Log in

No account? Create an account
Money - De File
Does Collecting Make You Feel Dirty?
I really dislike the Lakers. Not as much as I dislike Republicans, naturally, but the animosity is strong. Nevertheless, I have to give it to Kobe. Steven has repeatedly deconstructed the idea of "clutch performance" -- the whole game counts, after all -- but if there is such a thing, Kobe may be its most impressive exemplar. And, though I'm writing this wearing my Cal shirt, I have to give props to Luke Walton too. He just makes everyone around him play better.

Mode: reluctant
Muse: Rockets - Cat Power - Protest Records

8 comments or Leave a comment
masoo From: masoo Date: June 9th, 2004 06:56 am (UTC) (LINK TO SPECIFIC ENTRY)
There are descriptive methods, and predictive methods. A descriptive method describes what has happened: "Man, Kobe sure came up big, didn't he?" We all agree about that ... nothing really to argue with. But if something doesn't have predictive power, it's pointless to use the description to pump up artificial notions of how athletes perform. In every sporting event, someone does something that amounts to "coming up big." But "clutch performance" as a concept assumes predictive value: Kobe Bryant didn't just come up big in Game Two, Kobe Bryant is a "clutch player" and Game Two proves it.

But no one has ever demonstrated conclusively that "clutch" is predictive. You can say that Kobe Bryant is likely to play well in Game Three, but the reason for that is simple: Kobe is a great basketball player. He will have more great Game Twos or Threes than will Derek Fisher, because Kobe is BETTER than Derek Fisher. But there is nothing in the performance of any athlete who has been labeled good or bad "in the clutch" that can't be easily explained as normal variation within a player's range of ability. Kobe's Game Two doesn't prove anything: he had a great game, he's a great player, he'll likely be great in Game Three because he's a great player, but none of that is about "clutch."

And, like Charlie said, if the Lakers just scored 20 measly points in the first period, there wouldn't have been a need for overtime. But if Joe Blow had scored a basket for L.A. in the first quarter, and they ended up winning by two, no one would reward Mr. Blow with the label "clutch player."
cbertsch From: cbertsch Date: June 9th, 2004 07:27 am (UTC) (LINK TO SPECIFIC ENTRY)

Clutching At Straws

I can't really argue with anything you say here, Steven. But I still want to make my standard comeback, which is that the subjective perception of the situation is a factor that must be accounted for. Baseball is different from basketball in this regard, certainly. In basketball, you have to want to shoot, to seek out your shot. In times of crisis, some players shy away from the responsibility. Some of those players are good players, too. Chris Webber, for example, has often been harshed on for disappearing at the end of ballgames. Aside from a player wanting to shoot or not, though, I also think that the feeling of being under pressure matters. Sure, if I wrote the same amount every day, I wouldn't have to rush to make deadlines. Having said that, though, I know pleanty of people who freak out when they do have their backs against the wall, for whatever reason, and others, like me, who actually grow calmer. Returning to your example of Joe Blow, I'd make the counter-argument that Joe might be less likely A) to want to take the shot at the end of the game and B) to make that shot if he took it in comparison to a Kobe or Michael Jordan. The flipside of the whole "Two points is two points" idea is that the end of the game is still the end of the game. Even if you realize that a made free throw in the first quarter would have made the one you're shooting with two seconds left beside the point, you still know that the fate of the game hinges on your own performance. And that logic can be extended to baseball. Maybe I see things this way because I played organized sports more than you did, Steven. I vividly remember the difference between being at the plate in the first inning and the last inning.
masoo From: masoo Date: June 9th, 2004 07:38 am (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

These are excellent points. I have a friend, Jonathan Bernstein, who I wish had a blog or something, because he's a great writer who also loves baseball ... he got his PoliSci doctorate at Berkeley, and I remember once checking in the office looking for something for John Brady, and there was a picture in the grad lounge of a PoliSci softball team, and there was Jon Bernstein. He's in Texas these days, but you can read him by the gobs in Usenet (the Giants newsgroup and rec.sport.baseball) ... Google him in the Google Groups section sometime.

Anyway, he has stated many times that what Charlie describes may well exist, but it gets filtered before the top professional levels. If you are one of those guys who "chokes," you don't get out of Double-A ball. When you watch the NBA finals, you are watching 24 of the very best basketball players in the entire world. My guess is, none of them are chokers. You or I might have "choked" when we were 16 years old and it was the ninth inning, but we're not in the NBA or MLB.

Here's the thing: you're down by one point, eight seconds on the clock, you've got the ball and a timeout. You set up a play. If it's me, I'm not concerned with who is "clutch" ... I am going to try and get the ball to the player on my team who is best at scoring, because that is the person who gives me the best chance of winning the game. I remain unconvinced that there exists Joe Blows who average 10 points a game, shooting 42% from the floor, but who shoot %57 from the floor and score lots of points "when it counts." And I don't believe in the opposite, that there are players who average 25 a game and shoot 49%, but "when it counts" they shoot 41%. Sometimes good players do poorly, sometimes lesser players do well, but if I'm Phil Jackson, I get the ball to Kobe because he's my best player, not because he's clutch.

Chris Webber is a choker, nonetheless :-).
cbertsch From: cbertsch Date: June 9th, 2004 07:14 pm (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

But see, because of the perception that the end of the game is the end of the game, the defense tightens (or the concentration, in another sport). I don't think shooting 40% during the first period in the regular season is the same thing as doing so in the playoffs. And I certainly don't think that shooting 40% in the first five minutes of a playoff game is the same as doing so in the last five minutes. Does that make sense? Stats won't show it, but getting a shot off like Kobe did last night takes work that it wouldn't take earlier in the game.
masoo From: masoo Date: June 9th, 2004 08:30 pm (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

I certainly don't think that shooting 40% in the first five minutes of a playoff game is the same as doing so in the last five minutes. Does that make sense? Stats won't show it, but getting a shot off like Kobe did last night takes work that it wouldn't take earlier in the game.
I know what you mean, and that's mostly just me being sloppy above ... numbers like "41%" were made up out of thin air to make a point.

But I think you're wrong that stats won't show us about relative difficulty. In fact, it's an easy hypothesis to check. Take every NBA playoff game in history, or limit it to X number of years in our time if you want to filter out the days when basketball was different, but get yourself a large sample of playoff games. Run the stats on every one of the games in your large sample, and find out the shooting percentages for all players in the first five minutes of a playoff game and the last five minutes of a playoff game. If, after doing this, you discover that shooting percentages go up or down in the last five minutes of a playoff game, you can then use this discovery to adjust your analysis of what an individual player does. If (and again, I'm making the numbers up, it's the concept we're after, not the actual data) Kobe shoots 45%, and the entire population of NBA playoff players shoot 39% in the last five minutes of a playoff game, and Kobe shoots 43% in the last five minutes of a playoff game, you might say "Kobe is good in the clutch because he shoots better than the average player in the last five minutes of a playoff game," or "Kobe is bad in the clutch because he shoots below his average in the last five minutes of a playoff game," but the truth if these numbers were true would be that Kobe's pretty much the same whenever he plays, averaged out over time: he's better than everyone else in general, and like everyone else he's a little worse at shooting in the last five minutes of a playoff game, but relative to the expected drop in shooting percentage, he's pretty much where you'd expect him to be.

It's like adjusting for Coors Field, or for guys who played baseball in the 1920s vs. the 1950s ... stats DO tell you how to adjust for different playing environments. You figure out how a player performed relative to his environment, and compare those adjusted figures to those of other players' adjusted figures. In this specific case, the environment is "last five minutes of NBA playoff action."
cbertsch From: cbertsch Date: June 9th, 2004 09:01 pm (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

Yes, I guess you were over-simplifying a bit in your previous hyothetical scenario. That makes sense.

The thing is, someone would have to run the numbers. I don't know who has done that or could.

But what would you say if the difference between Kobe and the average were considerably greater in the last five minutes than overall?
masoo From: masoo Date: June 9th, 2004 09:32 pm (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

I wouldn't be surprised to find that someone has already run these numbers. I don't follow basketball analysis as closely as I do baseball, but I know it's out there ... there's even a Basketball Prospectus ... and the "could" part is easy, anyone could, the necessary stats are probably available.

If the data showed a statistically significant difference between Kobe's relative performance in the last five minutes of playoff games and the rest of his playing time, that would be reason to advance your hypothesis and start figuring out why the difference might exist. And one explanation would certainly be that Kobe is "clutch." I can point to several areas that would need attention in this case, based on the kinds of studies that have been done in this area with baseball.

First, sample size matters. Barry used to get creamed for being a choker, based on a very small sample size ... people thought the sample size was huge, because he'd gone to the post-season five times, and his team had never won a series, but he'd had a grand total of 97 at-bats in that time, and that's not enough. Of course, in 2002 he put that stuff to rest, just like Reggie Jackson, "Mr. October," did in '77 after an undistinguished post-season career up to that point. So you'd want to be wary of assigning "clutch" value to Kobe based on a small sample size. Having said that, he's played 116 playoff games and counting, so that's probably enough.

Then you'd have to look for all possible explanations, not just be lazy and say "see, I told you he was clutch." If clutch has never been proven to exist, and Kobe showed signs of being clutch, you'd need more evidence still. You might start looking at how he did in the last five minutes of regular season games, or how he did in important regular season games, or stuff like that. You could corroborate your findings that way.

But really, it's most important to keep pursuing it, because there might be other explanations than "he's clutch," so you don't want to get locked into one explanation. Maybe it turned out that defenses double-teamed Shaq more frequently in the last five minutes of playoff games, leaving Kobe more open. I don't suppose that's true, but you'd want to check. I'm always reminded of the way baseball fans like to compliment players for doing "the little things" ... one of their faves is moving the runner from 2nd to 3rd with less than two outs, do that and the announcer will say "that's a good at-bat, that's how the game is played, he's a real gamer, that guy!" I believe a study was done some time ago that showed that left-handed hitters were more likely than right-handed hitters to move the runner over, because they pull the ball to the right side of the field. If that's true, we're no longer talking about some intangible ability of the hitter, we're talking about how you were born, which isn't the same thing.

If, after all of that, you conclusively demonstrated that Kobe Bryant (or anyone else) stepped up their game in the last five minutes of playoff games, me personally, I'd say "that's very interesting, where the hell is Kobe the rest of the game?" If a guy can crank it up another notch, a skill I am not certain exists, but if he can, why save it until the last five minutes? The Lakers were seconds away from going down 2-0 in the NBA finals ... right now, Kobe is the great hero, but what business does this team have floundering around that badly in the first place? Every minute counts, is what I'd tell the Lakers. You want to avoid having to rely on last-second miracles, then get those players to play their best for 48 minutes, not just 5.

As a fan of the dramatic gesture in sports, I love that Kobe hit the shot. If I was a Laker fan, though, I'd wonder why it had to be dramatic. And if the Lakers lose this series, it won't be due to lack of clutchness, it will be due to lack of consistency.
cbertsch From: cbertsch Date: June 10th, 2004 06:52 am (UTC) (LINK TO SPECIFIC ENTRY)

Re: Clutching At Straws

The thing is, though, that the why-not-do-it-sooner logic works much better for statistically average robots than for people. Human beings rarely perform at anywhere near their top capacity. We seem incapable of that sort of consistency, for the most part. I think the most compelling case for "clutch" players' existence would be to argue -- assuming the statistics bear out the end-of-game superiority we're talking about --that the Kobes of this world rise to the occasion, not because they were lazy before, but because they realize that their fellow players' perceptions matter and save their best for the time when others have a hard time delivering the goods. In other words, Kobe might be saving his energy for a time when Shaq is less useful and the Rick Foxes on the team will be less likely to shoot well.
8 comments or Leave a comment