New York State of Mind, Part II
Mike K. of Knickerblogger vehemently disagrees with my recent article about the New York Knicks. There are many interesting issues to examine here and I will get to the basketball related ones shortly but first I want to touch on a different subject, namely who "owns" a piece of creative work? That question does not refer to copyright ownership but rather the issue of how a work is interpreted. Does a work mean what the creator says it means or once it is published does it mean whatever various readers interpret it to mean? For instance, Bruce Springsteen intended "Born in the U.S.A" to be a protest song about the suffering of Vietnam veterans but it was widely interpreted to be a patriotic anthem (not that the two ideas are mutually exclusive but the primary message that Springsteen meant to communicate is not the one that many people received). Should an artist, writer or performer explain what his art means to him or simply accept that once his art is viewed by the public it could mean many different things to different people?Some people respond politely but briefly to criticism, some people fly off of the handle and resort to making ad hominem comments and others ignore criticism completely. I've often been advised to do the latter and at times my reluctance to do so is interpreted to mean that I enjoy confrontation, which is far from the truth; I neither enjoy confrontations nor do I shirk them but I despise ignorance and misinformation, so my desire to correct those things inevitably leads me into the fray.
In other words, I think that the article I wrote about the Knicks is pretty straightforward and anyone who is intelligent and reads it with an unbiased, open mind can see that what I said is measured and logical. Here are rebuttals to Mike K.'s specific criticisms (his comments are presented in italics):
The author, David Friedman, talks about the Knicks hot start and recent cool down. He shows D’Antoni’s current win percentage to be similar to Isiah’s first year, which I thought was odd since it’s more relevant to use the previous season. However I let it slide because Friedman promised to use “several key statistics” to prove his point.
The reason that I compared D'Antoni's first season to Thomas' first season is that D'Antoni has only been in New York for one year, so by definition no second season comparison can be made. In Thomas' first season, the Knicks improved from a .280 winning percentage to a .402 winning percentage--and then the next year they dropped right back to .280. When D'Antoni has completed a second season in New York then we can compare his two years with Thomas' two years. My point is that despite all of the buzz about the Knicks they have yet to prove that they are better than they were initially under Thomas. Sometimes coaches have a honeymoon period and then the players tune them out. Injuries and roster moves also can obviously affect the won-loss record. As I wrote in my original article, "Quick quiz: Do the Knicks have a higher winning percentage this year in Mike D'Antoni's first season with the team than they did in 2006-07, Isiah Thomas' first season as New York's coach?" I think that there is a perception that the Knicks have made a complete about face under D'Antoni. All I am saying is that it is too early to know whether or not that is really the case; I also think that it is worth noting that the Knicks have been getting worse as the season progresses, particularly on defense.
Long time KnickerBlogger readers will know that Friedman’s choice of per game stats is a poor choice of rating a team’s ability. By using per possession stats, we can see that the Knicks are currently 15th & 23rd on offense and defense respectively. This is a clear improvement from last year's team which was 23rd on offense and 29th on defense. It’s true that the rebounding has slipped, although the Knicks are better on the defensive glass. Although it’s not true that the team is worse off in shooting percentage. Using eFG we can see that last year the team had a shooting differential of -4.3%, which has risen to -1.9% under D’Antoni.
Now we get to the crux of the problem: I cited actual real world numbers--New York's winning percentage, plus the team's scoring average, points allowed average, point differential, field goal percentage differential and rebounding differential. In contrast, Mike relies on "advanced" numbers to "prove" how much better the Knicks are. The problem is that in the real world of wins and losses, the current Knicks are no better than Thomas' Knicks were in his first year--and even though this year's team is winning more games than last year's team, the Knicks rank dead last in field goal percentage differential, have gotten markedly worse on the boards and have not improved their point differential ranking much, though the raw number has improved from -6.6 to -2.6; the reason that I emphasize the importance of ranking over the raw number is that if a team's differential ranks near the bottom of the league then no matter what the raw number is that team is not going to have a good record relative to the other teams.
The author moves from talking about the Knicks to an overall indictment of D’Antoni’s style of play. He accuses the coach of “neglecting the defensive end of the court” (something that was refuted by Kevin Pelton earlier this year) and launches into a defense wins championships attack on D’Antoni. His proof is that the “[Chicago Bulls] consistently rebounded and defended well… en route to the 1996 championship the Bulls won seven of the eight playoff games in which Jordan shot .440 or worse from the field.” It’s true that those Bulls teams played great defense, but let’s not forget that they were fueled by their offense. Of their 6 championship teams, Chicago was ranked #1 on offense 4 times and #1 on defense only once.
Pelton's article about the Knicks--which was written in January, before New York's recent swoon when the Knicks have regularly given up 100-plus ppg--does not "refute" anything and I will address that in a moment but first let's get rid of the notion that he is some expert on NBA defense. Pelton claimed early in this season that the Lakers were using some kind of revolutionary new defensive scheme; the reality is that the Lakers had a favorable schedule (10 home games in their 14-1 start) and that most of their so-called "new" scheme merely represented the coaching staff's attempts to take advantage of the length/quickness of their personnel in fairly standard ways: when I asked Lakers assistant coach Jim Cleamons what new things the Lakers were doing defensively, he told me, "The only thing we’re doing is what a lot of teams have decided to do: basically, playing a man to man defense that is actually a zone; we’re sending an extra defender over in situations that we feel threatened. There’s no big secret about it; that’s what we’re trying to do: give more help when we can and we’ve been fortunate thus far." When I followed up by asking Cleamons to compare the current Lakers' defense with the 1996 Bulls championship team (for whom he was also an assistant coach), he replied, "That (Chicago) team had a certain chemistry in that they knew how to help. That’s why we have gone to the scheme we are using this year: guys don’t know how to help—when to come over, when to get out. If these guys understood that schematic then we wouldn’t have to change up. We would have just gotten better at what we did" (emphasis added). In other words, the truth about the Lakers' defense is precisely the opposite of what Pelton wrote: the so-called "new" scheme was not some kind of defensive revolution but rather the coaching staff's attempt to organize the defensive efforts of some players who do not have great defensive instincts. As the Lakers have faced stronger teams and more road games, it has become apparent that their defense is not as great as Pelton suggested.
A couple months after my first interview with Cleamons, I caught up with him again and asked him to give a "mid-term" report card on the Lakers' defense. He said, "Anyone who watches film and is a student of the game would see that we don't play with the same intensity day in and day out, game in and game out. If you are going to be a championship caliber team, your defense is the one area that doesn't waver. We aren't good enough on a game by game basis to do what we need to do to say that we are going to be accountable in the end. Then, our rotations are not always what I like to call 'on point.' Sometimes, they are nonexistent, sometimes they are a little bit slow. If you are a good defensive team, then you play better on the defensive end than you do on the offensive end, because that (defense) is where you are really linked together; (in that case) the team has a feeling of when they have to help and a sense and a presence of how they need to get there so that when the ball moves and flows your defense is not always reacting. You are kind of ahead or you arrive right on the catch so the offense knows that you are there and there are no gaps in your rotations."
I did not initially call Pelton out by name regarding the Lakers but merely refuted what he wrote by publishing those interviews. However, the idea that Pelton's Knicks article "refutes" what I wrote simply cannot go unanswered. The assertion--by Mike K., Pelton or anyone else--that the Knicks are either good defensively this year or simply better defensively than last year is demonstrably false and you don't need so-called "advanced" numbers to prove this. I already mentioned the Knicks' poor point differential--the raw number is better this year, but the ranking is nearly as bad. The Knicks are last in field goal percentage differential (the difference between what they shoot and what their opponents' shoot); granted, some of that has to do with their own shooting percentage but they are also 28th in defensive field goal percentage. Last year, they were last in field goal percentage differential and 28th in defensive field goal percentage. Those numbers are terrible and when you combine that with the Knicks' rebounding woes the overall defensive picture is ugly. The Knicks have also dropped from 22nd to 28th in points allowed. I know that all of the "stat gurus" will scream that per game stats are not as important as pace. It is true that the Knicks are scoring at a faster pace but they are also giving up points at a faster pace, their point differential is still deep in the red and the shooting percentage that they concede is awful. Good defensive teams strive to hold teams to under .450 shooting and the Knicks are not even close to meeting that standard.
By the way, the so-called "advanced" numbers are viewed with derision by Cleveland Coach Mike Brown, who told me that the numbers he looks at most are "Opponent’s field goal percentage, first, and then opponent’s points, second, but the opponent’s field goal percentage is a big thing for me." Realizing that "stat gurus" mock the importance of total points, I asked Coach Brown about that again and he confirmed that he looks at total opponent's points even more than differential. He wants the Cavs to hold their opponents to fewer than 90 points on sub-.400 shooting and fewer than 20 free throw attempts--to defend without fouling. He stated plainly that he is not a numbers guy overall and he said that Gregg Popovich--who he served under as an assistant coach for the 2003 champion Spurs--"is not a stat guy either." When I mentioned to Coach Brown how much Houston GM Daryl Morey relies on stats, Brown replied with the line of the year about basketball statistical analysis: "Not to knock that, because I think it is great to use if you have some solid information, but how many championships has that gotten them?" Coach Brown later added, "It’s a thing that, yes, if you use it the right way it can be helpful, but if you try to use stats too much I don’t know if it’s going to bring you a championship, at least from what I’ve experienced. We didn’t need those types of detailed stats to win a championship in San Antonio."
The bottom line about D'Antoni's Knicks defensively so far is that they give up too many points, allow opponents to shoot too good of a percentage and they get killed on the glass.
Regarding Mike K.'s comments about the 1996 Bulls, they obviously were great on offense and defense; that is why they won 72 games. The important point that far too many people neglect to understand is that the great teams play defense on a game in, game out basis and that forms their foundation, because even the best offensive players and offensive teams can have off nights. As I noted in my article, the 1996 Bulls were able to win in the playoffs even when Michael Jordan had bad shooting games. The Bulls' high offensive rankings would not have meant a thing in those playoff games without their great defense; if you shoot poorly and lose a game seven because you cannot rebound and defend, all of the "advanced" numbers in the world won't make you a champion.
The relevance of all this to the Knicks is that there is a ceiling to how good the Knicks (or any other team) can be if they do not meet certain standards defensively in terms of point differential, defensive field goal percentage and rebounding. My article Recent NBA Champions by the Numbers discusses how much championship level success correlates with outstanding performance in those categories.
You can excuse Friedman for using archaic stats, but what’s not excusable is how he cherry picks the facts to support his argument. He specifically picks Isiah’s first season to compare with D’Antoni, because the numbers are much closer (.402 to .406) than comparing D’Antoni’s improvement over last year (.280 to .406). You have to wonder if he spelled out ‘fourth’ because saying the team improved from 21st to 4th is easier to process visually. And take for example his paragraph on the Knicks where Friedman ignores one key piece of evidence: point differential. By using points per game, he shows that the Knicks have improved by 4 points over last year (from -6.6 to -2.6). However this significant change is swept under the rug with “[it's] only a few spots better than last season.” You get the feeling that Freidman made up his mind long before he checked the stats out. As a statistical sports blogger, I get a lot of readers new to the field that have a general distrust of numbers. Statistically dishonest articles like Freidman’s helps to reinforce this skepticism, and are a disservice to all sports writers.
First, I have to enjoy the irony of being lectured to about proper journalism by someone who repeatedly misspelled my name. As for the stats that I cited being "archaic," Popovich and Brown would disagree. When the NBA's top defensive coaches evaluate their teams they look at exactly the numbers that I cited about the Knicks--that is why I cited those numbers in the first place. I'm not trying to look cool to the "stat gurus"; I'm trying to explain how and why basketball games are actually won and lost. Proclaiming that the Knicks are actually good defensively or that the Lakers have a new defensive scheme may get you "street cred" (or geek cred) in certain quarters but those are demonstrably incorrect statements.
Saying that I "cherry picked" facts is also false. As I stated above, I compared D'Antoni's year one to Thomas' year one simply because D'Antoni has not yet had a year two in New York. All D'Antoni has done in his first year is a "turnaround" comparable to the one that Thomas did after Larry Brown's tenure. Maybe the Knicks will win 50 games next year and maybe they will regress to a .280 winning percentage, but we don't know now what will happen; the late season decline certainly cannot be encouraging to Knicks fans. I understand that the Knicks are trying to position themselves for the future in terms of the salary cap but if the team is lousy and does not employ the defensive minded philosophy that is proven to win championships then why would a max level player--who can get max level money from several different teams--come to New York? My article about the Knicks does not deal with hypothetical scenarios but simply details how the Knicks are performing this year.
Also, I certainly did not hide from the reader the fact that the Knicks have a better record this year than they did last year. I wrote, "Of course, Knicks' fans surely remember that in Thomas' second season as New York's coach, the team's winning percentage dropped to .280, which is exactly the winning percentage that Larry Brown had in his only season as New York's coach (in the year prior to Thomas taking over the coaching duties); the hope/expectation in New York is that the Knicks will steadily improve under D'Antoni and not regress the way that they did under Thomas but that is why the team's late season collapse should raise eyebrows: Several key statistics suggest that the Knicks are what their record says they are--a lower tier team, albeit one that now plays at a much faster pace." If I were "cherry picking" numbers then I obviously would have left out the direct reference to last year's record. The previously discussed recitation of defensive stats followed that paragraph.
The comment about spelling out "fourth" just betrays Mike K.'s ignorance of basic journalistic practices; numbers less than 10 are generally written out as words. To suggest that I chose that formulation for some nefarious reason is silly.
Obviously, my article was neither "dishonest" nor a "disservice." Those terms much more accurately apply to people who are so blinded by either their support of a particular team and/or their nearly religious faith in "advanced" numbers that they are unwilling to even entertain an alternative viewpoint. Compare how Mike K. concluded his screed with the measured tone that I used to conclude my article:
"The Knicks have not had a winning record since 2000-01. They have been bad for a long time and it may take a while before they are good again. No one should rush to judgment after D'Antoni's first season with the franchise but there are two interesting dynamics to watch with the Knicks, namely what roster changes new team president Donnie Walsh makes in the next year or two and whether or not D'Antoni is willing/able to coax a better defensive performance out of this team.
'Defense' may be a four letter word to D'Antoni but if the Knicks want to spell a certain 12 letter word --'championship'-- for the first time since 1973 then defense will have to become a part of their collective vocabulary, as should be obvious by watching the teams who currently sit atop the Eastern Conference, Cleveland and defending NBA champion Boston."
Mike K. declares that I "cherry picked" numbers in a "dishonest" attempt to tell a biased story but the reality is that I simply cited the relevant numbers regarding the 2007, 2008 and 2009 Knicks, indicated that the D'Antoni Knicks have yet to surpass the level that the Thomas Knicks reached in 2007 and suggested that the Knicks need to make personnel and philosophical changes in order to become a good team.
Labels: Isiah Thomas, Mike D'Antoni, New York Knicks
posted by David Friedman @ 12:03 AM
34 Comments:
I think the 4 point swing in their point differential is huge. It might not matter in the standings, but it has more to do with the other eastern conference teams getting better. These factors are out of the Knicks control.
More importantly, the 4 point swing is a small but important step in creating a winning culture. Remember when Isiah caused some ruckus about opposing team running up the score or something?
D'Antoni didn't just get praises for the early season success against crappy teams, up until they gave away Crawford and Randolph, they were actually in the playoff discussion.
Up until they gave away their 2 most talented players, I think that the Knicks this season were more successful than during Isiah's first year.
I think that after they gave away their 2 most talented players, they are in a better position to improve themselves than a couple of years ago.
So if you asked me would I rather have the Knicks now or the Knicks from a couple of years ago, it's a pretty easy call to make.
They got rid of Marbury too!
Z
Z:
Addressing your last point first, getting rid of Marbury is definitely a plus; that is a move that I had been advocating for a long time.
The four point swing in point differential has some significance but that only moved them up a few spots in the rankings; the Knicks are still securely in the bottom third of the league in that category and if they stay in that area they will have a difficult time making the playoffs even in the East--not just this year, but any year.
In 2007 under Isiah, the Knicks started out 26-32 before collapsing down the stretch to finish 33-49. This year, the Knicks were 24-34 after 58 games and have since gone 4-8. Obviously, there are differences between the 2007 and 2009 teams in terms of the salary cap situation and so forth, but my point is simply that in spite of all of the buzz the Knicks have yet to accomplish anything under D'Antoni that they did not accomplish under Isiah. Clearly, that could change, as I readily acknowledged in the conclusion of my article--but there is no reason for anyone to get as bent out of shape as Mike K. did in response to an article that simply stated the facts objectively.
There are two subtexts to this discussion:
1) Isiah Thomas became such a polarizing figure in New York that the team's fans are quick to criticize him but much slower to objectively consider the strengths and weaknesses of the current Knicks team.
2) The stats that I used in my article are not considered to be "advanced," even though point differential, field goal percentage allowed and even total points allowed--one of the stats that "stat gurus" hate the most--are stats that NBA coaches use to grade their teams, as noted in my interview with Mike Brown, who is one of the top coaches in the NBA.
Also, the whole contretemps about me comparing 2009 to 2007 is just a red herring, because I plainly noted the Knicks 2008 record in my article and when I compared the defensive stats I compared 2009 to 2008, not 2007.
No matter how anyone attempts to spin things, D'Antoni's year one strikingly resembles Thomas' year one and the Knicks defense and rebounding have regressed since last season.
Which Knicks team you would rather have--2007 or 2009--is not the subject that I discussed in my article. As I pointed out, D'Antoni deserves an opportunity to try to improve the team's defense and Walsh deserves an opportunity to upgrade the talent level. All I am saying is that neither of those missions have been accomplished as of yet.
All I know is with the great there is no D in DAntoni the Knicks have only won 5 more games than last year. Thats a disgrace. They let a Kings team who was 0-28 vs East coast teams win by 30 in the Garden. The Knicks scored 44 last night in the first quarter and then gave up 140 to the Clippers (who lost 9 straight road games)and some no named guard score 35. The Knicks continue to be a joke and its a shame what happened to New York basketball. St Johns is also terrible.
No matter how anyone attempts to spin things, D'Antoni's year one strikingly resembles Thomas' year one and the Knicks defense and rebounding have regressed since last season.
By saying this, you are clearly not watching the games. You (and Mike K) can spout out all the stats you want. But by saying these knicks resemble any of the knicks that Thomas coached just proves you have not been watching. This team is very different. The results have not yet come... as most good things take time... but they will. This is a different team.
Which Knicks team you would rather have--2007 or 2009--is not the subject that I discussed in my article. As I pointed out, D'Antoni deserves an opportunity to try to improve the team's defense and Walsh deserves an opportunity to upgrade the talent level. All I am saying is that neither of those missions have been accomplished as of yet.
Well... the bailout hasnt worked yet either. In the 1960's when they started the space program, it would have been asinine to suggest 1 year in that since they have not yet shot someone into space nothing has been accomplished. Talking to a second year Med Student and saying that they are not yet a doctor so they havnt accomplished anything is also just as silly. Saying that 3/4 (or maybe 4/5) way though season 1 that Coach D and Walsh has not yet accomplished anything is just the same.
I'm not saying give D'Antoni the CoY award, but given what he has to work with, I think he has done a pretty good job.
They were hovering around .500 before they gave away their 2 most talented players so I simply cannot point the finger at D'Antoni for their "collapse."
As I pointed out, D'Antoni deserves an opportunity to try to improve the team's defense and Walsh deserves an opportunity to upgrade the talent level. All I am saying is that neither of those missions have been accomplished as of yet.
Agreed.
There's a slight problem with that though... Walsh's idea of improving talent was to get rid of as much salary as possible. In a strictly basketball sense, it makes no sense sure.
So it's difficult to label this team as a success or not, since clearly, their priority was to tank...er...get rid of as much salary as possible.
Did Walsh upgrade the talent? No. But they are in a better position to do so next year.
Did Mike upgrade their defense? No, but you have to remember how the Suns "defended." Their threat of the fastbreak prevented opposing teams from exploiting their rebounding weakness too much. Their pace also disrupted what the opponent was trying to do. Their shooting prowess ensured that the opponents would be inbounding the ball. The Suns were a middle of the pack defensive team despite his reputation as a coach who totally ignores defense. He wasn't given the opportunity to do so with the Knicks. If you give away a 10rpg player, expect to be pounded on the boards.
Oh and that Marbury thing bears repeating. He has done absolutely nothing for the Celtics.
Z
Mike Brown didn't know about any of the advanced stats in San Antonio because the front office and coaching staff didn't think enough of him to share them. He was "encouraged" (to put it nicely) to look for a new job because of his stubbornness and aversion to change.
D'Antoni also has a track record of success(which Isiah does not possess).
I completely agree that you cannot break down the value of players into numbers, but discounting points per possession stats in favor of points per game seems silly. What advantage could points per game have in that matchup?
Shooting percentages can be looked at as well just because of the fact that the 3pt shot is worth 50%more.
There are tons of similar situations where the stats have definite benefit and ignoring them seems like a poor choice.
Not everyone needs the stats, and they definitely cannot provide a replacement for basketball knowledge. Brown and Popovich possess enough basketball knowledge to be successful without the crutch of statistics(most people are not however).
There are plenty of people who wouldn't mind some help evaluating talent and could probably learn something.
I think the stats angle here is getting blown out of proportion. The Knicks are not a good defensive team any way you slice it. They are 29th in points allowed or 23rd in defensive points per possession, both in the bottom third (just like the points differential). In fact, the tempo-free offense and defense numbers are shockingly close to Isaiah’s first season (D was 25th that year vs 23rd now, O was 16th then 14th now).
Now as for Mike Brown not being a numbers guy. He says he wants to keep opponents under 90 PPG. Well of course he does, his teams play at a slow pace. Since they only play a 89 possessions, wanting them to allow little more than one point per possession is demanding statistically sound defense. (it should be noted that the Cavs offense is great when you take pace out, but they are 19th in PPG)
If you were a coach using stats all that means defensively is pushing your team not give up a high shooting percentage, keep the opponent from getting offensive boards, not sending opponents to the free throw line and forcing turnovers without giving up high percentage looks. Shockingly any defensive coach (probably including Mike Brown) does that.
I’ll cite the Memphis Grizzlies as evidence of this. They let opponents shoot 47.4 percent and have a rebound differential of -1.5yet they are above average in defensive points per game. Again, this is because they depress pace. Simple as that.
Gbaked:
All I am saying is that the 2009 Knicks and 2007 Knicks resemble each other in their records. I am quite aware that the two teams play different styles and that Walsh is hoping to use cap room to add talent--but whether or not he is able to do that is hypothetical at this point.
Down the stretch with a playoff berth within reach, the Knicks have lost several ugly games at home. If that makes you optimistic about the future, so be it.
Mike Brown looks at points per game. Well, I am sure that that is just a piece of the puzzle he uses to evaluate a team, given his experience with basketball at a high level.
Mr. Friedman, you rightfully cite opposing field goal percentage as a measure of team defense. (Of course, it doesn't account for free throws and the difference between three-point and two-point shots.) FG% is a rate statistic, the number of shots that go in per shot attempts. It is a useful measure of efficiency. What is wrong with using another rate statistic to measure defense by looking at the number of points (which obviously matter more than FG%) per possession?
It doesn't make sense to assess a team's offensive prowess by how many possessions they have, but by citing simple PPG you are depending upon pace.
Z:
We agree completely about Marbury.
As for what the Knicks will do in the future, we will see. My article is about what the Knicks are doing right now and how that is being perceived/portrayed. I think the latter aspect is important but overlooked. D'Antoni and Walsh may very well save the Knicks but they have not done so yet.
Madnice:
Yes, you are right that the Knicks have lost some very ugly games recently.
G. Wolf:
You seem to have a "scoop" that has not been reported anywhere else, to the best of my knowledge. What is your source of information regarding Coach Brown's relationship with San Antonio's front office? Cavs GM Danny Ferry also came from San Antonio and both he and Brown are close to Coach Popovich, so what you are alleging does not sound very plausible. Hank Egan, who is one of Coach Brown's assistant coaches, was Popovich's college coach and worked for Popovich as an assistant in San Antonio. Coach Brown told me that when he was on the San Antonio staff the only "stat guy" on the bench was P.J. Carlesimo.
Rickjr82:
How do you define a "track record of success"? Isiah won two NBA titles and one NCAA title as a player. He has proven to be a very good evaluator of draft day talent. He led the Pacers to the playoffs for three straight years and their record improved despite having to rebuild after retooling the team that had made a Finals run. Granted, his time in N.Y. did not go well.
D'Antoni won a lot of regular season games in Phx but never reached the Finals despite having some very talented teams. I think that he is a good coach. All I am saying is that in his first year the Knicks have not accomplished anything more than they did in Isiah's first year.
It is Cleveland Coach Mike Brown who told me that he looks at opponents' ppg as a key stat (not in reference to this thread but in general).
I agree with you that stats can be useful but, like anything else, you have to understand what you are watching and what you are talking about in order to make full use of them.
Ben:
I agree with you completely that the Knicks' defense is bad by any statistical (or visual) measurement. It surprised me greatly that anyone would even try to dispute that point.
The point about Coach Brown is that he does not use the so-called "advanced" statistics. For more details about this, simply read the complete interview.
Birdman:
In the interview, Coach Brown states quite clearly that he coaches by "feel, not numbers." Again, all of this is spelled out, in his own words, in the interview.
Most defensive-minded coaches who I have talked to place value on point differential and defensive field goal percentage. As my research about recent NBA champions shows, those stats correlate well with winning (though we know that correlation does not necessarily equal causation).
Coach Brown told me that he looks at opponents' points scored more than he looks at differential. I understand your objection to this but I think that it is worth noting what Coach Brown thinks, particularly since his approach of coaching by feel instead of numbers is something that he says he learned from Popovich.
I don't have some objection to pace based stats per se. What I am saying is that in order to understand how/why a team wins or loses it is important to look at the total picture, not just one number. Pace based stats are not the be all, end all that some people seem to think that they are. They are one tool.
Ah, didn't read your last comment on pace-based stats until just now. I'd only read the article. So my apologies for saying you're rejecting them. I still maintain they are better than the metrics you cited for evaluating how well a team played defensively. I think you'll find that a team's pace-based stats correlate even more closely with winning than the ones you mentioned.
Idiots who think points per game measures are more important than points per possession measures must think that Westhead's Denver Nuggets teams of the early 1990s must be the greatest offensive team of all time, and Mike Fratello's mid-to-late Cavalier teams are the best defensive teams of all time.
After all, they have the best per game numbers of all time.
NTMR/KD (I'm assuming that you are not the real KD):
I don't think that Cavs Coach Mike Brown is an idiot. It is important to understand what he said. I asked him which stats he looks at to gauge/evaluate his team's performance and he said that points allowed (as opposed to differential or points per possession or anything else) is very important to him and that his target is to hold teams in the 80s. If you read the interview, you know that he said that in faster paced games the Cavs may give up more than 90 points while still playing good defense but that in general holding teams below 90 is the target.
I think that it is very interesting to know how a leading NBA coach thinks about the game and what methods he uses to prepare his team.
Why must raw points per game and points per possession be at odds with each other? Theyre supposed to go hand in hand in evaluating games.
Points per game are not "archaic." Without PPG, efficiency rating is removed from its context.
If Bynum shoots 5-5 while Kobe goes for 12-25, does that mean that you should be giving player of the game honors to Bynum???
So if Bynum can go 5-5 why didn't he attemt ALL of the Laker's baskets then?? Efficiency, when removed from its context, is meaningless.
So what if your team scores a few more points per possession than the other team, if the other team averages a dozen more possessions?
If you spend more time playing than looking at numbers, this is an easy concept to understand.
The objective is to outscore your opponent. You do so by a combination of having more possessions and using them efficiently, while preventing the other team from doing so.
Z
I can understand why coach Brown looks at opponents' points scored more than he looks at differential. Because he is evaluate his own team. Because his team's style has been established, total points allowed does reflect the players effort. He never said it is important to use that number to compare team to team.
Just a fan, not even a player and definitely not an analyst. But for me, all these stats improvements - from last year, from previous coach, etc. - are nothing if the team continues its awful standing through out the league.
Imagine, your team improves in almost all aspect/stats, but the team is still near at the bottom end of the standing.. how's that for everyone?
Z:
I don't think that those stats have to be "at odds with each other." In my original article about the Knicks, I cited an array of stats to make the case that the Knicks' defense this year is not good.
I agree with your point about "efficiency" being meaningless without context but I think that most "stat gurus" do indeed draw the erroneous conclusion that a player who shoots 5-5 is always more valuable than one who shoots 12-25. Certainly I think that kind of error goes a long way toward explaining the often bizarre ratings that WoW spits out.
Anonymous:
That is a very good point.
Just a Reader:
That, in a nutshell, was the point that I made about the Knicks in my original article: despite the hype/buzz, despite the career-high individual numbers for some players, despite the higher team scoring average, the Knicks are still a bad team--and, what should be more worrisome to Knicks fans, they are getting worse as the season progresses.
So what if your team scores a few more points per possession than the other team, if the other team averages a dozen more possessions?
In a game, this cannot happen. A possession ends, when the ball is turned over to the other team. Be it via a missed shot + defensive rebound, a score, a blocked shot or turnover. There can be a small difference through end of quarter possessions, but in the end, no team can have decidedly more possessions than the other team.
As for general concerns about what stats to use: I cannot understand how you could disapprove of per possessions stats, those take everything into consideration. Raw stats can e skewed by pace, allowed field goal % means nothing if you cannot secure the rebound´, turnovers forced can be a product of gambling and a faster pace, blocked shots mean nothing as long as the ball returns to the opponent directly. There can be no such clouding with points per possession. There is no scenario, in which you play very good defense and still give up a high points per possession margin.
Jalfa:
I think that he was referring to averages over the course of a season, not one game. Obviously, the number of possessions in one game is the same for both teams, as you indicated.
Per possession stats can certainly be interesting to consider but the actual unit that matters is a game; the object is to outscore your opponent over the course of a 48 minute game, so that is why a coach who is looking at the bigger picture for his team would be interested in scoring X amount of points or holding his opponent to Y amount of points. If the coach or someone else is trying to figure out how efficient his team is or why his team is scoring so much (or so little) then the per possession stats and some other stats might be ONE useful tool. Coaches also have to look at matchups and skill set considerations (for their team and for the opposition).
I don't question the worthiness of the goal of creating accurate basketball metrics, nor do I question the value of per possession stats per se. I do challenge the idea that basketball statistical analysis is right now a complete science that offers definitive conclusions about the value of players or the ranking of teams. The stats can be a useful tool and contribute to making such evaluations.
If stats indicate something that goes completely against visual evidence--or vice versa--then it is important to try to figure out why that is the case and not just assume that either the stats are wrong or the visual evidence is wrong. Too many "stat gurus" have a tendency to insist that their numbers--and only their numbers, of course, not the numbers of a competitor "stat guru"--provide definitive answers.
@dave
sorry, my comment wasn't directed at you, but to the general "don't use PPG/FG%, they are archaic!" crowd.
@jalfa
"As for general concerns about what stats to use: I cannot understand how you could disapprove of per possessions stats, those take everything into consideration. "
I do not disapprove of them, but I disagree that they take EVERYTHING into consideration.
Efficiency tells you part of the picture. Raw points/game gives you another part of the picture. FG% gives you another part. I'm not bashing efficiency. I like John Hollinger's articles. I think it's a very good indicator of a team's performance but I think using it out of context is flat out wrong. It is also not a substitute for watching the game with the intent of improving one's own understanding.
If Portland increases their pace to 100 pos/game, do you think they'll be scoring 110? NO! If Phoenix slows down their pace, their efficiency would NOT remain the same, it would drop dramatically.
Efficiency will only take everything into account in a video game. It's not the holy grail of statistics, we are not even close to getting there yet.
Ask any coach if they think Denver is a top 8 defensive team. Ask any coach if they think that San Antonio is only the 14th best offensive team in the league.
How do coaches use stats? They use it to evaluate their current performance against their past performance and the future performance that they want to achieve.
Blocked shots do help your team even if they go out of bounds. The shot clock doesn't reset, but the offense has to.
You're actually saying deflections don't help your team. Dwight Howard probably leads the league in out of bounds rejections, but the Magic are a great defensive team.
Going for steals does not automatically equal to gambling. Michael Jordan led the league in steals and steals per game three times, #2 all-time in career steals and #3 all-time in steals per game. How many defensive first team selections does he have? Didn't he win DPoY? There was also an article here a while back, I forgot if it was about Dr. J, discussing the merits of going for a steal.
Z
Z:
I have done several articles discussing the merits/drawbacks of going for steals, but perhaps the most thorough examination I have done of this subject is my Pro Basketball's Greatest Ball Hawks article, which can be found in the right hand sidebar under "Articles Originally Published at NBCSports.com."
As for shotblocking, it is obviously preferable to keep the ball in bounds and gain control of it but even blocked shots that go out of bounds have some value in terms of intimidation and making the offensive team inbound the ball and run a new offensive set.
Blocked shots do help your team even if they go out of bounds. The shot clock doesn't reset, but the offense has to.
You're actually saying deflections don't help your team. Dwight Howard probably leads the league in out of bounds rejections, but the Magic are a great defensive team.
Going for steals does not automatically equal to gambling. Michael Jordan led the league in steals and steals per game three times, #2 all-time in career steals and #3 all-time in steals per game. How many defensive first team selections does he have? Didn't he win DPoY? There was also an article here a while back, I forgot if it was about Dr. J, discussing the merits of going for a steal.
Well, I did question the effectiveness of a blocked shot, which is, on second thought, wrong. What I really should have said is something along the lines of: "the blocked shot is a good indicator of a defensive presense, however, it is, in my opinion, just as important to change a shot, which is a statistic that I have yet to see somewhere."
Moreover, I never said that steals are only accumulated via gambling, I said that they *can* be accumulated that way. No one ever accused Allen Iverson of being a good defender, for instance, or called the smallballing Warriors of the last few years a defensive juggernaut.
My point really is though that if you change a lot of shots, force turnovers without gambling and limit your opponents attempts at making a shot, you will give up fewer points per possession. You will, of course, also give up less points per game, but, as I said, those numbers can be skewed. Take, for instance, those Nuggets team that you mention. Their defense has always been underrated and there was a big deal made out of how it was their defense that limited their chances, although it was, and still is, primarily their offense that needed fixing.
If Portland increases their pace to 100 pos/game, do you think they'll be scoring 110? NO! If Phoenix slows down their pace, their efficiency would NOT remain the same, it would drop dramatically.
Yes, which is probably why such emphasis is layed upon controlling the pace of a game (which is essentially what coach Brown says he's doing, when he wants his team to not give up a certain amount of points).
Anyway, I think that there isn't much about our two positions that is really too different. I completely agree with you that there is absolutely no substitution for sitting down and just watching the games. I don't even want to argue about using advanced stats too much because it has really gotten out of hand in a number of cases. It is just that I find point per possession stats the most useful out of all the stats that are out there (except, of course, for winning percentage) and quite a lot more useful than any of the traditional stats that are being kept.
Rebounding and FG percentage are only a part of defensive efficiency, and whether the omission of other factors is deliberate or not, they certainly need addressing,
EFG% is a much better indicator than FG%, as it takes into account 3-point shots; certain teams defend 3-pointers differently. TS% is even better, as it takes into account allowing FTs.
No issues with your analysis of rebounding. But why, and where, are they getting fewer boards? Is it because they cannot force misses (defensive teams rebound what, about 70% of misses)?
Or is it because their opponents are not even getting a chance to shoot?
Turnovers are not mentioned once in the context of the article, and this is startling. Are the Knicks forcing turnovers, or not? The Spurs this year have been a very good defensive team, and a big part of that is forcing turnovers. They've been quite pedestrian in terms of FG defense, actually.
You cite Mike Brown continually as justification for relying on FG percentage and rebounding differential, and aversion to more sophisticated statistics. I expect that those are the main factors for his team's success (and most teams, for that matter) on the defensive end.
I was unaware that Brown had made any peer-reviewed publications touting FG percentage and rebounding as the sole contributing factors to teams success, however.
The Cavs are a pretty straightforward defensive team, and their defense is focused upon contesting shots and securing boards. Do they foul on every single interior shot attempt, where most players score at a very high rate? No, and neither do 28 of the NBA's other teams. (Take a bow, Utah!) That's pretty sound to me - most players shoot better from the FT line than anywhere else. Do the Cavs allow their teams opponents tons of 3-point shots? After all, even very good shooters typically are around 40% from beyond the arc, whereas almost everyone is a better shooter inside of it. No, they are pretty good at contesting threes. Do they take chances for steals, or leak out early before a rebound is secured? Rarely, but on occasion.
I believe he focuses on these factors because they are the ones that are most within his control. His players know not to allow open 3s, not to gamble, etc. Reiterating to contest shots and secure boards is the strongest message that he can give.
My point is this: FG percentage are rebounding are certainly important factors, but they are not the only factors that determine a team's defensive prowess. If you disagree, surely there is a better way to illustrate the point than "Mike Brown said so". I think I learned something about the appeal to authority in freshman philosophy class.
Philip:
Accusing me of doing nothing but "appealing to authority" seems to be the charge du jour against me this week. Are all of you stat guys comparing notes?
There is a big difference between simply appealing to authority and to quoting acknowledged experts in a field as just one part of a larger, overall body of work. If you check out the right hand sidebar of the main page, then you will see that researching NBA defense has been an ongoing project here for quite some time and that I have interviewed many Hall of Fame coaches and top defensive players about this subject, in addition to compiling data about what makes championship teams successful. My research combines numbers, anecdotes and intelligent observation of games and does not overly rely on any one of those areas.
To cite just one example, my NBCSports.com article about the statistical profile of a championship team (which can be found in the right hand sidebar of the main page) shows that eight of the 18 NBA champions from 1990-2007 ranked first in ppg differential and 15 of them ranked in the top five. Four of them ranked first in defensive fg%, 12 ranked in the top five and 15 ranked in the top 10. Also, of those 18 champions, 15 of them outrebounded their opponents by at least 1.6 rpg. There is a very solid numerical case to be made that being near the top of the league in these categories is an important part of being a championship team. Could other numbers be important? Sure. However, the fact that championship teams not only tend to rank highly in these areas but that leading NBA coaches like Popovich and Brown explicitly state that they focus on these numbers almost to the exclusion of anything else is a powerful argument that these numbers are a good barometer for how close a team is to being a championship contender. The quotes that I provided are meant as support for my case, not a mere appeal to authority.
Of course, it is funny that "stat gurus" deride an appeal to authority when their modus operandi largely consists of appealing to the authority of their chosen high priest of numbers (such as Berri) despite the fact that said high priests often know nothing about the sport other than the exotic numbers that they crunch and that the high priests completely dismiss the accumulated basketball knowledge of the people who are actually in the trenches. If Popovich or Brown showed up at an economics conference and said not only that they know more about economics than any professor but that they don't even want to hear what the professors have to say what kind of response would they get? That, essentially, is what Berri et. al do: they show up in discussions about the NBA, they blast the decision makers for allegedly being idiots who know nothing about the value of players and they demonstrate no interest whatsoever in what coaches have to say. Just reread your comment; I have the great opportunity to ask questions of coaches who have won championships and taken a team to the Finals and when I post their responses here the stat lovers basically mock Brown and Popovich as if they know nothing about basketball. That is just absurd and demonstrates hubris, a lack of intelligence and, frankly, a lack of common sense. Few writers are actually asking intelligent questions of these coaches and thus getting intelligent replies but instead of getting credit for the fine work that I am doing I get anonymous knuckleheads who don't know anything about journalism or basketball saying that I am appealing to authority or cherry picking numbers. Give me a break.
You may think that the numbers that you like or that a "stat guru" brainwashed you to like are the best indicators of how a team is performing but doesn't it interest you in the slightest that a four-time NBA championship coach and the coach of the team with the best record in the NBA this season disagree with you? Or do you just assume that because you read something about "advanced" basketball statistics that you are some kind of expert on the subject of NBA defense?
Turnovers are not mentioned in the PBN article about the Knicks because forcing turnovers is not an essential part of creating a great defensive team. In fact, a cornerstone of the Spurs' defensive philosophy is to refrain from pressing or even from excessively pursuing offensive rebounds in order to get back in transition and wall off the paint; the Spurs focus on holding their opponents to a low shooting percentage, not necessarily forcing turnovers. The Spurs rank dead last in the NBA in steals this season but they are ninth in defensive fg%--and that is actually a bad ranking by their standards, due largely to injuries to Duncan and Ginobili.
Check out my interview with Hank Egan (former Spurs assistant, current Cavs assistant) for more information about the Spurs' philosophy regarding steals/forcing turnovers. Some coaches believe in pressing, trapping and forcing turnovers but that is not true of all coaches--and you can get a lot of steals and still be a bad defensive team: last season, the Warriors ranked second in steals and 25th in defensive field goal percentage. In your "expert" opinion, which stat tells us more about their defensive prowess? It is true that some of the top defensive teams--like Boston--get a lot of steals but just looking at steals or forced turnovers only provides a small (and often misleading) picture of a team's defense.
Your line about Mike Brown not having written "peer reviewed" articles made me laugh out loud. You want credentials for Coach Brown? I'd say that a 66-15 record with one game to go while ranking first in points allowed, first in ppg differential and tied for first in fg% defense are good credentials. By the way, since clicking on your name does not lead anywhere, I'd love to know your credentials as an expert on NBA defense.
David,
Thanks for the timely response. My main contentions were your use of FG over TS percentage, and complete ignoring of defensive role in causing turnovers. When did I accuse you of nothing but "appealing to authority"? You spend a lot of time focusing upon this point, when it was a fairly minor issue.
I agree with your characterization of the Spurs. But even though they are not a team that goes for a lot of steals, they have this year created a lot of turnovers, albeit not an exceptionally high amount of live-ball turnovers. Have the Spurs declined defensively, and is a lot of this due to defense against FGs? Yes. But forcing turnovers has mitigated some of the decline.
I agree that forcing turnovers is not the most important factor of being a good defensive team, but it is a factor. If you had argued that the Knicks defensive struggles will continue until their FG defense and rebounding get better, I would agree with you.
If you wrote an article saying that the Knicks defensive improvement is misleading because none took place in two of the most important aspects of defense, then again, I would agree with you. If you said that forcing turnovers can grant minimal gains at best, once again we could find an accord.
There are only three factors that affect defensive efficiency: turnovers, TS percentage, and rebounding rate. There are a ton of outside variables that impact these three factors, but they are it. Why you completely ignored one of these factors and focused only on part of another is beyond me. It's like trying to
(This is assuming that teams have equal possessions, which can be slightly altered by end-of-quarter possessions. I do not know any any significant difference between teams in this regard, but correct me if I am mistaken.)
Does it interest me what Gregg Popovich thinks? Of course. But if all e really cared about was FG percentage, then why would so much of the Spurs defense be focused upon taking away 3-pointers, despite it being a low percentage shot in terms of FG percentage? Because 3-pointers can be very effective from a TS and eFG perspective. And why do they not foul on every interior shot, despite it being a high-percentage shot for most players? Keeping their key players on the floor is part of it, but most players are far better scorers from the line than anywhere else. (Suns fans know of one exception, I am sure). And why do they even get steals at all, if they are so unconcerned with them?
Even though the Spurs might use different tools of evaluating their own play, that they are nonetheless focused upon other aspects as well.
I do not disagree with your assessment of the main attributes of a good (or bad) defense; only that they are the only ones worth mentioning.
I'm glad I could make you laugh out loud. And I should hope that my ideas are what matter, not my nonexistent credentials.
Philip:
Perhaps you intended the "appeal to authority" remark to be a throwaway line but I have recently come across that accusation in a few other threads and I find it to be completely absurd; quoting top NBA coaches is one way of citing reliable evidence regarding how NBA games are won and lost and calling that an "appeal to authority" simply begs the question of who the accuser considers to be a greater "authority" on these matters than NBA coaches. Certainly, David Berri assumes that he knows more about the NBA than anyone and he blithely dismisses the insights of NBA coaches/GMs but that is just one example of why even other "stat gurus" consider his work to be seriously flawed.
I would be interested to know your source for the Spurs' "forced turnover" data. The Spurs rank dead last in the NBA in steals. Forcing turnovers has never been a focus of their defensive scheme under Popovich. Would they like to cause turnovers? Of course, but not at the expense of compromising their primary defensive principles of walling off the paint and contesting shots.
As I said, some coaches believe in using pressure defense and forcing turnovers but simply looking at turnover stats in isolation tells you very little about how effective a team is defensively. The Knicks are so horrible defensively and on the glass that even if they led the league in steals they still would have big problems; they ranked 11th with 7.44 spg, while top ranked Utah averaged 8.78 spg, so we are talking about a difference of only about one possession per game. I think that non stop pressure defense works a lot better in the collegiate game, for a variety of reasons; good NBA teams tend to only selectively use elements of pressure defense.
While I may not have explicitly stated the things that you claim I should have said in the article, those things are certainly implied by what I wrote. That said, since a lot of people apparently either did not understand my article or chose to deliberately misinterpret it, I may have assumed that my readership has a greater understanding of NBA defense than they actually do. Certainly, the folks at Knickerblogger and WoW turned out to be completely clueless.
As for your analysis of the Spurs, it would be more precise to note that they especially focus on denying corner three point shots. The corner three is only 22 feet from the hoop, much closer than the 23'9" distance of the rest of the three point arc. On the flip side, that is why if you look at Bruce Bowen's shot chart he takes a very high percentage of his three pointers from either corner. In general, the Spurs want to run other teams off of the three point line, keep them out of the paint and force them to shoot contested two point jump shots.
Popovich told me that in addition to looking at defensive field goal percentage he also looks at any numbers in a given game that are unusual and I think that this answers your question: if the other team is making a lot of three pointers then Popovich would be concerned even if the overall defensive field goal percentage remained at an acceptable level.
I assume that your question about fouling on every interior shot is rhetorical but on the off chance that it is not, a team that does this would not only foul out several of its own players but would also put the other team in the penalty, resulting in two free throws after every foul. The best defensive coaches almost universally stress defending without fouling (Pat Riley's Knicks had a "strategy" of committing five fouls on every play with the idea that the refs could not call every one and would eventually relent and call a looser game but that is a story from another era).
Your question about why the Spurs get any steals at all is similarly naive/rhetorical in nature but the answer is that sometimes the other team simply throws the ball away and/or the Spurs can get a deflection in one on one defense without leaving the paint vulnerable. What the Spurs avoid doing is taking risky gambles for steals that could break down their overall defense, though Ginobili is the one guy who Popovich gives a little bit of license to do some defensive freelancing. If you look at Bowen's numbers, he was a top defender for years while getting very few steals (Dumars, in a previous era, also did that).
I won't judge you by your credentials (or lack thereof); my point is that anyone who considers himself a serious NBA commentator and/or fan should take heed to what NBA coaches say instead of dismissing their comments out of hand. Coaches are not always right--and they can of course disagree with each other--but in any field you have to understand the fundamentals before you try to blaze new trails. I just get fed up with these "stat gurus" who think that they have "solved" the NBA game on their spreadsheets. I respond so vigorously to their nonsense that I am sure some people consider to me to be anti-stats, which is ironic considering that I actually love numbers; I am simply opposed to the incorrect usage of numbers and I would think that any number-loving person would agree with me about that.
Post a Comment
<< Home