I mistakenly thought he said something to that effect. I still disagree with his attitude that you can't measure clutch and, at the same time, he can argue so strongly as to why current measurements are flawed. If you have a strong opinion on how clutch a player is, that means you should have a very clear idea of what it means to be clutch. And if that is the case, then you should be able to measure it. Wrong. Show me the dataset out there that declares other players to be superior to Kobe at shooting off the dribble, with defensive pressure, at the end of games. I have not seen any such stats, but if they exist and that's what it shows, then that could sway my opinion. Right now I would expect Kobe to be the best at that, though I wouldn't go so far as to say "no matter in any situation" (you added that yourself). Isn't it the case that we are all ignorant to some extent? Have they seen every single NBA game? Is every play fresh in their memory? Are they not more likely to remember the made shot that made Sportscenter rather than the missed shots that are glossed over? And, yes, existing methodologies are not perfect. That doesn't mean we should just throw out all the quantitative evidence out the window. Use the stats and use your subjective judgment to fill in the blanks where the stats are lacking. You need to understand what the quantitative analysis is really telling you. If you're defining clutch in one way, and the stats are defining it in another way, that is the source of the discrepancy. I don't consider that a limitation of quantitative analysis any more than it is a limitation of subjective judgment.
Kwame, define clutch for me. Boil it down to the essentials. How is this: The player I would regard as most clutch is the one I'd want on my team when the game is in the balance, especially when the stakes and the pressure is high (e.g., playoffs). Does that pretty much capture what we're looking for (at a high level)? Anything else I should tack on? Any caveats? If we can agree on that, then we can start discussion on what type of player is desirable in close games. He would need to be a player that is adept at creating good shots. Versatile. Good at drawing fouls (and making the foul shots). Etc. Edit: And let me just add one other thing. We should distinguish between players that have performed the best in clutch situations, and players that we expect to perform the best in future clutch situations. What we really care about is the latter category, right? For that, I think we need not limit ourselves to only look at crunch time stats. Weight that more heavily, sure, but how the player plays overall also is relevant in making predictions.
durvasa: I think all the article, and Kwame (an ironic nick to be discussing 'clutch' fwiw) are saying is that 'clutch' is about the intangibles. A 'you know it when you see it' argument. Simply, that you cannot fully judge a game by the boxscore, or the player by his stats. Basically...as you so adeptly put it....you need to assess it, in part, on a subjective basis. If your analysis of the dataset yields results which don't make sense -- the analysis (or dataset) may be flawed. He was taking issue with John Schuhmann's article that rated Kobe 38th on some clutch matrix. And he didn't say it couldn't possibly be measured -- see his paragraph on what would be needed to 'really have a clutch stat.' I think he was saying that there are just too many variables, and its too complex to narrow it down as simply as Schuhmann did. And I think I agree.
I agree with that as well. But I think its possible to "measure clutch" in a more meaningful way. The claim "No stat can measure clutch players" is not supported simply by discounting one person's analysis.
I was looking at 82games.com for numbers on a completely different subject (how much Rafer sucks :grin: ) when I happened to find this article. Background: -A "game winning shot" was defined as a shot taken with 24 seconds or less on the clock, with the ball-holding team either being tied, down by 1, or down by 2. -FIVE full regular seasons and playoffs (03-04 to 07-08), as well as half of the 08-09 season (the point at which the article was written), were analyzed. That's a huge period of time. -League average FG% on these "game winning shots" (as defined above) was 29.8% On to Kobe: -Kobe was the league leader in these "game winning shot" attempts, with 56 FGAs. Kobe's FG% on this type of shot was...25%, which was a very poor and only beat 4 out of 76 players who attempted these "game winning shots." Despite being the league leader in attempts, his low FG% resulted in him ranking just 4th in game-winning shots actually made (behind LeBron, Vince Carter, and Ray Allen, all of whom had much higher FG%). -Kobe doesn't pass on taking these last shots. He only recorded one assist in thes five and a half years under the defined "game winning" conditions. For comparison, LeBron had 6, Vince and Wade had 3, and Pierce had 9. Small sample size, I know. If you restricted the study to "game winning shots" made in playoff games (since people always want to dismiss regular season stat analysis by saying that playoff games are what really matter), Kobe is indeed #1, though tied with LeBron with 4-8 shooting. Dirk was #3 with 3-6 shooting. Pretty small sample size, if you ask me. This article predated the second half of 08-09, or the first half of 09-10, but Kobe's "clutch" reputation preceded that end point, and five and a half seasons is a massive scope. Kobe is an amazing player, one of the best in the league. His "clutch" reputation is overrated though.
They're ignorant. Most people do not see even 2% (That's about 50 games) of the regular season games played league-wide in a single season, yet they feel comfortable to judge one player above the hundreds in the league? Without even attempting to count the kind of shots that they're forming their judgment on? Being inside the league, or the media (and the media essentially means nothing, since the only qualification is to be able to write or talk in an entertaining way) doesn't mean you're immune to error. Look at all the bonehead players out there. Look at all the GMs who destroy their team's cap by overpaying mediocre players. Look at all the coaches who give minutes to blatant scrubs (when they're not getting fired, NBA coaching jobs have high turnover). This isn't to say that I, or any other average person, could do their jobs better if I were inserted in their place. This is to point out that they make plenty of mistakes too, so their opinions shouldn't be taken as evidence without any examination. Stats do not lie, it's just people that do. The true meaning and scope of a set of numbers is discernable as long as you actually take some time to actually look at how the data was compiled. You criticize stats...but you think "traditional" alternative of using your eyes is better? That method is essentially being lazy and not putting in the effort to actually count and compare a large amount of data. Again, many people might not even see 2% of the league's games in a year. Are you absolutely sure that you're not favoring the players who get nationally-aired games? Are you paying equal attention to every player in the games that you are watching? Do you actually keep count of things (more than the hardcore stat guys do) instead of just forming a general impression in your head?
I wonder why 03-08? Kobe was in the league since 96. He didn't practice that much in 03 and 04 offseason due to lawsuit, had worst shooting fg% in 03-05. He didn't get decent help until late 08. He didn't like to pass to Kwame and Smush, but he trusted Fisher and Gasol. You are always more willing to pass in a good team, and you'll be more open too. Kobe had worst career shooting fg% in 03-07. Ain't these things supposed to be tightly related to game winning shots as well? I think the author of this article is biased by using selective stats. It would be convincing if that's career number instead.
2003-2009. That is five and a half seasons, a huge sample size. Kobe was in the league since 1996? Well, it's not all about Kobe you know. The original article was written in 2006, and went back two years to 2004 for a more significant sample size. This article expanded on that 2006 article by bringing it up to date to 2009. Are you willing to go to such lengths to make excuses for the other players on the list? LeBron, Carmelo, and Wade were rookies in 2004. Vince, Pierce, Ray, Joe Johnson, and Iverson were on some awful teams. Gilbert basically missed all of 2008 and 2009. Some of these guys are a few years older than Kobe, and reached the wrong side of 30 earlier than he did within this time frame. Some of these guys were traded, some of these guys were injured, etc, etc. I could go on. I think you're making excuses for Kobe. Kobe not practicing much in one summer is supposed to be a big deal in a study spanning half a decade?
The question is who is the most clutch today, right? Wouldn't data from recent years be more relevant than data from a decade ago?
This is ridiculous We always evaluate a player's clutchness based on his entire career. MJ was 1-9 in playoff for his first 3 years and didn't make playoff for his 2 years in Wizards, that's 5 years, half a decade too. Do we evaluate MJ's playoff success because this sample is big enough? LBJ etc. were in the league since 03, fair enough, that's his career numbers. Compare career numbers to career numbers.
The numbers just shocked me. I knew that Kobe was overhyped from reading the "Clutch Play" sections of 82games.com before. But I didn't expect to see anything close to the stats that I saw when I first clicked on that article. Kobe was near bottom of the barrel in last or near-last possession accuracy with the game on the line. Of the 76 players who took such shots in those five and a half years, Kobe was only more accurate than 4 of them. Of course he took a ton more shots than most of those guys (some of whom were in single-digit attempts), but Kobe didn't hold up well compared to other high-volume star players either. Those other stars managed FG% in the 30s... Of course Kobe is probably a better overall player than all but maybe a couple of those guys on that list. But the numbers suggest that he's a terrible last shot option.
OK, how about pick this season or the last 30 days coz they reflect current status better? Didn't Kobe hit 3 buzzer beaters and no one else comes close? If the author was not biased, I am certainly not biased here either.
Do you have problems with the flip-side of that, which is cherry picking MJ's championship years? Because that's what most people do. No one has compiled the data on game-winning shots from 1996 to 2003...But the work has already been done for the recent half decade. The numbers for that period (which is the period in which Kobe's "clutch" reputation took off) show Kobe to be overrated. Not satisfied with that? I would have no problem if you did the work and analyzed hundreds of game winning shots from the late 90s as well. Until that work is done, that period is a question mark. What we do have is the article I linked to, which is far better than making assumptions about a time period that no one has the numbers on.
I think defining, specifically, how clutch someone is as difficult as defining how good someone is. Now, you can measure points, shooting percentages, rebounds, etc in a given time period but I wouldn't say that is the gist of clutch. Of course, many great basketball minds have said they would prefer to have Bryant in what they deem the clutch.
It wouldn't be biased to just present the facts. Kobe was really hot when it came to buzzer beaters...for the recent month. Not disputing that. Now, would you seriously argue that that is more representative than the last 5 years?
Man, you are getting more and more desperate defending Kobe. I don't know about Kobe's numbers in his earlier years. What I remember is that Kobe as a rookie was nowhere near Jordan as a rookie. Not a knock on Kobe. He's a lot younger. But if you insist on including Kobe's early years stats, I am quite sure that it would LOWER his career average, let alone those years of his prime that were used in the article JimRaynor55 posted.
I just show that the fg% of game winning shots are highly positively correlated to team quality. And since that 5 years were among the worst for Kobe in terms of team quality. It's a bad indicator if you want to use it to evaluate his overall career. Simply put, you shoot better in better team. Game winning shot or not.
On the contrary, he's going to be much better coz he wasn't the go-to guy. If accuracy was the only indicator for clutchness, Kerr easily beat Jordan if you compile the data.
Did you actually calculate the correlation, or are you just stating your impression? Kobe's 2004, 2008, and 2009 teams were terrific. His 2005 team was lousy. His 2006 and 2007 teams had 45 and 42 wins, respectively. Average, not bad. Be fair. Are you willing to make the same team quality excuse for Pierce, Iverson, etc.?
