Daily Thunder.com

@DSMok1

Maybe so, however this presumption gets to the research questions- is clutchtime distinctive and does giving that data elevated status in some degree and fashion help predict future clutchtime better than career numbers alone. More that could be done...

Crow :
Can’t control responses, even to direct questions researched and answered…
so here is a follow-up answer…
I would think that a model for future clutch shooting that included “both” his career average shooting (or a time sequence of shooting percentages) and how he’s performed in clutch situations in the past “might” do better than either alone.

Probably, but probably not much better than the career numbers.

@dream catcher

I dunno, I strongly disagree. There are plenty of 'calm assassin' types in the NBA who have had a lot of success. Tim Duncan is a perfect example. You don't have to be visibly pumped up to perform in big moments. I think KD does just fine, and his mediocre performance in the clutch is mostly due to our isolation offense we run - and the defenses keyed specifically to stop it.

Can't control responses, even to direct questions researched and answered...

so here is a follow-up answer...

I would think that a model for future clutch shooting that included "both" his career average shooting (or a time sequence of shooting percentages) and how he’s performed in clutch situations in the past "might" do better than either alone.

Sammy :@DSMok1

No, not really. To tell you the truth, a lot of that stuff is over my head or else the discussions get so granular that it’s hard for me to make sense of the bigger picture. I’ve never taken a stat course, not even in HS, so that probably doesn’t help.

I haven't either... I'm a structural engineer that taught myself statistics. I was homeschooled in high school and did algebra/calculus. No stats anywhere. I'm no expert on stats from a theoretical side at all...

@DSMok1
No, not really. To tell you the truth, a lot of that stuff is over my head or else the discussions get so granular that it's hard for me to make sense of the bigger picture. I've never taken a stat course, not even in HS, so that probably doesn't help.

BTW: Sammy, do you post over at APBRmetrics? You'd fit in well over there, and any work you do would be read by people with lots of experience (I'm not among those).

Given the changes in how coaches call plays in the fourth, I think you'd have to go through Synergy data and compare at efficiency stats in specific situations (isos, PnRs, spot ups, screen curls, et al) in order to get any sort of apples to apples number comparisons. At that point, the samples are probably so small as to be useless.

I agree: clutch is different in basketball. You can run every play to your "clutch" players if you'd like. In general, production seems to suffer in clutch time, perhaps because teams use more predictable plays.

Personally, I believe players probably do have a certain amount of "clutchness" but I don't think there is enough data to determine it statistically. I think perhaps, I say perhaps, this is an area where coaches' feel may be more perceptive than the numbers.

Sammy :@dream
catcher
This is the attitude of most basketball fans. The clutch numbers on this page do bear out your observations: http://www.82games.com/0910/CSORT11.HTM.
What I’m asking is, going forward, in these same situations, should I expect a player to perform closer to his career averages or should I expect him to perform how he’s performed in clutch situations in the past? For example, KD shot 36% from the field and 26% from three in clutch situations last year (as defined by that 82 games chart). In future clutch situations, which is more likely: that he hits at those percentages or that he hits at percentages based on a weighted career-average stat line?
As DSMok1 noted above, clutch stats in baseball are pretty much completely useless, despite most fans perceptions. This has been proven statistically. How Derek Jeter or A-Rod has fared in clutch situations in the past has no bearing on how well they will fare in clutch situations in the future.
A couple interesting points have been raised above, though, which differentiate basketball clutch situations from baseball ones. In baseball, the game is exactly the same in a clutch situation as it is in the top of the 1st. It’s still just you vs. the pitcher. In basketball, coaches will completely change the gameplan in clutch situations, often going ISO with their star player on every play. This is a different game than the players are playing in the first quarter, so it follows that statistical markers should be different. I didn’t consider this when I first raised the question.

you bring up a good point in the difference between baseball and basketball clutch. Basketball not only has coaches draw up plays not seen before, but everyone's intensity raises a notch and their sense of urgency raises. KD seems to play at one level all the time, only on a rare occasion do you see him play with any urgency. When KD is focused and his sense of urgency is higher then normal he is a special special player and a treat to watch. Unfortunately that doesn't happen very much at all, and I wonder if that will ever come out of him in clutch situations.

@justin
Lets not pretend KD hasnt had clutch plays blown. Not saying he cant be a clutch player but not yet.
And dont say Green had the ball in his hand on accident at the end of those games. Because I assure you he was the 2nd option.

Sammy :@dream
catcher
This is the attitude of most basketball fans. The clutch numbers on this page do bear out your observations: http://www.82games.com/0910/CSORT11.HTM.
What I’m asking is, going forward, in these same situations, should I expect a player to perform closer to his career averages or should I expect him to perform how he’s performed in clutch situations in the past? For example, KD shot 36% from the field and 26% from three in clutch situations last year (as defined by that 82 games chart). In future clutch situations, which is more likely: that he hits at those percentages or that he hits at percentages based on a weighted career-average stat line?
As DSMok1 noted above, clutch stats in baseball are pretty much completely useless, despite most fans perceptions. This has been proven statistically. How Derek Jeter or A-Rod has fared in clutch situations in the past has no bearing on how well they will fare in clutch situations in the future.
A couple interesting points have been raised above, though, which differentiate basketball clutch situations from baseball ones. In baseball, the game is exactly the same in a clutch situation as it is in the top of the 1st. It’s still just you vs. the pitcher. In basketball, coaches will completely change the gameplan in clutch situations, often going ISO with their star player on every play. This is a different game than the players are playing in the first quarter, so it follows that statistical markers should be different. I didn’t consider this when I first raised the question.

I think KD and Russ can improve their clutchness with more experience in those situations. I doubt they ever become Kobe, but they should be at least average with time.

justin :@dream
catcher
This is coaching, IMO. It’s easy to appear ‘clutch’ when you have a system and you’re running plays. Watch the Spurs play. They have clutch performers because those guys are put into a position to succeed.
Kobe is thought of as ‘clutch’ but I bet his FG% in clutch situations isn’t much higher (if at all) than his overall percentage. Jeff Green is perceived as clutch because he’s made plays – but many of these plays (Atlanta is a good example) only happen because of the attention Kevin Durant receives, as Keith alluded to. The whole isolation offense in clutch situations is a doomed strategy, it’s not Kevin Durant’s fault his clutch performance is not up to par.

good point. But really how many teams run a play at the buzzer? Most often its get your best player the ball and let him go to work.

@dream catcher
This is the attitude of most basketball fans. The clutch numbers on this page do bear out your observations: http://www.82games.com/0910/CSORT11.HTM.

What I'm asking is, going forward, in these same situations, should I expect a player to perform closer to his career averages or should I expect him to perform how he's performed in clutch situations in the past? For example, KD shot 36% from the field and 26% from three in clutch situations last year (as defined by that 82 games chart). In future clutch situations, which is more likely: that he hits at those percentages or that he hits at percentages based on a weighted career-average stat line?

As DSMok1 noted above, clutch stats in baseball are pretty much completely useless, despite most fans perceptions. This has been proven statistically. How Derek Jeter or A-Rod has fared in clutch situations in the past has no bearing on how well they will fare in clutch situations in the future.

A couple interesting points have been raised above, though, which differentiate basketball clutch situations from baseball ones. In baseball, the game is exactly the same in a clutch situation as it is in the top of the 1st. It's still just you vs. the pitcher. In basketball, coaches will completely change the gameplan in clutch situations, often going ISO with their star player on every play. This is a different game than the players are playing in the first quarter, so it follows that statistical markers should be different. I didn't consider this when I first raised the question.

Clutch players have no pressure when the game is on the line. Their veins run ice cold. Calm cool and collected.

Shoot my palms sweat and Iam sitting on the couch watching games.

It's easy to be clutch when you are Jeff Green and you have no responsibility to handle the ball and score most of the time.

@dream catcher

This is coaching, IMO. It's easy to appear 'clutch' when you have a system and you're running plays. Watch the Spurs play. They have clutch performers because those guys are put into a position to succeed.

Kobe is thought of as 'clutch' but I bet his FG% in clutch situations isn't much higher (if at all) than his overall percentage. Jeff Green is perceived as clutch because he's made plays - but many of these plays (Atlanta is a good example) only happen because of the attention Kevin Durant receives, as Keith alluded to. The whole isolation offense in clutch situations is a doomed strategy, it's not Kevin Durant's fault his clutch performance is not up to par.

Using 82games' definition of when it is clutch and their data from 09-10 and 08-09 regular season and choosing to limit the list to those with 100+ clutch minutes in each season and throwing out any names I couldn't easily confirm as exact matches I was left with 62 names and found the correlation of the simple raw FG%s from season to season in the clutch to be .53. This is probably a good deal higher than it would be for the full population. Converting to eFG%, TS% and Clutch points per minute had equal or higher correlations. This was fast and very dirty but I think I'll settle for thinking there is medium correlation for the bigtime clutch shooters though maybe not as strong for clutch shooting as for overall shooting.

Clutch players like Micheal Jordan, Kobe Bryant, Reggie Miller, Larry Bird......would disagree

clutch is not just numbers, but a feel and attitude during pressure situations. Durant is not a clutch player because he seems to tighten up, same for Russ as he loses a lot of balls in crunch time.

being a great clutch player is to not feel any pressure and just execute under duress. Which is why when you watch Jeff Green you can tell he is clutch just by his body language, he doesn't put any extra pressure on himself. Eric Maynor IMO is clutch as well. You need both extreme confidence in the clutch and the ability to focus.

@Keith
Are you saying a close game might as well be lost since it came to that play, or that you just wouldn't select a player based on the ability to make that play?

@justin
I imagine one could get a wide enough sample size if you opened the pool beyond game winner/tie attempts. Say the last 2 minutes of the fourth when the score margin is ≤ 5. As DSMok, noted though, that data's pretty hard to cull without going through the PbP data for each team/player.

@DSMok, thanks for that link.

I've asked Pelton if he knows of any studies; I'll pass along anything he finds.

@justin
I agree. Clutch exists so far outside the average that it is mostly worthless to me. Clutch exists in an NBA where everyone knows who is going to take the shot 90% of the time, and thus defenses can adjust to it. Further, clutch is tremendously affected by the creativity and execution of a coach and worn out team (on both sides). Kevin Durant isn't clutch because our offensive gameplan isn't good and he's not a strong ISO player. Kobe wasn't as clutch as he is perceived. The best "clutch" players are generally the ones who receive rare passes from their team's star and have wide open looks.

Whenever someone starts with, "All right, if the championship was on the line, with 5 seconds left... " I like to stop them and say, "Why did it come to this?" If it comes down to the very last shot in the very last game, it doesn't really matter who I have. The shot is going in less than 50% of the time. Give me someone who makes a difference for the first 47:55, and leave the last second heave to another team.

I've always thought clutch performance is such a small sample that it's basically worthless.

Also a lot of clutch attempts happen as desperation, low percentage type shots. I'd imagine this colors the percentages somewhat.

@Sooners
Nah man we need to wait on Green and see what he can do in this new improved offense. Krstic is not my favorite player but serviceable as long as he also can hit those 20 footers better then last game. Then again I would trade him if the opportunity came:) Why you going to get rid of Maynor he is a solid backup PG who is lately hitting 3's. And he is not a shooting guard. If anything Cook, Sef or Harden should be hitting the deep ball.

Here's a cool study on clutch: http://www.hardballtimes.com/main/article/with-the...

Sammy :
Justin, DSMok, Crow, AC or anyone else:
Does anyone know of any studies on the correlation between previous clutch performance (using whatever metric) and future clutch performance? Is the r^2 higher than using overall performance as the predictive measure?

I don't know of many NBA studies--the data is just hard to get unless you parse PbP data. Here is one discussion: http://sonicscentral.com/apbrmetrics/viewtopic.php...

I would guarantee the R^2 of clutch to clutch is below that of regular season to clutch. The question is if it's even above 0. I don't know if it is. Basically, the methodology I might use would be to look at clutch shooting and see if the standard deviation of the z-scores is greater than 0. That's something Tangotiger uses regularly in baseball. I would guess there is so much small-sample-size variation that it just about drowns out the signal.

We need to get rid of Green, Krstic, Maynor and get a legit 3 point shooter and Big Man.

@Sammy

I am looking around a bit. There are things out there but I may not be able to remember or summarize it all. In baseball I have seen studies that say almost no correlation between past and future clutch performance and others with moderate correlation (.3). I might do my own quick regression if it doesn't take too much time.

@Sammy

Blatche will be a scary player if he ever learns how to play basketball.

That foam in the car is absolutely hilarious. It's not like "mean" hazing either. It's just funny. I love this team.

heh.
http://yfrog.com/b9k9qhj

This been posted yet?
Andray Blatche uses The Force on Rondo floater

Justin, DSMok, Crow, AC or anyone else:

Does anyone know of any studies on the correlation between previous clutch performance (using whatever metric) and future clutch performance? Is the r^2 higher than using overall performance as the predictive measure?

Looking at it further, the 82 game statements about win % while on and off the floor vary depending whether you look at shift or game level so the on /off comparison I made might not be exactly as stated or as severe but there is an on /off performance gap of some size that you would want to see lowered. Quality of opponents varies too, but if you care about that, then you can go to Adjusted +/- and he hasn't looked good on that yet.

@Greg

The Thunder do have the potential to have 3 - 5 2 way strong players or at least strong on one / near neutral on the other. That prospect could lead to great results if most or all of it comes to pass at the same time.

Green has been strong at offense at SF this season, weak on SF defense. Last season the reverse. Hasn't been both at the same time yet and good at team level there previously but there is room for hope.

His win % on the court at SF is at 50% for the first time.
His win % at PF this season is at 50% also, 53% last season. Team win % was about 65% last season when he was off altogether. And it looks like about 75% so far this season.

In year 1 he wasn't what I'd call good at any of the 3 measures at either position. There was been some improvement but how much and how much that is worth and how it compares with Ibaka and other options can be and has been discussed a lot and I'll leave it there.

Green should come off the bench and ball against scrubs, he's great at this. Do it!

The better solution is to start Ibaka with Krstic and start Harden for Sefolosha.

Greg :
From Hollinger’s chat today:
Jeremy (Portland)
Do the Thunder need to seriously think about moving Harden & Green for a better/more consistent player?
John Hollinger (2:33 PM)
What are they gonna get for those two? I’d love to see them start Ibaka at the 5 and keep Green at the 4 — that’s a small team but it’s the only way they have enough floor spacing. As for Harden, he’s been a total mess this year. I think you have to see if he can get it turned around, because right now they won’t be able to get fair value for him.

Ew. Why does it make so much sense to move Green to the Bench and play the majority of his minutes at the 3 and some at the 4, but yet nobody outside of this board even considers it an option? Kristic is already a floor stretching 5, and ibaka can hit the long two as well. Yea you can't call an isolation for either one of them on the perimeter, but who cares? Jeff Green will be a better player as a backup three, guaranteed, than he is as a starting 4. In turn, you might see Harden be a better backup two as well.

2 way strong players are rare or very rare depending on your standards.

Where does Beasley play on this team? Same problem we'd have with Jeff Green. I think Green could put up 20+ on that horrible Minnesota team.

Sammy :@justin

He certainly makes great adjustments in the playoffs. I like Adelman a lot. I became a big Alvin Gentry fan after last spring. Monty Williams is looking like a fantastic hire in NO.

word. ive been loving the way the hornets play under williams this season. I really don't like Mo Cheeks to be honest. Didn't like him in Philly, don't like our success with him on the Bench this year, and don't like him as a potential head coach for us. If we do get rid of Brooks, I'd like to see Rick Adelman come in, along with another defensive specialist assistant coach. But like it's been said on here multiple times, the likelihood of that happening is quite low...

But I still want either Scott Brooks out soon, or another quality assistant coach in even sooner.

@Cory
Evans would be too redundant on this team, he's very similar to Westbrook. Curry would be an intriguing fit here. And I don't see how Beasley would fit, he's clearly playing his best ball at the SF position - which happens to be the same position Kevin Durant plays.

Not sure how Green offers more floor spacing than Krstic when he's shooting 17% from 3PT range.

@Crow
Green at his natural position. Ibaka is getting there. It's a shame if Westbrook and KD aren't more consistently.

I know everyone focuses on the Oden/Durant debate. When will we start to mention the missed oppurtunities for the 3rd pick with Harden over Curry or Evans? Also, why could the Thunder not trump 2-2nds for Beasley - he's even boys with Durant. What a miss that may turn out to be if Beasley continues to improve...he's only 21. We have a great front office, but they tend to get a free pass on some glaring mistakes.

Do the Thunder have any on average 2 way strong players? Last season KD was. Not sure about right now.

From Hollinger's chat today:

Jeremy (Portland)
Do the Thunder need to seriously think about moving Harden & Green for a better/more consistent player?

John Hollinger (2:33 PM)
What are they gonna get for those two? I'd love to see them start Ibaka at the 5 and keep Green at the 4 -- that's a small team but it's the only way they have enough floor spacing. As for Harden, he's been a total mess this year. I think you have to see if he can get it turned around, because right now they won't be able to get fair value for him.

Adelman hasn't got much heat yet to my knowledge but if things continue there might be more looking for a change.

I probably won't go into detail about them but I will say they have few if any 2 way strong players right now.

Scott Brooks is fine. He is in some ways just like our players, still growing still learning. Sam Presti really believes in growing organically and being patient. He will be our coach for a long time.