Woj: Jrue to Celtics

[keevsnick]

Another excellent game by Jrue.

Yeah, who knew White was going to fall off a cliff? We've really needed Jrue's steady hand these past couple of games. Dude decided to step up at exactly the right time.

Didn't everyone know White was going to fall off a cliff? I mean White was always going to regress to the mean.  That is how those things work.

There’s a space between clear All-Star and 2022 playoffs disaster. We thought he had moved past the latter. Hopefully it really is just a mini-slump and not indicative of something greater. We are who we are this season in large part to White elevating his game.

But luckily Jrue suddenly became a much better player just as White fell off. That has been huge. If both of them just ‘regress to the mean’, then we will be fine, as well

well when you play well above your level, you are invariably going to play well below it at some point, so that you end up where you should be

No, that isn't how probabilities or regression to the mean work. There isn't some cosmic force that will make you play the perfect level of bad to instantly balance out the good, it's far, far more likely to have a bunch slightly below average games than a couple of real stinkers.

But even that isn't guaranteed, players have (and will continue to) go on unreasonably long hot streaks without suddenly having equally long/extreme cold streaks. And ditto for cold streaks like Al's not meaning that he's going to start shooting 57% on threes in the next 2 rounds.

Sure, they don't HAVE TO have unreasonably cold streaks after a hot streak. But they sometimes will. Shooting game to game is basically a normal distribution around your true average. most games you'll be relatively close to your average but for stretches you'll be crazy good or crazy bad, the tails of the distribution.

Following a crazy outlier shooting stretch one direction with a crazy outlier shooting stretch the other will happen sometimes. This is one of those times.

[Vermont Green]

When I say "regression to the mean" I am probably not using it in the formal statistical manner.  It is more general than that.  You just need to watch the player play.  We have all watched White play for about 2.5 seasons plus playoffs.  We all know he is not going to continue to shoot like that.  He has shot about 38%-40% from 3 over the last two seasons.  Of course he is not going to continue to light it up.

That doesn't necessarily mean he will have a bad stretch but to average say 39% from 3 means that you have stretches that are better and an equal amount that are worse.  That is just how an average or a mean works.

[Moranis]

Another excellent game by Jrue.

Yeah, who knew White was going to fall off a cliff? We've really needed Jrue's steady hand these past couple of games. Dude decided to step up at exactly the right time.

Didn't everyone know White was going to fall off a cliff? I mean White was always going to regress to the mean.  That is how those things work.

There’s a space between clear All-Star and 2022 playoffs disaster. We thought he had moved past the latter. Hopefully it really is just a mini-slump and not indicative of something greater. We are who we are this season in large part to White elevating his game.

But luckily Jrue suddenly became a much better player just as White fell off. That has been huge. If both of them just ‘regress to the mean’, then we will be fine, as well

well when you play well above your level, you are invariably going to play well below it at some point, so that you end up where you should be

No, that isn't how probabilities or regression to the mean work. There isn't some cosmic force that will make you play the perfect level of bad to instantly balance out the good, it's far, far more likely to have a bunch slightly below average games than a couple of real stinkers.

But even that isn't guaranteed, players have (and will continue to) go on unreasonably long hot streaks without suddenly having equally long/extreme cold streaks. And ditto for cold streaks like Al's not meaning that he's going to start shooting 57% on threes in the next 2 rounds.

Sure, they don't HAVE TO have unreasonably cold streaks after a hot streak. But they sometimes will. Shooting game to game is basically a normal distribution around your true average. most games you'll be relatively close to your average but for stretches you'll be crazy good or crazy bad, the tails of the distribution.

Following a crazy outlier shooting stretch one direction with a crazy outlier shooting stretch the other will happen sometimes. This is one of those times.

I'm not even sure that is true. Take Curry in his best 3 point shooting season. He shot 45.4%, but here are the number of games and the percentage he shotnin them

60% or better - 11 games
> 50% to 60% - 10 games
Exactly 50% - 13 games
> 45.4 to 50%  - 5 games
40-45.4% - 14 games
30-39.9% - 13 games
< 30% - 11 games

In other words, the greatest shooter ever in his best season was all over the map.  And basically had as many downright terrible games as he did amazingly awesome games.

[Vermont Green]

I'm not even sure that is true. Take Curry in his best 3 point shooting season. He shot 45.4%, but here are the number of games and the percentage he shot in them

60% or better - 11 games
> 50% to 60% - 10 games
Exactly 50% - 13 games
> 45.4 to 50%  - 5 games
40-45.4% - 14 games
30-39.9% - 13 games
< 30% - 11 games

In other words, the greatest shooter ever in his best season was all over the map.  And basically had as many downright terrible games as he did amazingly awesome games.

So pretty much a classic bell curve if you include the "exactly 50% in with the 45.4%-50%.  Not sure how you decided on the bands, but if you look at 40% to 50%, that is 22 games.  Then it tails off from there in both directions.  There is nothing surprising here.

White had a few ridiculously hot games.  He isn't going to keep that up for long.  Holiday had a few very flat games.  That isn't going to last either.  It is kind of like saying "you are what your record says your are".  If you have shot 39% over two full seasons, that is what you are.  Unless age or injury or somehting causes a step change.

[keevsnick]

Another excellent game by Jrue.

Yeah, who knew White was going to fall off a cliff? We've really needed Jrue's steady hand these past couple of games. Dude decided to step up at exactly the right time.

Didn't everyone know White was going to fall off a cliff? I mean White was always going to regress to the mean.  That is how those things work.

There’s a space between clear All-Star and 2022 playoffs disaster. We thought he had moved past the latter. Hopefully it really is just a mini-slump and not indicative of something greater. We are who we are this season in large part to White elevating his game.

But luckily Jrue suddenly became a much better player just as White fell off. That has been huge. If both of them just ‘regress to the mean’, then we will be fine, as well

well when you play well above your level, you are invariably going to play well below it at some point, so that you end up where you should be

No, that isn't how probabilities or regression to the mean work. There isn't some cosmic force that will make you play the perfect level of bad to instantly balance out the good, it's far, far more likely to have a bunch slightly below average games than a couple of real stinkers.

But even that isn't guaranteed, players have (and will continue to) go on unreasonably long hot streaks without suddenly having equally long/extreme cold streaks. And ditto for cold streaks like Al's not meaning that he's going to start shooting 57% on threes in the next 2 rounds.

Sure, they don't HAVE TO have unreasonably cold streaks after a hot streak. But they sometimes will. Shooting game to game is basically a normal distribution around your true average. most games you'll be relatively close to your average but for stretches you'll be crazy good or crazy bad, the tails of the distribution.

Following a crazy outlier shooting stretch one direction with a crazy outlier shooting stretch the other will happen sometimes. This is one of those times.

I'm not even sure that is true. Take Curry in his best 3 point shooting season. He shot 45.4%, but here are the number of games and the percentage he shotnin them

60% or better - 11 games
> 50% to 60% - 10 games
Exactly 50% - 13 games
> 45.4 to 50%  - 5 games
40-45.4% - 14 games
30-39.9% - 13 games
< 30% - 11 games

In other words, the greatest shooter ever in his best season was all over the map.  And basically had as many downright terrible games as he did amazingly awesome games.

The example you just provided is basically a bell curve. Here's the # of games broken down by 5% intervals:

0-5: 0
5.01-10: 2
10.01-15: 1
15.01-20: 2
20.01-25: 6
25.01-30: 5
30.01-35: 6
35.01-40: 6
40.01-45: 12
45.01-50: 19
50.01-55: 4
55.01-60: 7
60.01-65: 3
65.01-70: 3
70.01-75: 4
75.01-80: 0

The meat of his attempts are around his seaosn of average of 45 +/- 5%, with expanding shoulders out to there shrinking down to very few occurrences at the extremes. IE: A bell curve.

A seaosn full of shooting attempts will always have numbers all over the place, they'll just be more occurrences generally close to a players average.

[Moranis]

 25% of his games were more than 15% off his average. That isn't a typical bell curve

[Roy H.]

25% of his games were more than 15% off his average. That isn't a typical bell curve

I'm guessing that Probability and Statistics wasn't your best course?

[Moranis]

Another excellent game by Jrue.

Yeah, who knew White was going to fall off a cliff? We've really needed Jrue's steady hand these past couple of games. Dude decided to step up at exactly the right time.

Didn't everyone know White was going to fall off a cliff? I mean White was always going to regress to the mean.  That is how those things work.

There’s a space between clear All-Star and 2022 playoffs disaster. We thought he had moved past the latter. Hopefully it really is just a mini-slump and not indicative of something greater. We are who we are this season in large part to White elevating his game.

But luckily Jrue suddenly became a much better player just as White fell off. That has been huge. If both of them just ‘regress to the mean’, then we will be fine, as well

well when you play well above your level, you are invariably going to play well below it at some point, so that you end up where you should be

No, that isn't how probabilities or regression to the mean work. There isn't some cosmic force that will make you play the perfect level of bad to instantly balance out the good, it's far, far more likely to have a bunch slightly below average games than a couple of real stinkers.

But even that isn't guaranteed, players have (and will continue to) go on unreasonably long hot streaks without suddenly having equally long/extreme cold streaks. And ditto for cold streaks like Al's not meaning that he's going to start shooting 57% on threes in the next 2 rounds.

Sure, they don't HAVE TO have unreasonably cold streaks after a hot streak. But they sometimes will. Shooting game to game is basically a normal distribution around your true average. most games you'll be relatively close to your average but for stretches you'll be crazy good or crazy bad, the tails of the distribution.

Following a crazy outlier shooting stretch one direction with a crazy outlier shooting stretch the other will happen sometimes. This is one of those times.

I'm not even sure that is true. Take Curry in his best 3 point shooting season. He shot 45.4%, but here are the number of games and the percentage he shotnin them

60% or better - 11 games
> 50% to 60% - 10 games
Exactly 50% - 13 games
> 45.4 to 50%  - 5 games
40-45.4% - 14 games
30-39.9% - 13 games
< 30% - 11 games

In other words, the greatest shooter ever in his best season was all over the map.  And basically had as many downright terrible games as he did amazingly awesome games.

The example you just provided is basically a bell curve. Here's the # of games broken down by 5% intervals:

0-5: 0
5.01-10: 2
10.01-15: 1
15.01-20: 2
20.01-25: 6
25.01-30: 5
30.01-35: 6
35.01-40: 6
40.01-45: 12
45.01-50: 19
50.01-55: 4
55.01-60: 7
60.01-65: 3
65.01-70: 3
70.01-75: 4
75.01-80: 0

The meat of his attempts are around his seaosn of average of 45 +/- 5%, with expanding shoulders out to there shrinking down to very few occurrences at the extremes. IE: A bell curve.

A seaosn full of shooting attempts will always have numbers all over the place, they'll just be more occurrences generally close to a players average.

It is mostly flat with a real high peak in the middle.  That isn't a typical bell curve, but is typical for sports i.e. sports will have a lot more extremes than regular probability even for the greatest shooter of all time. And yes even in sports the largest number of games will be closer to the average, after all there is a reason the guy averages that number.

[keevsnick]

Another excellent game by Jrue.

Yeah, who knew White was going to fall off a cliff? We've really needed Jrue's steady hand these past couple of games. Dude decided to step up at exactly the right time.

Didn't everyone know White was going to fall off a cliff? I mean White was always going to regress to the mean.  That is how those things work.

There’s a space between clear All-Star and 2022 playoffs disaster. We thought he had moved past the latter. Hopefully it really is just a mini-slump and not indicative of something greater. We are who we are this season in large part to White elevating his game.

But luckily Jrue suddenly became a much better player just as White fell off. That has been huge. If both of them just ‘regress to the mean’, then we will be fine, as well

well when you play well above your level, you are invariably going to play well below it at some point, so that you end up where you should be

No, that isn't how probabilities or regression to the mean work. There isn't some cosmic force that will make you play the perfect level of bad to instantly balance out the good, it's far, far more likely to have a bunch slightly below average games than a couple of real stinkers.

But even that isn't guaranteed, players have (and will continue to) go on unreasonably long hot streaks without suddenly having equally long/extreme cold streaks. And ditto for cold streaks like Al's not meaning that he's going to start shooting 57% on threes in the next 2 rounds.

Sure, they don't HAVE TO have unreasonably cold streaks after a hot streak. But they sometimes will. Shooting game to game is basically a normal distribution around your true average. most games you'll be relatively close to your average but for stretches you'll be crazy good or crazy bad, the tails of the distribution.

Following a crazy outlier shooting stretch one direction with a crazy outlier shooting stretch the other will happen sometimes. This is one of those times.

I'm not even sure that is true. Take Curry in his best 3 point shooting season. He shot 45.4%, but here are the number of games and the percentage he shotnin them

60% or better - 11 games
> 50% to 60% - 10 games
Exactly 50% - 13 games
> 45.4 to 50%  - 5 games
40-45.4% - 14 games
30-39.9% - 13 games
< 30% - 11 games

In other words, the greatest shooter ever in his best season was all over the map.  And basically had as many downright terrible games as he did amazingly awesome games.

The example you just provided is basically a bell curve. Here's the # of games broken down by 5% intervals:

0-5: 0
5.01-10: 2
10.01-15: 1
15.01-20: 2
20.01-25: 6
25.01-30: 5
30.01-35: 6
35.01-40: 6
40.01-45: 12
45.01-50: 19
50.01-55: 4
55.01-60: 7
60.01-65: 3
65.01-70: 3
70.01-75: 4
75.01-80: 0

The meat of his attempts are around his seaosn of average of 45 +/- 5%, with expanding shoulders out to there shrinking down to very few occurrences at the extremes. IE: A bell curve.

A seaosn full of shooting attempts will always have numbers all over the place, they'll just be more occurrences generally close to a players average.

It is mostly flat with a real high peak in the middle.  That isn't a typical bell curve, but is typical for sports i.e. sports will have a lot more extremes than regular probability even for the greatest shooter of all time. And yes even in sports the largest number of games will be closer to the average, after all there is a reason the guy averages that number.

So out of curiosity I went and found the standard deviation of Curry's shooting performance that year in an excel spread sheet. His game to game three point shooting percentages that season had a standard deviation of about 15% with an average of about 45%. In a normal distribution 68.2% of all values will fall with standard deviation of the mean, so about 68.2% of all values should be between 30-60%. In this case 55 of 79 games fell within 1 standard deviation of the mean (ie he shot 30%-60%). Thats.....69.6% of his games. An additional 27.2% of values should fall between 1 and 2 standard deviations. Thats between 15-29.9 and 60.1-75. 21 of 79 games fall in that range. Thats....26.5% of his games. About 5% should fall on the extremes. 3 of 79 in this case do. Thats....about 3.8%

This is very close to your standard normal distribution. So yes, this IS a typical bell curve.

[green_bballers13]

Mo- you’re on a cold streak across multiple threads. Can you shoot your way out of this slump?

[smokeablount]

Mo- you’re on a cold streak across multiple threads. Can you shoot your way out of this slump?

I blame his never ending list of lower body injuries for this.

[mobilija]

Mo- you’re on a cold streak across multiple threads. Can you shoot your way out of this slump?

He’ll revert back to his mean eventually 

Hopefully boosted by a successful Finals prediction of Celtics in 6

[Vermont Green]

Mo- you’re on a cold streak across multiple threads. Can you shoot your way out of this slump?

He’ll revert back to his mean eventually 

Hopefully boosted by a successful Finals prediction of Celtics in 6

Well played mobilija.

I think it has been pretty well documented statistically, that there really isn't a thing as being hot or being cold.  NBA shooters' performances end up following traditional statistical patterns, bell curves or what ever.  The problem is that when you are playing or watching as a fan, you feel like you are hot or the player looks like he is hot, or vice verse.

In reality, players get hot and cold along pretty expected statistical trends.  It is kind of a natural feedback loop.  The player has good games and bad games, and these contribute to his average, and then the average is what informs prediction of future performance overall.  Not specific to one game or one shot, but overall.

And of course if you don't buy all this statistical stuff, you can just pay attention to the games and realize that a player like Derrick White is not going to continue to shoot over 50% from 3, as he did for 3 or 4 games, after you have seen him be a 39% 3-point shooter over the last few seasons encompassing hundreds of games.  You shouldn't need a degree in statistics to figure that out.

But this topic is about Jrue.  He is playing much better the last few games, as expected.

[Phantom255x]

At the end of the day, Jrue has had three great games in row and he's doing it in a variety of ways, but he's also been unselfish in the process.

I'll gladly take it.

And we've needed it with Horford, Hauser and White not looking so good most of the series.

[Roy H.]

Mo- you’re on a cold streak across multiple threads. Can you shoot your way out of this slump?

He’ll revert back to his mean eventually 

Hopefully boosted by a successful Finals prediction of Celtics in 6

Well played mobilija.

I think it has been pretty well documented statistically, that there really isn't a thing as being hot or being cold.  NBA shooters' performances end up following traditional statistical patterns, bell curves or what ever.  The problem is that when you are playing or watching as a fan, you feel like you are hot or the player looks like he is hot, or vice verse.

In reality, players get hot and cold along pretty expected statistical trends.  It is kind of a natural feedback loop.  The player has good games and bad games, and these contribute to his average, and then the average is what informs prediction of future performance overall.  Not specific to one game or one shot, but overall.

And of course if you don't buy all this statistical stuff, you can just pay attention to the games and realize that a player like Derrick White is not going to continue to shoot over 50% from 3, as he did for 3 or 4 games, after you have seen him be a 39% 3-point shooter over the last few seasons encompassing hundreds of games.  You shouldn't need a degree in statistics to figure that out.

But this topic is about Jrue.  He is playing much better the last few games, as expected.

How does this equate to there being no such thing as being hot or cold?