I don't live on analyzing NBA data, so no, I don't calculate it. But we've all seen enough times Kobe passed to Shaq/Horry, Gasol/Fisher for game winning shots. Wonder why that didn't happen for Smush/Kwame? 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.?[/QUOTE] I wouldn't use their lottery seasons to evaluate PP or AI's clutchness. As I said, use career numbers, that may be unfair to PP/AI, I get that, but that's how people evaluate clutchness.
If game winning shot percentage correlates so strongly with overall team quality that you can throw out the Laker's "awful years" (when they missed the playoffs a total of one time), why is Pierce's clutch shot percentage so much higher than Kobe's? Pierce's team was awful for much of that stretch, and overall the Lakers averaged 8 more wins per season for the four complete seasons considered here.
@durvasa: I think I've communicated my point and position fairly clearly and effectively. I don't think you can really quantify clutch play. I fully expect you to disagree since it seems you have this dogmatic view that everything in basketball can be explained through in-depth quantitative analysis. @JimRaynor55, @Easy, and @durvasa: Just answer this one question for me: If you had to pick one player right now, no matter what the situation called for, to have the ball in his hands during clutch moments, who would it be and why?
"If being "clutch" is indeed immeasurable (a sensible position), people should stop using language like "Kobe Bryant is the most clutch player in the league." If you can't measure "clutch", there's no sense in saying one player is more clutch than another. " "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. " How is that dogmatic? If someone's position is that you can't measure clutch, fine. Then just be consistent and stop trying to compare the "clutchness" of players. Otherwise, if you admit that stats can help measure clutch, then I think I made my position quite clear. I'll take LeBron James because he's the best player in the league. My reasoning is that's what my "feelings" tell me. I mean, isn't it obvious? (wow, basketball discussions are fun when you don't have to point to any actual evidence!) Who do you take?
I'd take most GM will take for last 7-8 years, and it's not even close. About 80%-90% GMs took Kobe in the survey in recent years.
I already said it: If you want to make an argument based on career numbers from a decade ago, then fine (I think that would be less relevant to now, but whatever). Do the extra work, pour over those hundreds of play-by-plays from every team from 10 or 12 years ago, and show us your numbers. Too lazy to do that work? Then don't dismiss someone else's existing work based on unverifiable assumptions that you don't have any supporting evidence for. Easy, LeBron. A better overall player, and much more accurate than Kobe in last possession shots given the data. If we're playing by your rules (i.e. ignore evidence and go by unsupported gut feelings) then I've already supported that statement more than I have to.
The irony is that a numbers-based argument can be made this season to support Kobe as a great clutch player. I suppose Kwame would want to ignore what follows as any sort of evidence (since stats can't be used to measure clutch), but based on it I'll say Kobe has done very well when his team needs a basket. The analysis I linked to here also supported the case that Kobe is one of the best clutch players. Below shows actions by selected players in 4th quarters and overtimes, when their team is down by 0-3 points. ast% and tov% is assists and turnovers per possession, respectively. pts/poss is points scored per possession scaled to 36 minutes at a 90 pace factor. <= 5 minutes Team player FGM FGA efg% FTM FTA assist tov poss TS% ast% tov% pts/poss CLE LeBron James 8 20 0.425 7 9 4 1 47 0.501 9% 2% 34.5 LAL Kobe Bryant 12 24 0.542 6 9 1 1 51 0.572 2% 2% 42.4 PHX Steve Nash 7 14 0.571 8 10 10 0 52 0.652 19% 0% 31.2 POR Brandon Roy 10 18 0.583 7 10 4 2 57 0.625 7% 4% 33.2 NOH Chris Paul 5 10 0.500 4 4 4 2 39 0.595 10% 5% 24.2 UTA Deron Williams 5 14 0.357 4 4 4 1 34 0.444 12% 3% 27.8 MIA Dwyane Wade 6 17 0.412 4 4 6 1 52 0.480 12% 2% 23.4 <= 2 minutes Team player FGM FGA efg% FTM FTA assist tov poss TS% ast% tov% pts/poss CLE LeBron James 4 11 0.364 7 9 2 0 25 0.501 8% 0% 40.5 LAL Kobe Bryant 6 12 0.583 4 4 1 0 24 0.654 4% 0% 50.6 PHX Steve Nash 3 9 0.389 5 7 1 0 19 0.497 5% 0% 42.6 POR Brandon Roy 6 10 0.600 0 1 1 2 23 0.575 4% 9% 35.2 NOH Chris Paul 2 5 0.400 3 3 3 1 20 0.554 15% 5% 23.6 UTA Deron Williams 3 7 0.429 1 1 2 0 13 0.470 15% 0% 36.3 MIA Dwyane Wade 4 12 0.375 2 2 1 1 26 0.427 4% 4% 28.6 <= 1 minute Team player FGM FGA efg% FTM FTA assist tov poss TS% ast% tov% pts/poss CLE LeBron James 4 10 0.400 3 4 0 0 14 0.468 0% 0% 53.0 LAL Kobe Bryant 4 9 0.556 1 1 1 0 16 0.583 6% 0% 46.4 PHX Steve Nash 2 6 0.333 4 5 0 0 10 0.488 0% 0% 54.0 POR Brandon Roy 5 8 0.625 0 0 1 1 15 0.625 7% 7% 45.0 NOH Chris Paul 1 4 0.250 0 0 2 1 9 0.250 22% 11% 15.0 UTA Deron Williams 2 4 0.500 1 1 2 0 8 0.563 25% 0% 42.2 MIA Dwyane Wade 3 6 0.583 0 0 0 1 13 0.583 0% 8% 36.3 <= 30 sec Team player FGM FGA efg% FTM FTA assist tov poss TS% ast% tov% pts/poss CLE LeBron James 4 10 0.400 2 2 0 0 10 0.460 0% 0% 67.5 LAL Kobe Bryant 3 6 0.667 0 0 1 0 12 0.667 8% 0% 45.0 PHX Steve Nash 1 4 0.250 2 2 0 0 7 0.410 0% 0% 38.6 POR Brandon Roy 4 7 0.571 0 0 0 1 10 0.571 0% 10% 54.0 NOH Chris Paul 1 2 0.500 0 0 0 1 4 0.500 0% 25% 33.8 UTA Deron Williams 1 2 0.500 0 0 1 0 5 0.500 20% 0% 27.0 MIA Dwyane Wade 2 4 0.625 0 0 0 0 7 0.625 0% 0% 48.2 <= 15 sec Team player FGM FGA efg% FTM FTA assist tov poss TS% ast% tov% pts/poss CLE LeBron James 1 6 0.167 2 2 0 0 6 0.291 0% 0% 45.0 LAL Kobe Bryant 3 4 1.000 0 0 0 0 7 1.000 0% 0% 77.1 PHX Steve Nash 1 4 0.250 2 2 0 0 6 0.410 0% 0% 45.0 POR Brandon Roy 2 5 0.400 0 0 0 0 7 0.400 0% 0% 38.6 NOH Chris Paul 1 2 0.500 0 0 0 1 4 0.500 0% 25% 33.8 UTA Deron Williams 0 1 0.000 0 0 0 0 2 0.000 0% 0% 0.0 MIA Dwyane Wade 1 3 0.500 0 0 0 0 4 0.500 0% 0% 50.6
Funny how many of these people who say, "I just know it" are the same people who cry about how unfair the all-star voting is. Hey a lot of fans just KNOW that McGrady is more deserving than Steve Nash. Who are you and based on what do you dispute THAT? The all-star voting is no more objective than a beauty contest. To say that "I just know Kobe is hands down the most clutch player by my own eye balls" is the same as saying "I just know that my girl is the hottest chick in the world." Yeah, sure. The other guy would say, "I know my girl is hotter than yours."
@durvasa: I don't think I or the person who wrote the article said that Kobe is the most clutch player in the league. Again, Kobe is being used as an example to demonstrate the flawed methodologies of the various quantitative analyses out there that purport to measure "clutchness." Are you disputing this? You say observational analysis is subjective judgment. Fine, but any dataset that tries to measure how clutch a player is also suffers from subjectivity and I've already explained why: ...the fact that statisticians who try and quantify clutch play come to different conclusions demonstrates the arbitrary nature of the quantitative approach. While on the surface it may seem "objective," it essentially boils down to which independent and dependent variables are included in a model (hence the subjectivity or arbitrary methodology). That could lead to a whole host of issues such as data manipulation, selection bias, omitted-variable bias, measurement error, etc... We all use stats in one way or another, whether we look at the box score, a players #s during the course of the season, and/or duration of his career. I know it's the new "in" thing, but I just hope you realize the limitations of the more hardcore quantitative approach to basketball. Don't get me wrong, it has its uses, but it's secondary to the more traditional observational approach in my opinion. It all boils down to what you do when there is a conflict between the two. If the quantitative analysis is saying so and so is more clutch, but your eyes are telling you that this other player is better down the stretch - what do you do? I would trust what I see as opposed to abstract and sometimes obscure statistics. The above also clarified my position regarding in-depth quantitative analysis vis-a-vis this issue. Maybe your emotions and the emotions of some of the other posters caused you guys to miss this. Stats are a supplement to observational analysis in my view and not the other way around. Still no answer to the question I posed in the aforementioned paragraph: It all boils down to what you do when there is a conflict between the two. If the quantitative analysis is saying so and so is more clutch, but your eyes are telling you that this other player is better down the stretch - what do you do? Robert Horry and Vernon Maxwell were clutch, but is there a dataset out there that would demonstrate that? None of this even begin to touch on some, for lack of a better term, intangible concepts that are part of clutch play that you can't even attempt to measure, such as leadership, ability to stay calm under pressure, toughness, determination, mental fortitude, and of course the "it" factor. There is no one comprehensive definition of "it", but we all know "it" when we see "it." Same with all those other variables that I just mentioned. How many times have we heard that so and so is a great leader, durvasa? I never once saw you jump in and demand that people should quantify their opinions. I guess you could ask the NBA to attach heart and blood pressure monitors to players to see if they rise or remain as is during crucial moments in games to measure a variable such as the ability to stay calm under pressure. But if we are to listen to you, we can't have an opinion if we can't use hardcore, in-depth quantitative/statistical analysis to justify it. In relation to the data you posted, it's arbitrary and the sample size is too small to be statistically significant. We haven't even gotten into a discussion of off the ball clutch play or what happens on the defensive end. Ultimately, there are too many variables, some that can't even be measured, that are a part of clutch play to be able to quantify it.
You keep saying the same thing. The only methodology referenced in the article was one used in a prior NBA.com article. That should not by any stretch be considered representative of all that stats can tell us about how clutch players have been or could be in the future. There is no set definition of what it means to be "clutch". If one is serious, they should think deeply about what that means, and think about which observable events are relevant to the discussion. Then take advantage of whatever evidence is out there to inform their evaluation. To me, this is only rational. Can we agree on that? I have no problems whatsoever in "observational analysis". That is science. But what is important is making clear what it is they are observing so that (a) we can ensure that it conforms with whatever evidence is available, and (b) so we have a rational basis for continued discussion. Your notion of "quantitative analysis" differs so much from mine that the very question makes little sense to me. If my eyes are telling me something, I need only record what I'm seeing on a piece of paper and ... presto! ... I now have a "stat" that is in agreement with my eyes. If there is a reasonable way to fit that observation, which I deem relevant, into my overall analysis, then I'll expand whatever "clutch model" I've devised to include it. What made them clutch? If there are facts to justify such a claim, then sure there could exist corresponding data. I remember Maxwell as a guy who hit a number of game winners over a stretch of a few season and who also was extremely competitive. He had some great playoff moments (e.g., against the Suns, on the road, in a must-win playoff game comes to mind). Horry's playoff resume is well established. These are things could certainly be reflected in the data. Leadership is something different. You can be a great leader without stepping foot on the floor. But to the extent we can observe leadership, it is also measurable. How often does a player come early to practice or stay late? That is measurable. How often does he take a younger player aside and give him helpful instructions or work with them in practice? That's measurable. How often does he make effort plays? That is positive leadership and also measurable. And I can go on. Anyone can have an opinion. But if its on a wishy-washy topic like "who's clutch?" and you don't bother and are unwilling to clarify what you're talking about, the opinion has very little value. But this is just repeating what I've already said a while back. Make use of the data that is available, fill in the blanks where it isn't available. Which of those should take precedence depends on how which criteria you consider most relevant. I do know that the data should not contradict observation. There may very well be a significant limitation in the data gathered, but if one observes one thing and the data says something different, then either your memory is faulty or whoever recorded the data made a mistake.
You still are evading the question: if clutch can't be measured, how can you say one player is more clutch than the other? Since it can't be measured, you don't know how much "Clutch" Kobe has right? Then how can you definitely state that Kobe is the most clutchest of them all then? Stats are based on what eyes can see, or objective evidence. If its leading you to different conclusions than its either you're looking at the stats the wrong way or you're just plain wrong.
THe problem with measuring clutchness is that part of being clutch is coming through under great pressure. And pressure is subjective. There is no objective measurement. There is no way to quantify pressure situations. Is it the last shot? Is it playoff games only? Is it the last minute? Is it an amount of expectation? Does a role player have as much pressure to hit a shot as a star player does? What about being the underdog vs. the favorite? What about if you are in the final year of a contract? Who feels more pressure? Jordan or Horry? Or Kobe? Or Wade? You see, it's qualitative. And just because something is qualitative or subjective doesn't mean it doesn't exist. Doesn't mean you can't say someone is clutch just because there's no quantitative measure. It's like saying you can't say it's a hot day or really sunny without having some qualitative measure. That's simply nonsense.
@durvasa: I'll let you have the last word since you're so obsessed with statistics that you feel you can quantify intangible concepts that are part of clutch play that you can't really measure, such as leadership, ability to stay calm under pressure, toughness, determination, mental fortitude, and of course the "it" factor. There's no point in even debating with someone who takes such extreme positions. From now on, however, if anybody on this site says so and so is a great leader, cool under pressure, and/or is clutch, I'm going to expect you to come in and ask them to quantify it. Again, you could also ask the NBA to attach heart and blood pressure monitors to players to see if they rise or remain as is during crucial moments in games to measure a variable such as the ability to stay calm under pressure. Maybe you shouldn't even watch the games and just run numbers through a simulator since you believe "everything" can be measured. @roslolian: I'm not evading anything and you have no idea what you're talking about. Here's what I said in the very 1st sentence of the post you quoted: I don't think I or the person who wrote the article said that Kobe is the most clutch player in the league. That negates your entire rant.
Sure. But saying someone is clutch, and saying someone is "hands down the most clutch player in the league" is different. The claim you made earlier in this thread demands something resembling actual evidence. If there isn't any evidence to support it, then that's a view I can't take very seriously. Its basically the same point I tried to make in my first post in this thread. If people want to say Kobe is clutch, fine. But when you start making statements like "Kobe is the most clutch player in the league, hands down" that implies you have some criteria by which Kobe dominates the competition. If you're saying something like that but can't come up with any such criteria, as far as I'm concerned "most clutch player in the league" is just a string of words without meaning.
I don't know what "it" factor is referring to and apparently neither do you, so I won't waste time debating that. As to the rest, if it is observable, it can be measured. To the extent you're able to observe a player's mental fortitude or his leadership, etc. -- those observations can be recorded as a data. I challenge you or anyone else to provide a counter-example. You should revisit the first sentence I wrote in this thread: "If being "clutch" is indeed immeasurable (a sensible position), people should stop using language like "Kobe Bryant is the most clutch player in the league." If you can't measure "clutch", there's no sense in saying one player is more clutch than another. " That is all. I would never take any issue with someone saying Kobe is clutch, cool under pressure, etc. But if you say something like "Kobe is clearly more clutch than LeBron James" or "Kobe is clearly more clutch than Brandon Roy" or "Kobe is hands down the most clutch player in the game" -- actually make a comparison -- then, yeah, I would expect some sort of argument/evidence to support that. You really think that's asking for too much? I watch games for the enjoyment, and because there's a ton of things that happen that are not being recorded very well, if at all. Computers are very good for storing and recalling information over a long period of time across thousands of games. But for real time processing of what's going on at the moment, you can't beat just watching the game. Our eyes and brains are very powerful machines too, with a different set of analytical strengths/weaknesses. I still believe there is a lot that can be mined from the data that is available, and in the future as far more intricate details are being stored in databases (imagine software that extracts information directly from game video -- its not as far off as you may think), computers will potentially give us even more insight.
You know, the author actually argues in favor of a clutch stat, upon reading it more carefully: [rquoter] Here's when we can really have a "clutch stat" -- when we have a mathematic/statistical equation that takes into account when a player like 'Bron or 'Melo is 4-for-16 until the last five minutes of a close game and then goes 4-for-5 from the field and 5-for-5 from the line. Or, when we place more weight on games against teams with winning percentages higher than .600; or when the "clutch stat" differentiates between a January game, a regular season "statement games" and a playoff game. Clutch is when Player X's team shoots 30 percent in the second half and in a "we're not losing"-type performance, scores eight straight points in the last five minutes.[/rquoter] He says no stat can measure clutch, but then proceeds to describe what he'd consider to be a useful "clutch stat". Two things I take from this: - put more weight on games that have greater playoff implications or may otherwise be considered "statement" games - give extra credit when a player's production increases as the possessions become more important The criteria he sets would indeed make for an interesting clutch stat. Not definitive (and I don't think there is a such a thing as a definitive clutch stat), but something that would inform the debate.
Huh? I don't need you to prove that Kobe is the most clutch, the mere fact that he can compare players' clutch ability already proves my point. From the article: "Because -- and no disrespect to Marion and his obvious talent and value -- he ain't Kobe. Kobe's clutch." ^That alone implies that Kobe is "more" clutch than Marion, yet his whole article discredits the stats which might show Kobe is indeed more clutch than Marion. How can he say that Kobe is "clutcher" than Marion when he doesn't know how to measure clutch? Because he saw Kobe make game winners? I'm sure if you go back to Marion's entire career he probably hit a game winner or two as well. You can argue that Kobe has hit clutch shots more frequently over a sustained period of time, but if that's the case then it should show up in stats, the same stats he says is worthless
Clutch is a myth IMHO. All possessions are worth the same, no bonus points are awarded at end game possessions so I don't know how you can quantify which possessions are more important.
If we neglect to take into account the running score, then I suppose it could be argued that all possession have equal value on average. But if you take the score into account, then clearly possessions are more valuable if its a 3-point game with 2 minutes remaining rather than a 3-point game with 42 minutes remaining. You could define the value of a possession as the % chance that the team would win the game if they score on the possession minus % chance they would win the game if they don't score.
