What the "Stat Gurus" Say Does not Correspond with Reality
The "stat gurus" say that LeBron James is by far the best player in the NBA and that Dwyane Wade is--depending on which "stat guru" you ask--at least a top five player and quite possibly the league's second best player. "Advanced basketball statistics" suggest that James and Wade are "worth" a staggering amount of victories all by themselves--yet despite the fact that they are paired with a third All-Star (Chris Bosh) and a more than adequate supporting cast (including defender/three point shooter Mario Chalmers, spot shooters Mike Miller and James Jones, defender/corner three point shooter Shane Battier and energy/hustle players Udonis Haslem and Ronny Turiaf) James and Wade have only managed to lead the Heat to the fourth best record in the NBA, hardly much of an accomplishment considering the breathless predictions that a James and Wade duo teamed with Bosh would produce a 70 win season (in an 82 game schedule, which is equivalent to 56 wins in a 66 game season--a number that the Heat will not come close to matching). The Heat have been fading fast since All-Star Weekend--posting a 13-9 record--and unless they turn things around in a hurry they will fail to win the title for the second year in a row after James boldly promised to lead Miami to "not one, not two..." but multiple championships. At some point, the "stat gurus" will simply have to explain why the gaudy individual "advanced basketball statistics" that James and Wade post--and the tremendous number of wins that James and Wade supposedly directly add to Miami's total--do not translate into more regular season victories in the real world and, more importantly, why they do not translate into playoff victories against teams that allegedly have clearly inferior talent."Stat gurus" generally do not believe in the value of shot creation; they just crunch numbers and note how many shots a player takes and how many shots that player makes, regardless of whether the player is a one dimensional big man who only dunks the ball or the player is a multi-faceted wing player who breaks down the opposing defense by attracting one or more extra defenders. A "stat guru" thinks that a team of five Tyson Chandlers would be unbeatable because such a team would shoot almost .700 from the field while hardly allowing the opposition to get off a good shot; it never occurs to the "stat guru" that a team of five Tyson Chandlers would not be able to dribble the ball up the court without getting it stolen and even if the Chandler Five somehow managed to get the ball into a half court set that group would not be able to create the point blank shots that enable the real Chandler to have such a high field goal percentage. Chandler's field goal percentage is high because he is a limited offensive threat who fully understands his limitations; the vast majority of his field goal attempts are dunks and putbacks, so it is ludicrous to assume that he would be able to maintain his field goal percentage if the volume of his shots greatly increased (an NBA offense can only generate a limited number of dunks and putbacks); it is equally ludicrous to use Chandler's field goal percentage (or his "true shooting percentage" or any other measurement of his shooting efficiency) as proof that he is a top five NBA player (according to at least one "stat guru") when the reality is that Chandler's field goal percentage is an indication of his limitations, not his dominance. "Stat gurus" insist that Chandler has somehow transformed the Knicks when the reality is that the team was 42-40 last season without him and the team is currently 30-28 with him. Yes, there are other factors that affected both seasons but if Chandler had truly transformed the team then the team's record would indeed be transformed and that is obviously not the case, much as I predicted before the season, declaring "Tyson Chandler's departure will likely hurt the Mavericks more than it will help the New York Knicks." Chandler is unquestionably a valuable player but, while it is good for the New York Knicks that Chandler understands that he should rarely attempt a shot outside of point blank range, the fact that Chandler has no shooting range makes it easier for opposing teams to guard him (or to double team a bigger offensive threat while making sure that a rotating defender puts a body on Chandler to prevent him from getting to the offensive glass).
Ever since Pau Gasol arrived in L.A., I have noted that he significantly improved in two areas in which an established veteran player generally does not significantly improve: field goal percentage and offensive rebounding. Gasol's numbers in those categories went up because Kobe Bryant attracts so much defensive attention that Gasol has a free run to the hoop for point blank dunks and putbacks. That does not mean that Gasol cannot score or rebound if Bryant is not on the court--but it does mean that over a large enough sample size of games a clear difference can be seen between Gasol's Memphis production and his L.A. production. Similar reasoning applies regarding Andrew Bynum, though the sample size with Bynum is smaller because injuries have limited his playing time until this season.
One of the big headlines in the NBA on Wednesday night was Bynum's 30 rebound game versus San Antonio as the Lakers defeated the Spurs 98-84 even though Bryant missed his third straight game due to a shin injury. It is certainly impressive that Bynum posted such a huge rebounding total (even though he padded that number a bit by tracking down his own point blank misses) and the Lakers would be well served if Bynum consistently focused on defense and rebounding as opposed to worrying about the number/distance of his field goal attempts but the real story about the Lakers during Bryant's absence--the story that the "stat gurus" and their friends in the media will ignore--is that Bynum and Gasol's field goal percentages have cratered. The "stat guru"/media theory about the Lakers is that Bryant is an inefficient gunner who should stop shooting so much and pass the ball more often to Bynum and Gasol; my theory about the Lakers is that Bynum and Gasol do not fight for post position as aggressively as they should on a consistent basis but when they do venture into the lane Bryant gets them the ball--and they are wide open because Bryant generally has two or even three defenders on him.
Why does Bryant shoot so much more often than Gasol and Bynum if I am correct that he is not just selfishly gunning? Bryant ends up with a lot of "hand grenades," shots that he has to take because the shot clock is about to "explode." Fans and media members often do not understand just how important a factor the 24 second shot clock is in terms of NBA strategy; veteran college head coach/NBA assistant coach Hank Egan once told me that the shot clock is "a monster" and that "some of the things that you do in the pros are, exactly as you said, trying to force the offense to burn time and get them in the late stages of the shot clock." When Bynum and Gasol "trot" down court (to borrow Jeff Van Gundy's quite apt description) and fail to quickly establish post position the shot clock is ticking; many times the ball goes into the post, the opposing team traps and when Bynum or Gasol pass the ball back out to Bryant there is not time for a re-post and Bryant is left with a "hand grenade." If Bynum or Gasol had posted up earlier then there would be time for a re-post and it is always more difficult to trap a second time without leaving someone wide open. After a few times of being stuck with "hand grenades" Bryant sometimes tries to preempt the "hand grenade" scenario by taking a shot earlier in the shot clock when he only has one defender on him and can clear out some space. Former Lakers Coach Phil Jackson used to refer to this as Bryant "filling the vacuum," meaning that if Bryant's teammates do not step up and do their part then Bryant will assume greater responsibility; Jackson often linked Bryant's so-called volume shooting to passivity on the part of Bryant's teammates.
It should also be remembered that during this season Bryant has played through a torn ligament in his right (shooting) wrist, a concussion and the recent shin injury (he battled that one for a few games before resting)--and, despite being a 16 year veteran who has already logged more than 50,000 regular season and playoff minutes, Bryant is averaging over 38 mpg because the Lakers generally fall apart when he is not in the game. The quality of Bryant's shots--judged, as Jeff Van Gundy would say, while those shots are in the air and not based on the outcome--has not been bad but Bryant's shooting percentage has suffered due to injuries/fatigue. Bynum and Gasol are clearly deadlier offensive weapons than Chandler but my theory would still predict that their shooting percentages would go down sans Bryant, while the "stat gurus" would insist that Bynum and Gasol's shooting percentages function independently of anything that Bryant does (otherwise, the "stat gurus" would have to change their formulas to somehow reflect Bryant's contributions to those shooting percentages but since no stat exists to do this the "stat gurus" either have to stubbornly insist that Bryant makes no such contribution or else admit that their formulas are flawed).
In three games without Bryant, the Lakers suffered their worst loss of the season to Phoenix--a team that will likely miss the playoffs--and then they barely beat New Orleans, which is not only the worst team in the West but is shorthanded due to the absence of Eric Gordon, Jarrett Jack and Trevor Ariza. The Lakers' win against San Antonio not only featured Bynum's extraordinary rebounding effort but an extremely anomalous shooting/scoring performance from Metta World Peace, who poured in 26 points (a season-high and just his second 20 point game of the season) while shooting five of eight from three point range; Peace's previous season-high for three pointers made in a game was three (on three different occasions) and prior to this game he had been shooting less than .300 on three pointers for the season. The Lakers needed those points from Peace because Bynum shot just 7-20 from the field and Gasol shot just 9-24. The Lakers' "problem" supposedly is that Bryant does not pass the ball enough to his big men but in three games sans Bryant those big men have seen their shot attempts skyrocket while their shooting percentages plummeted; Gasol has averaged 23.3 FGA/g without Bryant and shot just .471 from the field, while Bynum has averaged 21.3 FGA/g without Bryant and shot .375 from the field. Yes, Bynum's percentage on point blank shots sans Bryant is worse than Bryant's much criticized field goal percentage for the entire season! Overall, Bynum is shooting .564 this season on 12.9 FGA/g while Gasol is shooting .507 on 14.1 FGA/g (keep in mind that those numbers include the past three games, so with Bryant on the court their shooting percentages were even higher).
The sad reality is that far too many "stat gurus" are not conducting scientific research to figure out who are really the best players in the NBA; they are producing proprietary numbers that enable them to sell books and/or to write articles for big name magazines and websites. If a "stat guru" admits the flaws in his methodologies then he is killing the goose that is laying his golden eggs. That is why no "stat guru" who has published books and/or articles proclaiming that he alone knows how to rank players can possibly acknowledge that he is wrong (a "stat guru" who works for an NBA team and is trying to use numbers to help that team gain an edge can perhaps modify his theories to fit new data but you will never hear about this because teams do not want to publicly release such information and thus give away whatever edge they think they are gaining). Then there are the media members who have various personal and/or commercial agendas that prevent them from ever backtracking from supporting the "stat guru" orthodoxy (I would also not discount the possibility that some of the media sycophants for "stat gurus" are actually morons who really do not understand anything about math or basketball).
Therefore, I do not expect that if I write the perfect refutation of a "stat guru" he will respond by admitting the error of his ways; I have written many such refutations already and I have also pointed out the very real problems and inconsistencies with the box score data that forms the basis for "advanced basketball statistics" but I have yet to hear from a single "stat guru" who is the slightest bit interested in modifying his theories accordingly. What I hope will happen is that the large number of people who do not have a personal and/or financial stake in the success of a particular "stat guru" will consider the evidence on its own merits, draw the proper conclusions and be able to watch basketball with a better understanding of the strategic aspects of the game.
Labels: "advanced basketball statistics", "stat gurus", Andrew Bynum, Kobe Bryant, L.A. Lakers, Pau Gasol
posted by David Friedman @ 5:41 AM
21 Comments:
excellent post david.
been an avid reader for a long time. you have great command of the english language and your game/player analysis are almost perfect. good balance covering both sides of a certain game situation almost all the time. and your knowledge of nba history and facts is really amazing. quite a treasure trove your mind is. keep it up.
jax
Did you read this article? http://chasing23.com/tyson-chandler-the-backbone-of-the-new-york-knicks/
Not really worth it. But, this site sometimes links to your articles. Anyway, the writer actually thinks Chandler is an elite offensive player.
You're right about Chandler. Even if there were 5 players that could get the same stats as Chandler, shooting around 70% and scoring 70+ points in the paint every game, which isn't feasible, that team would still not score very many pts., because they rarely shoot the ball. He shoots something like 6 FGAs/36 min. I don't know why people can't understand that just because his shooting pct. is so great that that doesn't mean he's that good as an offensive player. The only thing I can say about chandler offensively is that he's a good off. rebounder, but that has a lot to do with your teammates drawing defensive attention to allow you easier chances for off. reb.
It is interesting to see Pau/bynum greatly increase their FGAs, but suffer in their accuracy, but not surprisingly. And Bynum is shooting atrociously. I mean, they should be shooting a lot, because that's what the lakers need from them while Kobe is out, much like the lakers need Kobe to shoot a lot when he does play, but I also doubt any statheads will pay much attention to this.
I think most of the statheads are just stubborn more than anything, and won't alter from their already-formed opinions/formulas. Unfortunately, some of it is them trying to earn a buck/keep their job/have an agenda, but I do believe a lot of them are just clueless about basketball.
MWP had his best game in years, and the lakers still failed to get to 100. I think it was more of the spurs just playing awful. The lakers have done this all year. They look great sometimes, then played like the worst team in the league at other times.
Jackson888:
Thank you.
Boyer:
No, I have not read that article.
The Knicks have got to be the most overhyped .500 team in the history of sports. They may not even make the playoffs yet tons of ink and bandwidth are devoted to them.
Chandler is a very good but limited player. Prior to Dallas' championship last season he was perhaps a bit underrated but now that he plays in New York he may be on the verge of becoming overrated. Chandler is a very good defender who also sets solid screens and crashes the offensive boards but he is by no stretch of the imagination some kind of elite player. Chandler is the kind of player who can put an already good team over the top (like he did in Dallas) but he is almost superfluous on a mediocre or bad team, as we have seen throughout his career. Chandler's contributions have failed to lift New York much above the .500 mark. As I predicted before the season, his departure hurt Dallas a lot more than his arrival helped New York.
The amazing thing about yesterday's game is that Gasol and Bynum combined for 16 of 44 shooting, which is only a shade of a percent better than Kobe's legendary 17 of 47 game against Boston, which is still to this day talked about and brought up every time someone tries to frame him as the "selfish gunner". And that was indeed a bad game, but that's the point - they just matched one of his worst shooting games ever, yet this is never going to get the same attention...
Sometimes I get the feeling that one way of determining the quality of a player is to look at the standards that he is held to.
No one receives as much scrutiny and criticism as Kobe Bryant and Lebron James. If Derrick Rose goes 1-13 from the field in a critical game for playoff positioning, no one cares. If Dwayne Wade makes a series of mistakes to put the team in a hole then Lebron misses one of those "hand grenades" at the buzzer or passes out of a double team to set up a wide-open teammate, it's still Lebron's fault the team lost. If Kevin Durant shoots a series of contested 20-foot jumpers while shooting his team out of the game and the playoffs, he's just learning and he will head the next great NBA dynasty.
Maybe the endless criticism and scrutiny is a backhanded way of acknowledging that those two players, indeed, are the best the league has to offer?
Amen, once again, David. :-)
Only caveat I would point out is the sample size of the splits here. You have three games sans Bryant vs. and entire season prior. The bigs need to learn to play with the extra attention they are drawing, and it's possible that Sessions also figures out how to get them easier looks closer to the rim.
Anonymous:
While Bynum's sample size sans Bryant may be small, Gasol's sample size sans Bryant consists of six and a half seasons playing for Memphis before the Lakers acquired him. Gasol's field goal percentage and offensive rebounding rate both increased as soon he joined forces with Bryant. It is important to understand that most "stat gurus" completely disregard the scientific method; they not only disregard sample size issues but they do not test their theories at all or even provide means to reliably test their theories. "Stat gurus" provide player ratings listed to one or two decimal points with no mention of margin of error.
I agreed with your statistical insight but I didnt like the undertone of the article, the "stat guru" paradox isnt that widespread, "stat gurus" can accept the limitations inherent in their own formulas even while citing it in their columns.
For example Hollinger often comments about PER not doing a thing to capture positional defense, plenty of analysts talk about skillset overlap between players, with the ball and on the defensive glass.
And Ive honestly never heard someone say 5 Tyson Chandlers would win, except from Henry Abbot when he makes the same argument, you 2 must talk to the same people. Most of the stat guys I respect acknowledge the flaws and make subjective arguments to what the stats could be missing, others try their best to actually incorporate the value of shot creation and spacing but since it is by definition an intangible aspect of the game, there will never be a consensus on its value.
----------------------------------------
As for Kobe and the bigs, Ive wondered for years how they (mostly Bynum) would produce without him. Sadly Kobe has never taken a prolonged break for those injuries (Clippers fan here) so the only data we have is limited to specific seasons on/off court sample. I think there is some truth to the notion that Kobe needs to reign it in some and feed more but I do agree its not as simplistic as High FG% = more shots.
That said, I would love for Kobe to take some more time off.
Alex:
Hollinger's willingness, at times, to acknowledge some of the limitations of his metrics is unusual among "stat gurus."
The point about Chandler is not that a specific "stat guru" explicitly said that a team of five Chandlers would be unbeatable; the point is that such a conclusion is the logical extension of what many "stat gurus" assert about the value of "efficiency" versus the value of shot creation.
Abbott is one of the most vocal advocates for "advanced basketball statistics" but I think that anyone who is doing legitimate work in that field must cringe every time Abbott pens one of his disingenuous, foolish rants; Abbott's writing is so over the top that he actually discredits the very ideas that he is trying to promote--and it is puzzling that he repeatedly cites Berri, a "stat guru" whose work is widely disparaged by many other "stat gurus."
I have nothing against the attempt to better quantify what happens on the basketball court but I am very opposed to Berri's arrogant assertions (supported by his sycophant Abbott) that he has already figured everything out.
Bryant does not need to "reign (sic) it in" ; Bynum and Gasol need to post up more aggressively and the Lakers need to acquire (or develop) players who can consistently make defenses pay for trapping Bryant/sagging into the paint.
The other day Abbott posted about talking with Morey about late-game situations. Morey, a stathead himself, acknowledged what a luxury it would be to have someone who could create his own shot('hero ball') late in games. Abbott basically dismissed his claim completely. So, Morey does seem to understand what is actually going on at least to some extent. And at least with last possession situations, isolating one player is the best option, especially regarding the game clock.
One of many problems I have with statheads is that they act like the stats they're using are facts and definitive proof what is actually going on. Advanced statistics are leading to improvements in the game, unfortunately most of the people using/creating them are clueless about the conclusions they make from them.
Most of the advanced stats are just various formulas devised from the raw stats. I love my example here: say westbrook never accumulates any stat, except min. played. On every possession, he passes to a teammate who immediately draws a shooting foul, but never makes the basket
David,
I think the reason the Shaq/Kobe combo worked so well, was because Shaq knew how to post up. Obviously, that was also when the duo were nearing or at the top of their games, but I absolutely loved watching the big fella running down the court and heading straight to the post. If the post wasn't there, he'd go weakside then come hard across the lane to repost better than anyone else i've seen play the game. Despite all the criticism that he doesn't pass the ball, there's a reason Kobe's been the team leader in assists 13 of the past 14 seasons. He knows how to get the ball in the post as long as his bigs get good position. My buddy and I, as season ticket holders sit there at the games complaining about how the bigs, Drew and Pau, can't seem to figure out what good post position is all about. As you mentioned, if the bigs would only be more aggressive getting post position, you'd see the offense flow much better.
As for all the detractors, if indeed, Lebron and Wade are that good, why is it that they could only manage a bronze medal in international play? Only after Kobe, the "low percentage, selfish gunner" as some stat-gurus like to refer to him, did team USA regain it's position at the top of the medal stand. It blows me away that for all the stat-crunching those folks do, they have this tendency to overlook the fact that players have much better numbers play WITH Kobe than without. Also, as "team-oriented" as Lebron might be, his teammates generally fare worse across the board playing with him than without him. Something to think about.
Boyer:
Abbott tries to carry the banner for "advanced basketball statistics" but he is so misinformed and tendentious that he actually sets back his own cause. It does not surprise me that he would dismiss out of hand a suggestion made by Morey. Morey seems to be one of the more reasonable "stat gurus" but Abbott is so fixated on certain concepts/biases that he refuses to consider other points of view; that is one of the things that I hammered him about in my two satires of his work:
What if ESPN's Main Basketball Blogger Wrote About the Miami Heat the Way that He Writes About the L.A. Lakers?
What if ESPN's Main Basketball Blogger Wrote About the Chicago Bulls the Way that He Writes About the L.A. Lakers?
I mentioned him being laughed at by GMs but instead of taking this to heart he dismisses what they say. That sounds like exactly what you heard him do in that interview.
EdPak:
When Phil Jackson became the Lakers' coach, he insisted that Shaq get in shape and run down the middle of the floor to post up quickly right in front of the rim; if it was not possible to get the ball to Shaq immediately then Shaq went to one of the low post blocks and the Lakers ran the Triangle. Later, Shaq got a bit out of shape and did not run the floor quite as well, which became a source of contention with both Jackson and Bryant, but you are right that Shaq had a much better idea of how to post up--and he was much more aggressive--than Bynum or Gasol.
I only alluded to this earlier but I don't think "advanced basketball statistics" is really a science. It's just a method that people use in order to push their own agendas. It's a tactic that's used consistently by Truehoop and other "analysts" to "prove" their points.
Here are the major themes that I see:
(1) The "stats" are used to denigrate Kobe Bryant, especially after he has a bad game or fails to win the championship.
(2) The "stats" are used to elevate Lebron James, which is merely a setup for the trashing he inevitably receives after the Miami Heat loses a game or fails to win a championship.
Almost every article is some variant of the two. Once in a while they talk about something different just to mix things up but it always seems to seems to come down to some agenda the writers have with the two players.
As for the Lakers and Heat, the reality is that they are two of the handful of teams that have above a 5% chance winning the championship. Even if Chicago, Oklahoma City, or San Antonio win the championship, this won't mean that Kobe and Lebron somehow "failed", or that Derrick Rose, Kevin Durant, or Tim Duncan are now "better" than Kobe or Lebron.
Daniel Song 39:
As I mentioned in this article, there is an agenda behind much of what the "stat gurus" and their media sycophants say.
Barring injuries to key players, I would be surprised if someone other than Chicago, OKC, San Antonio or Miami wins the championship. Memphis is a decent dark horse pick but I don't believe that the Grizzlies will win it all if one or more of the aforementioned teams are healthy. I don't consider the Lakers to be a serious contender and would place their chances of winning it all at less than 5%.
David:
While I would be extremely tempted to bet on the Lakers to win the Championship at 19-1 odds, I do concede that they will probably sitting at home by the time the NBA Finals roll around.
As for Memphis, they are playing well this regular season but I would have to place their chances of winning the championship somewhere between remote and nonexistent. It would be nice to see them pull off an upset or two, however.
The Celtics would be an interesting dark horse if they could capture the fountain of youth. They almost did it in 2010 but they are 2 years older now - so it will be up to guys like Rondo and Bass to pick up the slack.
No matter who wins, I don't think reliance advanced stats will play a major role in determining the champion.
Daniel Song 39:
I don't think that Memphis or the Lakers will win the title but I think that Memphis is more dangerous in terms of having the capability to win a series against one of the four legit contenders. The Celtics are another team that could win a series against any of the contenders but I am not convinced that the Celtics can win four straight series.
I agree with you that the eventual champion will not rely heavily on "advanced basketball statistics"--at least not the kinds of numbers that Abbott loves to throw around.
David:
Now Adande has jumped onto the Andrew Bynum bandwagon and is claiming that Bynum is better than Howard.
While I concede that a healthy Bynum is better than a Dwight Howard who sits on the bench due to back spasms, Adande's suggestion is so ridiculous that it probably does not merit a response.
If Bynum's performance in the playoffs actually trump Howard's performances in his prime, and Kobe performs like an all-star, the Lakers will be the 2012 NBA champions. Safe to say that neither of us see that as even a remotely possible scenario.
Daniel Song 39:
"Probably does not merit a response"?
You are far too kind when someone says something that is extraordinarily stupid.
Post a Comment
<< Home