Click to expand...

Hitting a bunch of shots or even having a career year increases a players fg%.

You said "My hypothesis was that hitting any number of shots in a row does not improve your chances on the following shot."

If this is true than a players chance of hitting his next shot is directly related to his shooting percentage. A players shooting percentage differs from year to year though and month to month. So when you say " a players normal shooting %", what do you mean by "normal"?
 
Plotting rolling 8 or 10-shot averages for players known as "scorers" should show if there is a trend. This ought to be easy to do for someone with access to the data.
 
Siro said:
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.
Click to expand...
This may depend on how you define 'normal field goal percentage.' If you remove shooting streaks from the total population of shots, the players field goal percentage will be lower than their overall average field goal percentage.
 
Siro said:
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.
Click to expand...
Except that by construction, the total field goal percentage on all remaining shots will be lower after removing series of makes than after removing series of misses.
 
Last edited:
Siro said:
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.
Click to expand...

Did you account for probability, 2 vs 3, and maybe somehow even missed shots leading to fouls that aren't recorded? Let's say a 2 point shooter hits .5, so the chances of making 3 in a row is .125. Now, it's obvious that the 3rd shot still has a statistical chance of .5 and not .125, but in games where someone goes on a streak did they hit that 3rd shot at a rate higher or lower than .125? Obviously they did.

So now we can establish a delta - how many times did a player shoot above and below their 3 in a row probability, then plot their standard deviations. I doubt if you did a bell curve of these standard deviations and calculated that standard deviation that players like Rodney Hood wouldn't be far from the middle. I'm almost certain you wouldn't find a 3 off deviation unless maybe you were plotting Shaq's free throws, but I'd bet you could find something approaching 2. I say those 2's are evidence of streak shooters/hot hand.
 
franklin said:
Did you account for probability, 2 vs 3, and maybe somehow even missed shots leading to fouls that aren't recorded? Let's say a 2 point shooter hits .5, so the chances of making 3 in a row is .125. Now, it's obvious that the 3rd shot still has a statistical chance of .5 and not .125, but in games where someone goes on a streak did they hit that 3rd shot at a rate higher or lower than .125? Obviously they did.

So now we can establish a delta - how many times did a player shoot above and below their 3 in a row probability, then plot their standard deviations. I doubt if you did a bell curve of these standard deviations and calculated that standard deviation that players like Rodney Hood wouldn't be far from the middle. I'm almost certain you wouldn't find a 3 off deviation unless maybe you were plotting Shaq's free throws, but I'd bet you could find something approaching 2. I say those 2's are evidence of streak shooters/hot hand.
Click to expand...

I didn't really account for anything. Three years ago, the question about hot streaks occurred to me one evening, so I spent a couple of hours investigating it and then posted the results (it had graphs and ****). It is by no means a serious attempt at providing an answer that can withstand serious scrutiny. The code simply compares the FG% of the whole sample, to the probability of a shot hitting after 2, 3, or 4 previous hits.

During the discussion then, and today, it turns out that some analysis has been done on the subject. Earlier attempts went with similar approaches to mine, and showed similar results. It was generally considered that the hot streak phenomena is a myth. But apparently, there have been newer and more sophisticated analysis, that suggests the opposite. I'm going to have a bit about this to see where the thinking is on this.
 
GVC said:
Except that by construction, the total field goal percentage on all remaining shots will be lower after removing series of makes than after removing series of misses.
Click to expand...

Nothing is removed. Like I explained above, it's just the total percentage of makes in the whole sample, compared to the percentage of a 3rd, a 4th, and a 5th straight hit.
 
Siro said:
I didn't really account for anything. Three years ago, the question about hot streaks occurred to me one evening, so I spent a couple of hours investigating it and then posted the results (it had graphs and ****). It is by no means a serious attempt at providing an answer that can withstand serious scrutiny. The code simply compares the FG% of the whole sample, to the probability of a shot hitting after 2, 3, or 4 previous hits.

During the discussion then, and today, it turns out that some analysis has been done on the subject. Earlier attempts went with similar approaches to mine, and showed similar results. It was generally considered that the hot streak phenomena is a myth. But apparently, there have been newer and more sophisticated analysis, that suggests the opposite. I'm going to have a bit about this to see where the thinking is on this.
Click to expand...

I just don't think that you can statistically prove this either way. You can only disprove it with stats but human nature and the eyeball test say otherwise. Coaching says otherwise when they tell 3 point shooters to shoot of the hop. Rhythm and momentum are things that we can't quantify.
 
franklin said:
I just don't think that you can statistically prove this either way. You can only disprove it with stats but human nature and the eyeball test say otherwise. Coaching says otherwise when they tell 3 point shooters to shoot of the hop. Rhythm and momentum are things that we can't quantify.
Click to expand...

I don't agree. The world is full of things that people swear are so, but under examination they are clearly not so. If the phenomenon is true and has an effect on the real world, then that effect can be quantified. I don't see how it can be otherwise.
 
Siro said:
I don't agree. The world is full of things that people swear are so, but under examination they are clearly not so. If the phenomenon is true and has an effect on the real world, then that effect can be quantified. I don't see how it can be otherwise.
Click to expand...

Can we bring Dutch back to comment "Settled Science"? ;)

You damn well know that some things cannot be quantified because you cannot always have a control group. Your question here lies in the realm of "proof of impossibility" and proving non-existence. Sometimes the eyeball test tells more than we can quantify.

Run my numbers. I gave you a simple formula that might possibly quantify hot hand but at minimum give a value to streaky shooting.
 
sahlensguy said:
The Hot Hand fallacy.

https://en.m.wikipedia.org/wiki/Hot-hand_fallacy

Data does not support the hot hand hypotheses exept at the free throw line.
Click to expand...


That's just it. The FT line, which, in a live game, has the least number of variables to show that in fact a hot hand can exist. During regular play, there are too many variables that come into play. I hit a few shots, my defender guards me closer, goes over the screen, etc. I hit four in a row, I might take a more difficult "heat check" shot. If I miss, it does not mean I don't have a hot hand.
 
JAZZGASM said:
That's just it. The FT line, which, in a live game, has the least number of variables to show that in fact a hot hand can exist. During regular play, there are too many variables that come into play. I hit a few shots, my defender guards me closer, goes over the screen, etc. I hit four in a row, I might take a more difficult "heat check" shot. If I miss, it does not mean I don't have a hot hand.
Click to expand...

I thnk shooting streaks are real but it's almost fruitless to discuss. Has a distinction between between a hot hand and a streaky shooter even been established? Rodney Hood was a streaky shooter for us. If a hot hand and a streaky shooter are one in the same, hot hands absolutely exist.

Ricky Rubio hit on like 52 consecutive TFTs that ended earlier this season. There comes a point where people take notice and then there is pressure just to keep the streak alive.

Joe DiMaggio had a streak of hitting in 56 consecutive games in 1941 while hitting .408. At some point the pressure is too great to continue the streak. This is a variable different in every athlete, but until that breaking point, they absolutely are "hot".
 
You must log in or register to reply here.

Share this page
Back
Top