David Thorpe – of ESPN.com – recently wrote a column comparing Kevin Durant to Carmelo Anthony (insider access required). Thorpe’s analysis considered a host of factors including shooting, scoring, making teammates better, on-the-ball defense, secondary defender, rebounding, and intangibles. For each category the players were graded on a 10 point scale, and the player with the most points was…
Wait, before I get to Thorpe’s answer, let me comment on the word “intangible.” This word means “not tangible” or something that we cannot discern or measure. And yet, Thorpe is able to tell us that Durant offers more “intangibles” than Melo (by a score of 7 to 5). So we can’t measure “intangibles” but we know Durant offers more?
Thorpe argues that Durant’s value – according to Thorpe’s scoring system – is 48 while Melo scores a 44. So ½ of the difference between Durant and Anthony can be linked to something that – by definition – cannot be measured. Thorpe is not the only person to abuse the word “intangible”. But it’s odd to see someone assign a number to something that by definition, isn’t tangible.
Okay, let’s take a more tangible approach. We begin with Durant. Table One reports what Durant – and his teammates with the Oklahoma City Thunder – have produced after 60 games in the 2009-10 season.
Table One: The Oklahoma City Thunder after 60 games in 2009-10
As one can see, Durant leads the Thunder in Wins Produced. Of the team’s 36 wins, 13.3 can be linked back to Durant.
Moving away from the subject of Durant for a moment… one can see that the Thunder are led by a collection of young players. Durant, Russell Westbrook, James Harden, and Serge Ibaka are all younger than 22 years of age. And this quartet are on pace to produce 36 wins this season. As has been noted in the past, young players (younger than 24 or 25) tend to get better (so although Thabo Sefolosha is already quite good, he is not likely to get much better). This means that prospects for Durant and the Thunder are extremely bright.
Now let’s turn to Carmelo Anthony and the Denver Nuggets. This team already is quite good. But as Table Two indicates, Denver’s success is not really about Melo.
Table Two: The Denver Nuggets after 61 games in 2009-10
Of Denver’s 40 wins, only 4.7 can be tied to the production of Anthony. And four players – Chauncey Billups, Nene, Chris Andersen, and Kenyon Martin – have done more for the Nuggets this season.
Now we “know” what Thorpe said (remember he favored Durant over Melo) and what we learn from this analysis must be wrong. On Wednesday night, Anthony and the Nuggets crushed Durant and the Thunder. And when we turn to the box score, we see that Melo posted a 16.5 Win Score. Meanwhile, Durant only posted a 1.5 mark. So there you have it. Melo is clearly better than Durant.
Okay, obviously one game is not much of a sample. Let’s look at Table Three, where what Durant and Anthony have done with respect to all the box score statistics across the entire season is noted.
Table Three: Comparing Kevin Durant and Carmelo Anthony in 2009-10
When we look at the entire season we see – as we saw when we looked at Wins Produced – that Durant has done more. Durant is currently offering more with respect to shooting efficiency, rebounds, blocked shots, and personal fouls. consequently, we shouldn’t be surprised that Durant is producing so many more wins than Melo.
When we focus strictly on Anthony we see that he definitely scores points in large quantities. But his shooting efficiency is only average. So yes, he is above average (because he rebounds and gets to the free throw line). But Anthony is not quite as valuable as his scoring average suggests.
Let me close by once again noting how far Durant has come. Despite being named Rookie of the Year, Durant had a disastrous rookie season. Last year, though, his production was above average. And now – at the age of 21 – he is a star. If he continues to improve – and the same happens with his young teammates – Oklahoma City is going to be a dominant team in the NBA for many years to come. So although the Thunder were crushed on Wednesday, the Thunder will be more like to be the crushers – as opposed to the crushees (crushees???) in the future. And that outcome should be quite tangible.
– DJ
The WoW Journal Comments Policy
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
khandor
March 4, 2010
David,
———————————————————
re: As has been noted in the past, young players (younger than 24 or 25) tend to get better (so although Thabo Sefolosha is already quite good, he is not likely to get much better).
———————————————————
Based on the fact that your metricsare concerned with a player’s “level of statitistical production” it might be more accurate for you to say, instead, that,
“younger players tend to improve their ‘individual production numbers’ as they get older,”
as opposed to using qualitative words like “better” and “good” to describe their relative abilities.
As I’ve noted for you in the past, just because an individual player’s level of easily discernible statistical production might actually decrease as he grows older in the NBA does not necessarily mean that he is not also becoming a “better” basketball player, as his individual game continues to mature.
Nick
March 4, 2010
Simply because an intangible cannot be measured does not mean that it cannot be evaluated.
For instance, beauty is intangible and certainly we can evaluate variations in quality even if we disagree about those evaluations. But the fact that we can all acknowledge that there is variation of some kind means evaluation is possible. This is the same for artistic merit generally.
Or if you want to define intangible as something unable to be perceived by any of the senses entirely (as opposed to just touch) then you can use something like love. Love varies in quality and experience. You can’t put a ruler to it. But you can evaluate these differences.
So if you’d like to argue that rating intangibles will inevitably be subjective, that’s fine. It’s true. But you seem to be making the false point that one can’t evaluate them because one can’t quantify them or at least not literally. Yet this is exactly what everyone does when they say someone has a nice looking jumpshot or that was an awesome dunk.
Mo
March 5, 2010
intangible != immeasurable
in·tan·gi·ble–adjective
1. not tangible; incapable of being perceived by the sense of touch, as incorporeal or immaterial things; impalpable.
2. not definite or clear to the mind: intangible arguments.
3. (of an asset) existing only in connection with something else, as the goodwill of a business.
—Synonyms
2. vague, elusive, fleeting.
im·meas·ur·a·ble–adjective
incapable of being measured; limitless: the immeasurable vastness of the universe.
bduran
March 5, 2010
Nick,
His point stands. Thorpe did measure Durant’s and Melo’s intangibles. Durant got a 7 of 10 and Melo got a 5 of 10.
Khandor,
These are stats. He’s saying that in general the box score stats correlated with winning games improve until the age of 25, then decline. So while it might be true that indivdual palyers peak later, his statement stands. Hence the words “tend” and “likely”.
Your counter argument seems to be that while the measurable stats peak at 25, the intangibles continue to improve.
brgulker
March 5, 2010
But to say that player X is better than player Y because “he has a nice looking jumpshot” would not be a supportable claim.
This seems to be the type of thing that Dr. Berri is arguing against and the thing Thorpe is arguing for.
I liked your analogies, and I think there’s a good point there. I’m just not sure it’s directly related to the “intangibles” of basketball, especially when those intangibles are weighted and valued in direct relation to the tangibles.
Italian Stallion
March 5, 2010
I think the word “intangible” as commonly used to describe athletes refers to human qualities that are difficult to measure and tangible qualities for which statistics have generally not been available.
Things like leadership ability, ability to handle pressure, competitiveness, self motivation, basketball IQ, ability to get along with others, experience under fire, ability to gain the respect of others, etc… are qualities that to some degree show up in personal statistics, but to some degree can also benefit others on the team.
Things like taking charges, altering opponents shots, help defense, keeping a rebound live without actually getting it etc… are things that are measurable, but that have typically not been part of the boxscore. So they have been lumped into the intangible category because everyone recognizes their value.
Italian Stallion
March 5, 2010
I should add that I also think is kind of funky to try to assign a number and value to less tangible human qualities. But I think we can recognize somewhat general levels of them in individuals. So there’s no reason to ignore them. It’s just tough to measure and value them as accurately as rebounds, assists etc…
Nick
March 5, 2010
bduran,
Assigning a number to something doesn’t mean that one measured something. It’s simply an evaluation.
When I give a movie two thumbs up or someone else gives a restaurant five stars, they didn’t measure the food nor did I measure the movie. The evaluated it i.e. compared it to others.
The numbers there are essentially metaphorical comparisons, which makes sense in regard to intangibles.
Nick
March 5, 2010
brgulker,
I think it’s perfectly fair to say that evaluating intangibles isn’t likely the best way to identify the most productive player on the court because intangible things like hustle or leadership do not have any great effect on productivity outside of what is already translated into the box score. I’d agree with that for the most part.
But to take the extra step and state that one cannot evaluate intangible characteristics like hustle or leadership on their own, regardless of their final effect upon productivity, appears to be a plain falsehood. I cannot quantify who hustles more but I can perceive differences between two player’s activity levels and so can you.
And once there’s a recognition of differences, there’s the possibility of evaluation.
khandor
March 5, 2010
bduran,
1. It is always interesting when one person chooses to criticize another’s use of a particular word in the english language … e.g. using the word “intangible” in conjunction with a specific numeric measurement tool … and then that same person him/herself falls prey to using other words in a context that might be considered as inappropriate, as well … e.g. associating specific numeric measurements with qualitative terms like “better” and “good” [etc.] instead of numerically relative words like “more” and “less” [etc.].
2. Not only do the “intangibles” associated with a “good-great” player improve/increase as s/he matures … but, so, too, does an assortment of other specific statistical measures, which elite level coaches keep in their personal intellectual databases over time, that are difficult for the average “fan” and/or “economist” to judge with any degree of accuracy, e.g. #1. Greater Number of Low Percentage Shots Passed Up which Lead Indirectly to Enhanced Scoring Opportunities for Less Skilled Teammates; #2. Fewer Number of Blown Individual/Team Defensive/Rebounding Assignments which Lead Indirectly to Enhanced Scoring Opportunities for Opponents.
3. Indeed, what some [many?] basketball “fans” and “macro statistical analysts” might think of as “intangibles” aren’t that, at all, in the eyes of an elite level basketball coach who considers the performance of those highly specific skills to be concrete, fundamental aspects of the game.
bduran
March 5, 2010
Nick,
I would accept this argument if he didn’t then use this number in his final total. He scored intangibles and assigned them a relative value to tangibles.
Khandor,
The use of terms like better and good are clear in this situation. Improving win score means a player is getting better … in general.
David claims his metric is highly correlated with winning, not that it’s an absolute. It’s entirely possible that a coach sees something valuable to his team not reflected in the box score, all though your shooting example isn’t the best. Better shot selection leads to more efficient scoring which leads to a higher win score.
If , in general, box score statistics decreased after the age of 25, but more important intangible quantities kept improving, then there wouldn’t be a strong correlation between win score (based on box scores) and winning. It seems like you are trying to focus on possible excpetions.
Nick
March 5, 2010
bduran,
That’s not exactly true. It’s not as though Thorpe summed up intangibles alongside tangible things like rebounding. In other words, Thorpe didn’t state that Durant had five hustles and forty-seven three pointers while Carmelo had eleven leaderships and twenty seven rebounds.
Instead, what he did (rightly or wrongly) was evaluated tangibles as though those were intangibles. Normally, NBA players shooting isn’t rated on a 1 to 10 scale as some sort of holistic measurement. Instead, you just look at the field goal percentage (or eFG% or what not).
So the problem here isn’t that Thorpe tried to measure intangibles as though they were tangible quantities; rather, the error is that he evaluated tangible quantities on a bizarre intangible scale when there was clearly quantitative data that covered those issues.
The general point I’m making isn’t that Thorpe got it right or that intangibles should affect the measurement of a player’s productivity.
I’m only say that simply because intangibles “by definition cannot be measured” (Berri’s phrase) does not mean that they cannot be evaluated and compared.
bduran
March 5, 2010
Nick,
“Thorpe argues that Durant’s value – according to Thorpe’s scoring system – is 48 while Melo scores a 44. So ½ of the difference between Durant and Anthony can be linked to something that – by definition – cannot be measured.”
This is the key point. He assigned points, Durant won by four. 2 of these point were due to “intangibles”. This is silly. Now laugh with me.
Nick
March 5, 2010
It is maybe silly but I don’t think for the reasons you’re claiming, and once again I’m not defending Thorpe.
But he didn’t say that Durant had seven intangibles while Carmelo had five. He graded the quality of their intangibles at that level. That is something entirely different and it is not, by definition, a measurement.
Measurement is quantity. No where did Thorpe argue that a certain quantity of intangibles occurred. Instead, he used a point system, as a teacher might when grading a research paper, to evaluate the quality of the intangibles.
Measurement, unless you’re being metaphorical, does not seek to directly determine quality.
evan
March 5, 2010
if i were moderating this forum, i would bring the banhammer down on khandor
brgulker
March 5, 2010
Nick,
Wouldn’t it make sense for the “point system” to correlate with actual “quantities”?
That seems like the only logical way to evaluate the intangibles at all.
Ethan Harning
March 5, 2010
Does it bother you that like Wayne Winston your opinion of Kevin Durant has changed so dramatically?
Last season in https://dberri.wordpress.com/2008/11/16/durant-and-company-miss-for-the-thunder/ you were saying Durant was:
“Speaking of Durant, much has been made of his rookie performance. According to the coaches and the media, Durant was the top rookie in 2007-08. According to the Wages of Wins metrics, though, he was well below average. Although Durant did not play well last season, there was hope he would get better. After 10 games, though, that hope has not been realized.
…Durant was “well below average” in your opinion in his rookie season.
After 10 games this season Durant has the same profile. But now he has declined further with respect to every statistic except for shot attempts and personal fouls. Yes, Durant went from bad to really bad
…a little ways into his second season you had decided he was “really bad”
Then you can conclude with “Despite a plethora of bad players, there are some bright spots on this team. The first is Jeff Green.”
…I almost lost my lunch! Green is the star and Durant is really bad??????
Since you maintain players in basketball are so consistent, doesn’t it trouble you now to think if you had been the GM of Oklahoma City you would have gladly traded Durant for say Reggie Evans? (like Winston would have traded Durant for Sebastian Telfair!)
todd2
March 5, 2010
If we adjust for the Halo Effect (being associated with a team with a better record), Durant easily surpasses Anthony. ;-D
brgulker
March 5, 2010
Ethan, have you ever heard the saying about the Forest and the Trees?
Robert O'Malley
March 5, 2010
Ethan, Berri also said Durants college numbers projected him to be a very very good pro player. He has also said that rookies tend to produce on average a lower win score than typical players.
So while I haven’t talked to Berri about this in any way I doubt he would have traded Durant for Evans after his rookie year. Especially when he was pretty good by the end of his second year and now really really good. I know he expected Durant to be very good coming out of college and expects rookies to under-perform relative to the whole.
dberri
March 5, 2010
Duran’t improvement can be explained in terms of how performance of NBA players change over time. So I am okay with saying
1. Durant was very bad as a rookie (and he was)
2. Durant got better as he aged
For adjusted plus-minus, we see player performance changing (according to this metric) for no apparent reason. So Durant’s performance being evaluated differently, especially when that evaluation ran counter to what we see in the box score numbers, should be troubling to people.
John Giagnorio
March 5, 2010
Only one post calling out dberri for “hating” Durant in the past? Something must be wrong here!
Man of Steele
March 5, 2010
Ummm, not to take away from the discussions on the proper definition of “intangible” or the varying evaluations of Kevin Durant, but I thought I’d note that this article reminded me of an article in ESPN about a year ago along the same lines. Although I don’t think that article mentioned Serge Ibaka, it basically argued that as Durant and Russell Westbrook improved, along with solid contributions from Thabo Sefolosha, the Thunder could be a good team within a year or two.
Italian Stallion
March 5, 2010
If we can agree that things like leadership ability, ability to handle pressure, ability to gain the respect of others, willingness to be the hero/goat, basketball IQ, experience in huge games, competitiveness under fire, desire to get better etc…. are qualities that not only impact a player’s own statistics, they can influence the success of other players and the team (however marginal) I think we are half way home.
Obviously if we can’t agree on that much, then the conversation is over. :)
The next step is to break away from a purely mathematical way of thinking about things just for a moment.
Even though measuring and valuing things like this may be somewhere between difficult and impossible, if you agree that they do have value, then ignoring them would seem quite foolish.
So what some people do in situations like this is try to intuitively quantify and value these things. Then if they are smart they measure their success over time and adjust their thinking as they learn. (trial and error)
That intuitively ability and way of learning is similar to mathematical ability in some ways. We are not all equally blessed with that skill either. But it is a method of learning that can also be successful.
I built a model that successfully beats horse racing primarily using intuition, very basic stats, and trial and error. I know others that have also. I am still waiting to meet someone with advanced mathematical skills that can beat the game with numbers alone. (I heard there is one)
So when you read something like the Thorp article, the idea is to use your own experience and intuitive ability to decide whether he’s in the ballpark and not worry about the decimal points.
Palamida
March 5, 2010
Ethan, You have to remember that Durant is a very special case. Most players tend to “carry over” their production in terms of WP from the college game to the Pro game. Durant had a stellar College season in terms of WP and was indeed as someone here mentioned “projected” to be a significant win producer. However following the combine workouts in which Durant failed to bench press a single time he was deemed by many to be too “weak” to play inside at the next level. That in itself is rather ludicrous since it uses faulty logic: Durant was “weak” that much seems to be a fact. His “weakness” however didn’t stop him to be a very productive player at the college level – much more productive than other “stronger” players.
I’ts not like there’s a critical point\line of strength that’s needs to be met in order for one to play in the NBA. sure, Most inside players use strength (that Durant doesn’t possess) to produce. Obviously he uses something else :p otherwise he wouldn’t have been this productive! What Durant actually uses is his great length and talent. Durant has a standing reach of 9’2 which is above the average Center in the NBA! he’s long!
following the ballyhoo about the combine some “genius” (P.J Carlesimo?) decided he has a solution to the problem (which doesn’t exist, obviously) – we’ll deploy Durant at SG. That is something that does not happen every day, and thus makes Durant’s case, a special one.
Seemingly Durant has the “entry level” skills to play Sg that would usually limit players Durant size: He can shoot from outside and has the ballhandling skills of a guard. This was a grave error: Durant played his whole life as a “big man”, that’s his mentality and that’s what he knows. He’s a poor passer in comparison to other Guards in the league and all his natural abilities and unique talents could not come to fruition in the SG position. His ability to rebound\Block was greatly diminished away from the basket and he didn’t know\wasn’t able to use his superior size to his advantage as a Guard. Also it exposed his glaring weakness as a passer (SG’s carry a larger load of passing than Sf\PF). And so Durant was terrible. When Brooks took over he immediately switched durant to SF\PF and the improvement was immediate. I’ts not like he “improved” he was simply deployed correctly and was able to showcase his true ability.
In his second season these are durant’s stats as a SG Per 48: 24 fga with Efg% of 48.6; 5 FTA ; 4.7 REB; 2.6 AST ; 5.4 To; 0.4 Blk; 2 PF; 27.9 PTS (82games.com don’t provide steals stats)
AS a SF in that season he posted : 23 FGA at 51% Efg%, 8.9 FTA! ; 8.2 REB!; 3.6 AST; 3.5 TO; 0.9 BLK; 2.2 PF.
That’s incredible. He upped his shooting % significantly, almost doubled his FTA (very easy for him to use his length to draw fouls from bigger players compared to Guards) almost doubled his Rebounds and Blocks while Turning the ball over almost half as much and even adding an extra Assist per 48 mins. That’s incredible. Imagine Shaq being assigned to take the ball over halfcourt, run the pick and roll (as the PG), taking outside shots, etc. He will be abysmal. Offcourse i’m just making the Shaq example to prove a point and There was no reason to believe Durant would be so poor as SG (most players who switch from SG to SF or vice versa do not exhibit such a radical change in production), in Durant’s special case – that’s exactly what happened. durant now splits his time between Sf and Pf depending on matchups and lineups and is sometimes guarded by big wings and sometimes by inside players. In all of those cases he comes out on top.
In summation I don’t think Mr. Berri needs to be concerned about the fact that Durant made such a remarkable improvement; players do tend to offer consistent production, but i’m guessing Duran’t unique skill set and physical properties makes him an exception.
P.S this season, durant played about 30 mins of SG (a laughably small sample) but i’d like to note he was utterly terrible in those very limited mins, meaning even the “superstar” 3rd year Durant finds it impossible to produce as a guard.
Thoughts?
Ray
March 6, 2010
Nick is absolutely killing it in his argument. Everybody keeps sidestepping it or missing the point. I’m gonna try this caveman style to help Nick out because Nick is right.
Intangibles (leadership, off-ball, etc.) cannot be measured; there are no stats for leadership. They can be evaluated, however. Like choosing if you love your car more or your guitar. You can’t get stats for your car or guitar, but you can evaluate with reasoning which one you love more.
Tangibles (FG%, Rebs) can be measured because they keep stats for that.
Nick is saying that evaluating jump shooting (for example) is stupid, because they keep track of jump shooting %. So instead of Thorpe givng Durant an 8 and Melo a 6 for jump shooting, why not just say that KD shoots _% on jumpers and Melo shoots )%.
I believe Nick is saying that evaluating intangibles was what Thorpe did right, and evaluting something that can be kept track of is pointless. Like letting a machine grade a multiple choice test and then having a teacher go in and grade how THEY thought you did on the test. It’s just meaningless.
So Thorpe’s scoring system is essentially correct and consistent. He grades everything as though it is abstract and intangible. The problem is, though, many of that grades could be measured because of stats and it basically made everything pointless, though not incorrectly scored.
Italian Stallion
March 6, 2010
>I believe Nick is saying that evaluating intangibles was what Thorpe did right, and evaluting something that can be kept track of is pointless.<
Exactly
stephanieg
March 6, 2010
All I know for sure is that whenever Durant has a big game the collective Portland fanbase weeps openly. I wonder if Roy/Durant would have led to conflict.
Too bad Oden’s bones are made of rice krispies.
mrparker
March 6, 2010
I suspect Carmelo would have the same numbers as Durant in the same role. I don’t think he would do as well on a different team and we shouldn’t get carried away. Players don’t just start putting up monster numbers after two ho-hum seasons. Let’s not underestimate getting to play with Westbrook and Harden. Most will probably disagree with me.
Michael
March 6, 2010
Durant is a joy to watch. I don’t know about the rest of you guys but I am getting more excited about his future each season.
todd2
March 6, 2010
When Bob Knight was coaching he was known to spend less than 10 minutes in a gym evaluating a kid. It’s probably a common practice among elite coaches. If you’ve been around the game long enough, you can also get a pretty good read on a player during warmups.
khandor
March 6, 2010
Palamida,
—————————————
re: I’ts not like there’s a critical point\line of strength that’s needs to be met in order for one to play in the NBA.
—————————————
Unfortunately, I will need to disagree with this point you attempted to make.
There is most defintely a “tipping point” at which a specific player either has the level of “physical strength” required TO EXCEL at the NBA level, given the specific requirements of a given position, as a multi-dimensional player, or he does not.
In Durant’s case …
He did not have the “strength” it takes to succeed as a PF when he first entered the league, which is why Coach Carlesimo made the correct decision to shift him to a new position, at the beginning of his pro career.
Unfortunately for PJ, however, he made an equally poor decision, and fruitlessly attempted to employ KD as an elongated SG … which was doomed to fail from the get-go, not because of Kevin’s relative lack of strength for that position, as well, but due to his relative lack of quickness and his inability to defend adequately against some of the smaller SG’s and most of the PG’s in the league [i.e. in cross match or various switch scenarios].
When, Scott Brooks took over and made the decision to use Durantula, almost exclusively, at the SF position, he illustrated just how important PROPER COACHING actually is in the game of basketball, by putting his most talented offensive player at the position for which he is bested suited in the NBA game.
===============================
Ray,
I agree with your take that Nick [and IS are] is eating up ground in this instance.
===============================
IS,
re: The relative worth of “evaluating” something for which a statistic is kept; versus “evaluating” something which is properly considered an “intangible”
Classic stuff, right there … my friend.
khandor
March 6, 2010
todd2,
—————————-
re: “When Bob Knight was coaching he was known to spend less than 10 minutes in a gym evaluating a kid. It’s probably a common practice among elite coaches. If you’ve been around the game long enough, you can also get a pretty good read on a player during warmups.”
—————————-
In last night’s broadcast of the Spurs vs Hornets game, Jeff Van Gundy made reference to the fact that … when HE says an athletic high school prospect “can actually play”, either in basketball or football, no one should ever doubt that he actually knows what he is talking about [in this instance, when it comes to be able to evaluate accurately the chances of Emeka Okafor’s younger brother to succeed as either a college basketball or football player].
In my experience, as an expert evaluator of athletic ability, if you are someone who truly knows what you’re doing it shouldn’t take you more than 10 minutes to assertain correctly whether or not a prospect has what it takes “physically” to succeed at a specific level of competition.
It’s when it comes to the many “intangibles” that also factor into the equation, however, that the process takes a longer amount of time, sometimes, even for the best in the business.
Travis
March 6, 2010
“I suspect Carmelo would have the same numbers as Durant in the same role. I don’t think he would do as well on a different team and we shouldn’t get carried away. Players don’t just start putting up monster numbers after two ho-hum seasons. Let’s not underestimate getting to play with Westbrook and Harden. Most will probably disagree with me.”
Dude, what? I hope you’re trolling. You have to be trolling. Just in case you’re not, here’s the following:
You know who else had an unbelievably dreadful rookie season? LeBron James. He was even worse than Durant was offensively in his rookie season: 99 ORtg to Durant’s 100. Here are LeBron’s (left) and Durant’s (right) first three seasons, using ORtg, eFG, TS%, rebound rate, and usage rate:
99 : 100
114 : 111
115 : 115
.438 : .451
.504 : .510
.515 : .514
.488 : .519
.554 : .577
.568 : .601
7.6 : 6.9
10.2 : 9.6
9.8 : 10.8
28.2 : 28.1
29.7 : 28.3
33.6 : 32.1
LeBron was miles ahead of Durant in assist rate. That’s it. That’s all that LeBron has on Durant statistically by any significant margin in their first three seasons. This comparison is especially valid as they are separated in age by only a few months; they are almost the exact same age during their matching NBA seasons.
The real fun part will come next season; LeBron dropped off fairly considerably, both in efficiency and usage rate. With the emergence of Westbrook and Harden, Durant’s usage may stagnate and very slightly drop, but I don’t see his efficiency going anywhere but up.
Travis
March 6, 2010
*LeBron dropped off fairly considerably in his fourth season, both in efficiency and usage rate. With the emergence of Westbrook and Harden, Durant’s usage may stagnate and very slightly drop next year, but I don’t see his efficiency going anywhere but up.
mrparker
March 6, 2010
I wasn’t trolling. When you watch Carmelo play on athletic teams (i.e all star, team usa) he always looks like he might be the best player on the floor.
NoTimeToRead
March 6, 2010
There are a lot of interesting comments, and I believe this thread highlights the main objections to the econometrics analysis of sports performance. I will try here to help people reconcile the times when models and eyeball information say different things.
I agree with Ray’s point, that treating performance measures, for which we have statistics, as if they were intangibles is ridiculous. Just use the statistics. So Thorpe’s main contribution is to use intangibles as a means to differentiate between Durant and Melo. Thorpe had ranked the 2 as essentially statistically similar (i.e. having produced similar performance scores, whatever they may be.)
Berri seemed to be annoyed that the Intangibles score accounted for 50% of the difference. It doesn’t annoy me at all. It seems to me that Thorpe simply thought Durant and Melo are similar, and cast a tiebreaker (with his intangibles score), if you will.
As many posters have noted, econometrics aren’t the only way to analyze the game. Berri has never said that one should ignore aesthetic factors, nor has he ever said that one cannot bring more information to the analysis (such as the breakdowns for players favoring one side, better success on a drive through the lane from the left or from the right, physical size matchups, passing success when using the dominant vs weak hand, etc.) As a matter of fact, that’s the exact job that coaches should be doing.
However, I do think that Berri also thinks that a fair amount of these human-tendency statistics and so-called intangibles do in fact affect box score statistics.
If you have low basketball IQ, one might reasonably expect bad passes, bad coverage and compensation by committing hard fouls, coughing up the balls, taking bad shots, and so forth. But in all these cases, there are in fact correlates to statistics: increased turnovers, increased fouling, low shooting percentage, below average number of assists, low blocks and steals.
I agree with Italian Stallion that observation and generating some sort of intuitive (ahem) model for player performance is a skill and talent. However, successful scientists can do this as well as encapsulate these intuitions into a quantitative model. You can’t really have one without the other. To generate new research, most scientists must observe data from their own experiments and read between the data published by other scientists to detect discrepancies and incorrect conclusions. So I would say that in general, scientists must be observant and have some good intuition about their system, about their experiments, and about their data and analysis.
At any rate, scientists tend to use a so called reductionist approach and numerical models, not because numbers are better in and of themselves, but rather because the mathematics force scientists to use precision in defining a model. This is what is meant when a scientist says that a mathematical model of a natural phenomenon is “rigorous.”
I would say that Berri’s approach (and by extension the scientists’ approach) takes the whole intuition model one step further. Berri asked, which of these observations really matter? His answer, as we should all know by now, is that one can come up with a model (a mathematical estimation, essentially), that uses very basic statistics to capture some correlation between an individual’s performance with point difference (and that is well correlated with win-loss records.)
Does that mean his model is perfect? By no means, and Berri has never said his model is perfect. What Berri has harped on is that his model differs from traditional player evaluations. And of course he would focus on the most dramatic differences. No one cares if Sebastian Telfair is ranked below average, because both eyeballs and econometrics agree. So we get his pronouncements that both Allen Iverson and Kobe Bryant are overrated, as their WP48 do not match conventional evaluations.
And another important point is that even if his WP48 model is empirically “better” (i.e. has a better fit to point difference) than PER or Adj +/-, it isn’t as if the correlation coefficients are so dramatically different, or that WP48 will predict the correct win-loss records 100% of the time.
So there is room for improvement.
However, claiming simply that we can simply give equal weight to some concept called “intangibles”, or to bundle some “intangibles” numeric factor to operate on WP48/PER/Adj plus minus is the wrong way of modifying models.
Even as a GM/fan/analyst who favors scouting, eyeball information, and hyper-detailed player tendency statistics, wouldn’t you wonder which of these scouts or observations are the most relevant? As commenters like Italian Stallion point out, basketball is complex. So, I would, as a rule, wonder about *your* observations that are you are using in constructing your “intangibles” evaluation.
I would question whether your intangibles of leadership, basketball IQ, dribbling, movement, positioning, coverage, and so forth would somehow remain hidden from box scores. As I’ve mentioned, it seems strange that a player who can move, can cover his man, can get open, can cut to the basket, can draw double teams, takes few low percentage shots, and hustles, wouldn’t get some benefits in terms of box scores.
But you know what? I have no idea if that is in fact the case. I have made a guess, and some commenters clearly feel that these things would not show up on the score sheet.
So let us assume that there is some affect of “basketball IQ” on box scores (which happens to be my point of view). Can I quantify how much of an effect there is? What would I use as a proxy for basketball IQ? Would my proxy be a good one? How would I choose among all the possible proxies? As for IS and Nick’s idea that there are in fact intangibles that do not lead to better box scores. I can pose pretty much the same questions.
It’s a complex game with a lot of movement. What are you keying in on that let’s you say, “Hey, this guy is a leader and is a positive influence on his teammates, gets his teams wins, but doesn’t get a lot of assists, steals, shots, rebounds”? And once you identified a characteristic, are you willing to say that this trait, if identified in any other player, will generally lead to immense success for his team (and somehow with that player *not* getting obscene rebounds/shot percentage/steals/blocks and so forth)?
And don’t forget, there are a lot of things you can identify as relevant factors. Such as his disposition on the bench (this is a Bill Simmons favorite), how “smooth” he is (and how would you rank this? Can you make a consistent rule to teach others to rank players in the same way?), how many double teams he draws, how often he dribbles on his weak hand, whether his feet are turned out or in, if he reaches out to grab passes, if he chases after loose balls, how close he stays on his assigned man, and so on.
We could all go on, but the point is that the indicator space is rather large, and it is conceivable the GM may not have the resources to keep tabs on everything. I should be clear and say that you don’t have to quantify it, but you should be able to articulate what it is the scouts and coaches should look for.
And yes, you must use some type of metric (it seems like we can all agree on point difference?), not other GM’s or scouts player evaluations. Otherwise, you are basically trying to explain some other person’s idea of how basketball players should be ranked, rather than coming up with a new and improved model.
So you end up doing something like Berri does: which is to build the model (in IS’s and Ray’s case, it would be a model based off a standard set of observations) and measure how often the appearance of these observations correlate to point-difference and win-loss. So you might keep a count on how many players you see with that particular trait, and how often they are on winning teams and perhaps how their teammates do when those special players are on the floor. You may find that model performs worse, as well, or better than models like WP48, PER, and Adj+/-.
One note: you can’t just focus on your team or one test case player. You need to apply your intangibles to all players. Otherwise, you might end up overfitting your model (i.e. making your model describe and explain one particular individual – which is actually kind of neat, but… it would reduce the model’s utility in estimating other players’ performance.)
However, when the commenters here talk about intangibles, I am not sure I see that they have these thoughts in mind. I am not sure if they are thinking about the degree to which existing statistics (either box-scores or some model like WP48 or PER) are influenced by the intangibles model, which intangibles are the most important, the reproducibility of detecting said intangibles, or how well these intangibles predict win-loss records and/or point difference.
antony
March 6, 2010
I doubt one persone actually read the above post. That is the longest most ridiculous thing I have ever seen.
mrparker, are you really trying to compare the Thunder to team USA or an all-star team. I mean… wow.
And talking about the effects of playing with James Harden, who only plays 23.2 minutes a game? And are you calling players such as Chris Anderson, Kenyon Martin and Nene not athletic players? And he has a great point guard in Chauncey Billups who can get him the ball (Chauncey is also a much more effective player then Westbrook).
I know you claimed that your “system” said Durant would never be above a wp48 .200 player, and now you are having trouble you are admitting you are wrong (despite the fact that you have never revealed ANYTHING about your system), but deal with it. Durant is playing at a star level and is only going to get better. Stop making excuses for you being wrong
dberri
March 6, 2010
Thanks NoTimeToRead.
Good to see the voice of science chime in on the debate. Once again you did a great job explaining the nature of an objective approach (as opposed to the subjective approach often favored outside of science).
mrparker
March 7, 2010
antony,
Its one season. Ariza had a .25 last season and look at him now. Durant is not a .2 player for his career yet. And even if he plays at this level for his career my system has still been pretty accurate and I can live with it. I didn’t think anyone would remember my system nor did I bring it up.
mrparker
March 7, 2010
And no, I’m just saying there a difference between playing with alot of guys who can pass and shoot and get out on fast breaks and make things happen.
Time will tell.
Michael
March 7, 2010
“You know who else had an unbelievably dreadful rookie season? LeBron James”
Seriously? You think Lebron was ‘unbelievably dreadful’ in his rookie season?
Really?
reservoirgod
March 7, 2010
NoTimeToRead – that was an excellent comment. Wow. Well done.
Palamida
March 7, 2010
Personally I Think the WoW model has more merit than all it’s “competitors”.
To be fair, IMO What most commentators that harp about “intangibles” mean is that some\most aspects of the game are too complex (5 on 5 and all that jazz) and thus aren’t represented (correctly) on the Box score.
naturally, we can expect a player with high basketball IQ, Leadership and so forth to gain from those qualities some gains that are represented in the Box score. They will argue however that some qualities mostly affect OTHER players (be it the opposition or a team member). What if a high level of “leadership” for example, leads teammates to higher levels of say… motivation?
This can manifest itself through higher degree of energy and effort leading to more Rebounds, Steals or what have you. Berri’s model in this example would not assign “production” to the “leader” whom by their logic was responsible for the production (or at least to some of it) but rather to the players who got the rebound, steal etc.
I think that’s the sort of thing most commentators who harp on “intangibles” are alluding to.
Personally I think such an effect does in fact exist but overall it’s rather marginal in effect.
The fact that some aspects of the game cannot be derived directly from an individual’s stats doesn’t change the fact that the majority of player’s impact on wins (and losses :P) is captured by Berri’s model and as in such, makes it an excellent starting point in evaluating the production of players in the NBA.
Surely it’s not an end-all model (not that anyone claims differently) but if you fell off the moon, never even seen a basketball let alone a basketball game and all you had was Berri’s model and it’s results, you’d know pretty darn much about whose a “good” basketball players and whose a “bad” one.
antony
March 7, 2010
mrparker,
As a team the Thunder have the 25th best assist ratio while the Nuggets have the 17th best. And despite the fact that Durant is a much more efficient player than Anthony, the Nuggets have the 3rd most efficient offense while the Thunder have the 17th. Denver also plays at the 5th fastest pace in the league, while the Thunder only play at the 13th. This tells me that Durant does not play with a lot of guys who can pas, shoot or get out on the fast break. At least certainly not in comparison to Anthony.
antony
March 7, 2010
And it is still frustrating to hear you mention this system and refuse to even give a general summary of what its about and how you go about evaluating players. If you don’t plan to divulge, then just don’t bother mentioning it. Nobody wants to hear about a “system” that can tell us A or B when we do not know how the system reached its conclusions
mrparker
March 7, 2010
antony,
You brought it up. I don’t bring it up if its not draft time. I’m not a professor so I have no incentive to “publish”.
The Thunder have such a terrible assist ratio because the player who shoots the most passes the least. The Nuggets have a more efficient offense because their point guard is Chauncey Billups and the Thunder’s point guard is Russell Westbrook. None of that explains when/where/how Durant is doing most of his damage.
I’m not claiming that the Thunder have better players than Denver. I am claiming that the Thunder are a better situation for a player like Carmelo/Durant. And by better situation I mean players who are at least a threat to shoot from all over the floor.
Denver= 3pt shooters and rebounders
OKC= more versatile if not better players
mrparker
March 7, 2010
I forgot to dispute your pace factor. Cleveland Cavs are 25th in pace. Do you also want to argue that they aren’t good on fast breaks?
mrparker
March 7, 2010
I should finish browsing around before I hit the submit button.
Numbers from 82games.com
Finally, Durant is assisted on 51% of his baskets. 70% of his baskets are jumpers of which 51% are assisted on.
I’ll compare him to Kobe who has about the same usage and a a high off rating (kobe 1.12 vs. Durant 1.15) and about the same wp48. He is only assisted on 38% of his shots and 38% of his jumpers. Comparing him to Dwayne Wade(a bit more of a stretch) who is only assisted on 30% of his baskets and 26% of his jumpers.
Durant is greatly benefiting from teammates who have the skill to find him and are looking to find him.
Travis
March 7, 2010
““You know who else had an unbelievably dreadful rookie season? LeBron James”
Seriously? You think Lebron was ‘unbelievably dreadful’ in his rookie season?
Really?”
Yes. He was extremely inefficient; one of the most inefficient players in the league that season – refer to my earlier post for evidence. However, his potential was too obvious to ignore (his ability to drive to the basket, the ability to pass to teammates in a position in which they could score, his height, strength, and athleticism); there is a large difference between somebody who is just bad at basketball, like Gerald Green, and someone who is otherworldly athletic and also moderately talented at the incredibly young age of 18/19.
mrparker:
According to 82games, Carmelo’s close and dunk assisted rates (43% and 70%) are very similar to Durant’s (38% and 67%), with Carmelo actually getting assisted more frequently in those shot attempts.
Essentially, Durant takes and makes many more three pointers, and is assisted at a very similar rate in his made close and dunk baskets to Carmelo in his made close and dunk baskets; I haven’t seen anything that argues that if Carmelo made more three pointers/took more three pointers, that his percentage of assisted jumpers wouldn’t rise from its current 41% level.
For a fun exercise, I looked at Carmelo’s third season in the NBA, and found that his assisted rate was 63%, and has been trending downward to its current 44% level; Durant’s assisted rate has fallen from 59% in his rookie season to its current rate of 51%. What we’re seeing here is a trend that younger superstar players, as their skills develop, begin to attempt more shots originated from their own offense, rather than from offense originated from teammates.
Carmelo Anthony, assisted rate and offensive efficency, since 2003-04:
04: 55% – 102
05: 63% – 103
06: 63% – 110
07: 60% – 109
08: 59% – 109
09: 48% – 105
10: 41% – 111
Durant, since his rookie season:
08: 59% – 100
09: 57% – 111
10: 51% – 115
It appears that Carmelo Anthony has already benefited from teammates who have the skill to find him and are looking to find him, at a greater rate than Kevin Durant has ever had in his career.
Tom Mandel
March 7, 2010
One might also think of “intangibles” vs. “tangibles” along the lines of “causes” vs. “effects.”
Causes show themselves in effects. High basketball IQ (a classic intangible — especially because the term is framed as if it were something that did get measured) shows itself in effects which are indeed tangible, can be measured, and very largely show up on the box score.
A similar term is “little things”, as in “he does all the little things that don’t show up on the box score.” Do they show up in the final score? If so, how?
Another such is “chemistry.” And this is an obvious case where we *invent* a cause for a desirable effect.
We only point to successful teams as having good chemistry. But if “chemistry” exists one could also imagine an unsuccessful team — one that had few really good players and a bad record — having excellent chemistry (and therefore doing as well as they can, tho nonetheless not well). But no one uses the term that way, and in fact it would be difficult to point to good chemistry on a bad team — a fact that in itself indicates that it simply stands in for a good record — for an effect in other words.
Anon
March 7, 2010
I seem to remember seeing once that lebron’s WP48 in his rookie year was average, not bad. his numbers weren’t that terrible really.
Lots of interesting comments here :)
mrparker
March 7, 2010
Travis,
I think my point was lost somewhere. Maybe I should have stated a larger point of reference. 51% is incredibly high for a player with such a high usage percentage and offensive efficiency rating. Of course Carmelo would have a higher assisted percentage when his off eff was lower, he was missing the shots he was creating for himself alot. The percentage has come down because he’s made more baskets on his own.
In other words once a player has an above average offensive efficiency and a corresponding low assisted on percentage of baskets we can count on that player to have that level of production. Until then we cannot be sure how much of that players production(at least offensively) is tied to his teammates.
antony
March 7, 2010
“The Thunder have such a terrible assist ratio because the player who shoots the most passes the least. The Nuggets have a more efficient offense because their point guard is Chauncey Billups and the Thunder’s point guard is Russell Westbrook. None of that explains when/where/how Durant is doing most of his damage.”
Are you claiming that Carmelo Anthony passes the ball? He averages 3.6 assists per 40 minutes while Kevin Durant averages 2.8 assists per 40 minutes. In other words, the highest usage player on both teams are both terrible at passing (and their ratios are very similar), meaning the teams’ assist ratio largely reflects everyone else. And also, if you are talking about having great teammates around them, how is having a worse point guard a better situation for a player? Isn’t that the most critical position in terms of a player being able to get you the ball?
In terms of fast break, while my first stat was a bit misleading, the Nuggets are 9th in fast break points while the Thunder are ranked 10th. Doesn’t seem to be doing Durant any favors.
And of course Durant is going to have a higher percentage of his baskets assisted. He is the only option on offfense! That is why, despite the fact that Durant is a much more efficient player then Carmelo, the Nuggets have a more efficient offense. For the Thunder to be efficient, they HAVE to get the ball to Durant. For the Nuggets to be efficient, they have a number of players who they can be counted on. Every player is not constantly looking for Melo because they have many more capable offensive players (indicated by their number 3 ranked offense).
mrparker
March 7, 2010
Antony,
I don’t know where to start.
Denver has a middle of pack team in assists even though they have a great point guard because Melo is a black hole. A team led by Chauncey Billups should be expected to be much higher than 15th in the league in assists.
OKC is 10th in fast break point though they are a slow paced team. Doesn’t that probably mean that they are extremely efficient when they do run like I said.
1.12 vs 1.15 is not an extreme difference as you suggest in efficiency as you suggest.
Durant has a lower usage than Carmelo Anthony. If his teammates were looking for him more than Carmelo’s wouldn’t he have a higher usage?
Nick
March 7, 2010
Apparently, I’ve missed a lot during the weekend.
Notimetoread,
You state:
“As for IS and Nick’s idea that there are in fact intangibles that do not lead to better box scores. I can pose pretty much the same questions.”
I cannot speak for IS but this idea was never my idea.
To recap:
In this thread, I was making a simple point. Berri was making a fuss that Thorpe had measured “intangibles” which, Berri states, by definition cannot be measured.
In fact, Thorpe didn’t measure intangibles. He didn’t tabulate a quantity. He used a point system to express his evaluation of the intangibles, but he didn’t measure them.
Berri was apparently confused by the numbers, however, since he stated:
“But it’s odd to see someone assign a number to something that by definition, isn’t tangible.”
The error here is that Berri appears to think that use of numbers always implies measurement. It does not.
So Berri’s criticism on that point (that Thorpe had measured something that cannot be measured) was off. Thorpe had done no such thing.
My point went no further, and I certainly wasn’t defending Thorpe’s larger point about the importance of intangibles when evaluating a player’s productivity. In fact, I stated:
“I think it’s perfectly fair to say that evaluating intangibles isn’t likely the best way to identify the most productive player on the court because intangible things like hustle or leadership do not have any great effect on productivity outside of what is already translated into the box score. I’d agree with that for the most part.”
So generally that was a fine comment on your part, but it didn’t respond to mine. IS and I are talking about different things.
Your comment missed that.
notimetoread
March 8, 2010
I apologize to Nick; as I review my comment here, I realized that I wasn’t being clear enough in writing about how basketball analysis can encompass observations that may be difficult to measure or to quantify, but can in fact be evaluated and described in a rigorous and systematic way. I believe this last point was addressed within the reply at WoW blog. There is another point that occurred to me as I read Nick’s post, so thank you, Nick.
I am not sure if Berri was confused or made an error in thinking that numbers always imply measurement. Sure, in the strictest sense, you can simply use numbers in the nominal sense: to name something. In that sense, Thorpe probably wanted to work within his framework and gave an arbitrary number that reflected how he felt about the intangibles Melo and Durant brought (i.e. Melo brought more to the table than Durant).
I think Berri’s problem is where that evaluation came from. (And the following is for other readers’ benefit, since you clearly understand the distinction between ranking and measuring.) You can perform statistical analysis on data that are binned into assigned categories. These tests fall under “non-parametric statistics.” One example might be, “State your satisfaction for this new cola: 0 = extremely dissatisfied to 10 = extremely satisfied”. In this case, you would wind up with a “score” that isn’t really a measurement, but is an expression of, in essence, a subjective opinion. But a test subject would answer more than one of these questions, and one can operate on these data to see if in fact a group of subjects is satisfied. I’m sure Berri is aware of this type of data and analysis.
I think Berri’s problem is that for something to pass as analysis, it ought to encompass more than some vague feeling you have. Look at the problem from a different perspective: If Berri posted WP48 scores for 2 players, one around .150 and the other around .250, and he simply said, well, the intangibles player 1 offers (he of the 0.150 score) make him equal to the value of the second player, you’d question what intangibles he was looking at. You’d want something more to explain why he felt he thought the scores needed to be modified and why the performance scores were not enough. And if Berri did in fact write about which factors he considered, wouldn’t you want to know the importance he placed on each factor?
This type of analysis is not numerical; it just lays assumptions bare so that all of us can consider the argument fully. There is nothing wrong with this type of analysis. As a matter of fact, Berri does this all the time, especially when discussing why players have not performed as well currently than in the past (age and injury are standard justifications, but eminently reasonable and most readers here seem to agree with these reasons.)
Basically, I thought that Berri was just being sarcastic, riffing on the fact that Thorpe actually didn’t make his “intangibles” argument substantial enough.
On another note: I had the opposite problem with Bill Simmons’s Book of Basketball. It was an enjoyable read, and I think I can appreciate basketball much more than before I read the book. But in his ranking of the top 96 players of all time, he was fairly specific about the things he valued and how well one player stacks up against another.
However, the problem was that maybe the language he used couldn’t convey the nuances of why players significantly differed from each other. It could just be the nature of language, but it was difficult to see how his analysis was organize and how different factors were weighed. And there are only so many words you can use before one gets the sense of, well, he did everything well but his passing was above average. OK, but how does that compare with the guy who does everything well but is a better than average starter of fast-breaks and a slightly better shooter? Even as a descriptive, qualitative tool, he could have benefited from making a model where could have encapsulated his observations and created a shorthand metric …
khandor
March 8, 2010
notimetoread,
——————-
re: “This type of analysis is not numerical; it just lays assumptions bare so that all of us can consider the argument fully. There is nothing wrong with this type of analysis. As a matter of fact, Berri does this all the time, especially when discussing why players have not performed as well currently than in the past (age and injury are standard justifications, but eminently reasonable and most readers here seem to agree with these reasons.)”
——————-
Bingo.
Nick
March 8, 2010
Notimetoread,
First, thank you for responding.
But to the point:
Berri was being sarcastic, which is simply a polite way of saying snide. Unfortunately, for both him and his readers, he was also being, at best, misleading and, at worst, plainly ignorant.
I doubt the second, so what likely has happened here as that Berri, as he is prone to do, was just attempting to get a laugh at Thorpe’s expense with a condescending jab. And this is particularly disappointing, since a) this blog is normally very good and b) Berri didn’t need to go there to show that Thorpe’s “analysis” was less than ideal.
But oh well, just as we’re probably all being far too harsh on Thorpe (I’m not sure that one needs/should need to have a data set to state what they think of two players) and I’m probably being a bit too picky with Berri.
So it goes.
Travis
March 8, 2010
mrparker:
I made a mistake when listing Carmelo’s assisted rate for this season. It is 44%, not 41%. Not super important, in the grand scheme of things.
“In other words once a player has an above average offensive efficiency and a corresponding low assisted on percentage of baskets we can count on that player to have that level of production. Until then we cannot be sure how much of that players production(at least offensively) is tied to his teammates.”
Is your conclusion, then, that Carmelo Anthony was not a reliable offensive scoring threat before this season, since his three seasons with above average offensive efficiency (110, 109, 109) came paired with assisted rates that were well over Durant’s current rate of 51% (63%, 60%, and 59%)? I’m not quite sure what you’re trying to accomplish, as the numbers you chose to use to support Carmelo and damn KD don’t seem to be consistent with your argument.
There’s a larger point to be made: why are you choosing to fight this battle, anyway? Just sit back and enjoy the coming greatness, man. I thought Durant and LeBron were horrible in their rookie seasons as well, and quickly had to change my tune about halfway through their second seasons. There’s nothing wrong with being wrong.
mrparker
March 8, 2010
Travis,
I can admit being wrong. Ive said as much in my previous posts. All I’m stating is that I’ve concluded that there isn’t going to be much difference between Anthony and Durant over their careers. Only time will tell.
For clarification, I’m stating we can’t be sure exactly how reliable a scoring threat will be until we see what is done at a low assisted rate.
mrparker
March 8, 2010
Travis,
Rereading through your comment I think that my position is being mistaken. I’m not stating that Carmelo is superior to Durant. I think they are both kind of worthless compared to what people consider them to be.
Frokostordning
March 9, 2010
Hm hm.. that’s quiet interessting but frankly i have a hard time understanding it… I’m wondering what others have to say….