OT-ish: the "clutch" question

[handiman]

An increasingly popular topic of late on this board (free agency notwithstanding) has been whether there is any statistical support for the notion of clutch players, or if it's just a media creation for people that maintain their level of performance in clutch situations.

Minstrel pointed out in one of the threads that athletes naturally select themselves from the general population as types that are better suited to handle the pressure and not crumble in those situations. (Cliff Robinson wishes you could convince him of that.) That made sense on the surface...

It occurred to me the other day that the reason, IMO, that statistics cannot capture the essence of clutch performers is that there is no way to measure defensive intensity, which presumably steps up in the same situations and counters the offensive "clutch-ness."

Then I started thinking, why not look at a much "simpler" sport like Track and Field, where it's all about the raw performances ... other than little things like weather conditions and tactical racing, of course. There's no question that some athletes consistently save their best for last while others consistently crumble in the big meets, and training cycles and peaking/tapering can never explain it. I've worked with national champions, olympians, and world record holders, and it holds true at all levels. "Clutch" clearly exists in T&F, so it follows to reason that it also holds true in other sports such as basketball, just it is harder to quantify.

 

[huevonkiller]

An increasingly popular topic of late on this board (free agency notwithstanding) has been whether there is any statistical support for the notion of clutch players, or if it's just a media creation for people that maintain their level of performance in clutch situations.

Minstrel pointed out in one of the threads that athletes naturally select themselves from the general population as types that are better suited to handle the pressure and not crumble in those situations. (Cliff Robinson wishes you could convince him of that.) That made sense on the surface...

It occurred to me the other day that the reason, IMO, that statistics cannot capture the essence of clutch performers is that there is no way to measure defensive intensity, which presumably steps up in the same situations and counters the offensive "clutch-ness."

Then I started thinking, why not look at a much "simpler" sport like Track and Field, where it's all about the raw performances ... other than little things like weather conditions and tactical racing, of course. There's no question that some athletes consistently save their best for last while others consistently crumble in the big meets, and training cycles and peaking/tapering can never explain it. I've worked with national champions, olympians, and world record holders, and it holds true at all levels. "Clutch" clearly exists in T&F, so it follows to reason that it also holds true in other sports such as basketball, just it is harder to quantify.

If you're measuring statistics in clutch situations, then you are automatically accounting for this increased defensive pressure as well.

 

[handiman]

If you're measuring statistics in clutch situations, then you are automatically accounting for this increased defensive pressure as well.

How? All you can measure is the shooting percentages and what not. Increased defensive intensity counters increased offensive production... This might cause offense to remain steady (the claim of the clutch-is-a-myth contingent) despite clutch play.

 

[huevonkiller]

How? All you can measure is the shooting percentages and what not. Increased defensive intensity counters increased offensive production... This might cause offense to remain steady (the claim of the clutch-is-a-myth contingent) despite clutch play.

If you measure clutch statistics in relation to your peers, that's all that really matters to be honest.

From what I've observed though, you're right normally "clutch" players are the ones that have the same WS/48, PER, etc. in the playoffs. Not necessarily better just at a comparable level.

 

[Nikolokolus]

It's actually not that hard to quantify at all ... go to 82games.com and see how guys perform in the last minutes of a game when the score is close. The only thing that makes it kind of hard to evaluate is that not every game ends close, so the sample size is smaller compared to the sum of all regular season and playoff games. As huevon mentioned above, the simplest measure is to just compare the player to their peers to see where they stack up; if you really wanted to dig deep you could compare them to historically great players (if such data exists).

 

[handiman]

It's actually not that hard to quantify at all ... go to 82games.com and see how guys perform in the last minutes of a game when the score is close.

That does nothing to answer the question of how defensive intensity in late game situations negates offensive efforts, however. Statistics simply cannot measure it.

As huevon mentioned above, the simplest measure is to just compare the player to their peers to see where they stack up

The same inability to measure the above aspect applies to all players, so comparing one to another is just skimming over the same averaging problem. That's why the statistics yield at best inconclusive results...

 

[Nikolokolus]

That does nothing to answer the question of how defensive intensity in late game situations negates offensive efforts, however. Statistics simply cannot measure it.


The same inability to measure the above aspect applies to all players, so comparing one to another is just skimming over the same averaging problem. That's why the statistics yield at best inconclusive results...

I honestly have no idea what you are talking about ...

 

[handiman]

I honestly have no idea what you are talking about ...

Really? I thought it was pretty clear, but let me try again...

Player A is in a "clutch" offensive situation. Player B's team is dialed in for one last defensive stand, knowing the play that is likely to be run for Player A. If Player A does not go above and beyond (focus and effort-wise) what they've done throughout the course of the game, then the results will be worse due to the elevated defensive pressure. That's why the notion that clutch play is a myth -- due to those players at best holding even with their normal production -- is an erroneous conclusion due to statistic's inability to measure half of what's going on.

 

[rocketeer]

Really? I thought it was pretty clear, but let me try again...

Player A is in a "clutch" offensive situation. Player B's team is dialed in for one last defensive stand, knowing the play that is likely to be run for Player A. If Player A does not go above and beyond (focus and effort-wise) what they've done throughout the course of the game, then the results will be worse due to the elevated defensive pressure. That's why the notion that clutch play is a myth -- due to those players at best holding even with their normal production -- is an erroneous conclusion due to statistic's inability to measure half of what's going on.

which is why people are saying they should be measured against their peers, not just measured against their previous production during "non clutch" minutes of games. the assumption being that all superstar players would be receiving elevated defensive pressure as well.

 

[Nikolokolus]

Really? I thought it was pretty clear, but let me try again...

Player A is in a "clutch" offensive situation. Player B's team is dialed in for one last defensive stand, knowing the play that is likely to be run for Player A. If Player A does not go above and beyond (focus and effort-wise) what they've done throughout the course of the game, then the results will be worse due to the elevated defensive pressure. That's why the notion that clutch play is a myth -- due to those players at best holding even with their normal production -- is an erroneous conclusion due to statistic's inability to measure half of what's going on.

For the sake of argument let's say that Lebron James shoot 46% in clutch situations and the next best player after him shoots 42% in clutch situations and the average shooting percentage for all players is 36% in clutch situations, then what part of this comparitive analysis makes it impossible to say who statistically is better or worse in a clutch situation? Nobody is saying players are suddenly exceeding their "normal" in game performance or averages, but there is a way to quantify who is the best (relatively speaking) in these situations.

 

[handiman]

which is why people are saying they should be measured against their peers, not just measured against their previous production during "non clutch" minutes of games.

I get that, but it's only a side issue. That only answers the question of whether an individual does better or worse than average. But the overall question is whether "clutch" is a myth perpetuated by the media. Since "clutch" play is at least somewhat negated by comparable defensive effort, the individual comparisons will fail to reflect what's really going on from that perspective.

For the sake of argument let's say that Lebron James shoot 46% in clutch situations and the next best player after him shoots 42% in clutch situations and the average shooting percentage for all players is 36% in clutch situations, then what part of this comparitive analysis makes it impossible to say who statistically is better or worse in a clutch situation?

See above. If there are examples that cut and dry, I imagine there would be little argument...

Nobody is saying players are suddenly exceeding their "normal" in game performance or averages...

Hmm, I seem to recall lots of people saying exactly that and being countered by the stats argument...

 

[jlprk]

We need a PER for each quarter. PER would factor in how much the team outscores the other in the quarter. Then we need a Roland Rating for each quarter. That would include defense. Then we could see whether Aldridge performs when it doesn't matter, and Roy when it does. Then we need those stats for the last 5 minutes and 1 minute, but only for close games.

This can all be done, if someone had a case of coffee and a giant computer. A generation from now, maybe 5 years from now, it will all exist. Why doesn't ESPN transfer its money from its stupid moderator company to actual research?