Bill Simmons declares the following in the May 21st issue of ESPN the Magazine:
Fifty years from now, some stat geek will crunch numbers from Duncan’s era and come to the conclusion that Kevin Garnett was just as good. And he’ll be wrong. No NBA team that featured a healthy Duncan would have missed the playoffs for three straight years. It’s an impossibility.
Not to assume the label of “stat geek”, but had Simmons sent this article to me directly, he could have seen 50 years go by in a few minutes.
Let me start by noting that the above quote is taken from an article entitled “Tim Duncan is the Best Player of the Past Decade – And It’s Time You Noticed.” In this article Simmons “crunches” plenty of numbers. Yes, he does it in a non-systematic fashion, but his defense of Duncan does rest upon some numbers. Unfortunately, he doesn’t compare Duncan’s numbers to Garnett. All he does is declare that since Garnett’s team can’t consistently make the playoffs, he must not be as good as Duncan since the Big Fundamental is a playoff fixture.
Looking a bit deeper into the numbers, we see that there is a problem with The Sports Guy’s story. As has been noted in this forum more than once, the talent surrounding Garnett is generally quite bad. Consider the 2006-07 season. The top three players on the Timberwolves in terms of Wins Produced – without the initials KG – are Ricky Davis (5.5 Wins, 0.087 WP48), Craig Smith (3.6 Wins, 0.113 WP48), and Trenton Hassell (2.6 Wins, 0.057 WP48). Remember, average Wins Produced per 48 minutes [WP48] is 0.100. So two of Garnett’s top three teammates last year were below average.
Now compare this threesome to the top players on the Spurs who don’t have the initials TD. Duncan gets to play with Manu Ginobili (14.1 Wins, 0.330 WP48), Tony Parker (10.1 Wins, 0.194 WP48), and Brent Barry (7.5 Wins, 0.221 WP48). A bit of math reveals that Duncan’s top three teammates produced close to 32 wins. Garnett’s top three only offer about 12 victories.
And this is the same pattern we see throughout each player’s career. The following table details the production of each player over the past 10 years, as well as the wins production of each player’s teammates.
Table One: Comparing Kevin Garnett and Tim Duncan
Looking at these numbers we see that in the past decade Garnett has produced more than Duncan. And on average, Garnett has produced more than all his teammates combined. In contrast, Duncan’s teammates – on average – have nearly doubled Duncan’s production. The best season for Garnett’s teammates was 1999-00, when KG’s fellow Timberwolves produced 27.5 wins. The worst season for Duncan’s teammates was 2002-03, when Duncan’s fellow Spurs only produced 29.9 wins. As one can see, every single season of his career Duncan has had a better supporting cast than Garnett.
Bill Simmons, though, doesn’t see this. In the same article he states: “(Duncan’s) best teammates have been David Robinson (who turned 33 in Duncan’s rookie year), Manu Ginobili (never at top-15 player) and Tony Parker (ditto). In fact, Duncan has never played for a dominant team… Zoom through San Antonio’s past 10 rosters on basketball-reference.com some time. You’ll be shocked. Tim Duncan has never played on a great basketball team. Not once.”
Give credit to Simmons, he did correctly identify Duncan’s top teammates. But he has woefully underestimated the productivity of these players. Robinson played six seasons with Duncan. In these six seasons The Admiral produced 81.7 wins and posted a WP48 of 0.304. Ginobili has played five seasons with Duncan and in this time has produced 53 wins and a WP48 of 0.261. Like Robinson, Parker has also been a teammate for six seasons. Parker’s career wins production stands at 41.7 while his WP48 is 0.128. Amazingly, Parker has actually improved every season he has played, going from a WP48 of 0.040 his rookie season to 0.194 this year.
As for the “never a top-15 player” argument – Ginobili was the fifteenth best player in Wins Produced this past season (and the 16th best in 2004-05). In 1999-00, Robinson was ranked 9th in the league, while in 1997-98 The Admiral was the 5th most productive player.
In contrast, the best season by a Minnesota player (not named Garnett) was Sam Cassell’s performance in 2003-04. That season Cassell produced 12 wins, which ranked 20th in the league. In 2000-01, Terrell Brandon ranked 32nd in the league with 10.8 wins, while the season before Brandon ranked 25th with 11.3 victories. And those three performances complete the list of Minnesota players who managed to reach double figures in wins while being a teammate of Garnett.
As the above table noted, Garnett has been more productive than Duncan. The difference is not very big, but Garnett has offered more. Still Garnett has not seen much team success. Perhaps Simmons is right, though. Maybe if Garnett did more the Timberwolves would be just as successful as the Spurs.
Over the past ten years the Spurs have averaged 58 wins per season, while the Timberwolves have only averaged 45. How much more would Garnett have to do for his team to achieve San Antonio’s success? Garnett averages 23 wins per season. To move his team from 45 to 58, though, he would have to average 36 wins per campaign. Such an increase would move his WP48 from 0.358 to 0.559. Garnett’s productivity over the past 10 seasons is already more than any other player in the past decade. Still, if Garnett increased his productivity by 56% then the Timberwolves would be just as good as the Spurs.
Is it reasonable to expect on player to consistently produce 36 wins each season? Michael Jordan’s best season was 1988-89, where I estimate he produced 35.5 wins and posted a WP48 of 0.524. So we are only asking Garnett to consistently eclipse the production of Jordan at his best. Perhaps it would be easier if Kevin McHale – the only general manager Garnett has known – would simply find him better teammates.
The moral of this story is that players do not win games. Teams win games. If you surround an exceptional talent with other productive players, your team might contend for a title. But if you surround an exceptional talent with unproductive talent, your team has a good chance of missing the playoffs entirely. And there is nothing one exceptional talent can do to change that basic fact.
– DJ
Stephen Duncan Jr
May 15, 2007
It’s precisely this type of analysis that makes me wonder about Win Score and most other stats. I think it’s clear from watching Duncan and Garnett that Tim Duncan’s abilities and style of play lead to making his teammates look better, more so than Garnett (especially defensively). Isn’t it possible that one of the reasons Duncan can be considered better than Garnett is because he makes his teammates appear to be better? I
‘d love to see some analysis that tried to take this aspect into account, maybe comparing Duncan & Garnett’s teammates changes in Win Score playing with them vs without them (either in game, or maybe with previous/later teams).
Will
May 15, 2007
Stephen, I notice you use the words ‘watching’ and ‘look’ when explaining why you think Duncan is better than Garnett. This the point of Win Score (and economics/statistical analysis in general) – humans are prone to errors in judgement when using personal experiences alone to make conclusions. I do think your idea sounds logical, but so does the notion that the Earth is the center of the universe when you see the Sun and Moon revolve around it. Of course, statistical analysis of planetary motion proved this wrong centuries ago.
I also think your point about Duncan and Garnett’s differing ability to influence their teammates’ performance makes intuitive sense. However, I know there is a WOW post from about a month ago showing that KG’s teammates have historically been just as bad. Perhaps they could put up a post showing that Duncan’s teammates have been just as good.
In general, if you look at how accurately Win Score predicts a team’s actual Wins (see the post about their correct prediction of Denver and Philly’s records after the Iverson trade) you’ll see that it’s a much better metric than whatever sportscasters and columnists seem to be using. So if it’s better, why question it when it disagrees with popular opinion?
Stephen Duncan Jr
May 15, 2007
Yep, the words ‘watching’ and ‘look’ were meant to indicate that I have an unsupported hypothesis that says that Duncan actually is better than Garnett. From what I’ve understood of how Win Score is calculated, it doesn’t have any way of directly measuring this potential effect. Therefore I suggested that it might be interesting to come up with some way to analyze whether my hypothesis has any merit or not.
To reference another blog post here, the Wages of Wins does not say WHY players perform well or poorly. I’m curious to know if the statistics can be used to show (or imply at some level of correlation) that Tim Duncan’s teammates play well because they play with Tim Duncan, and that Kevin Garnett’s teammates play poorly because they do not play with someone like Tim Duncan.
anon
May 15, 2007
Stephen,
i actually have my reservations about Wages of wins, but I don’t think your argument is sound anecdotally.
As an abstract matter, there’s nothing that Duncan possesses that would make him obviously better than Garnett. Both are great defenders, efficient low-post threats, and great passers. Both play extremely hard. in the context of Nowitzki v. Duncan, you argument might have more bite. But I just don’t see Duncan as qualitatively better, ignoring stats.
Also, think about his surrounding talent. Wouldn’t manu and parker flourish similarly with Garnett? They are both fast, dribble penetration players. Things open up for them because people are worried about duncan. There’s nothing inherent in their play that would make them better with duncan than with Garnett.
I think wages has its limits. For example, I think wages doesn’t tell you that Nash was the reason the suns won last night. Wages doesn’t tell you that Nowitzki was the reason the Mavs lost. But in this case, I just think that deficiencies of Wages are inapplicable. Even anecdotally, the argument that Duncan is obviously better Garnett seems seriously flawed.
anon
May 15, 2007
another anecdote. Take the case of the Warriors. Now, if next the warriors were a .500 team that didn’t make the playoffs, no one would be really surprised. And no one would really blame baron davis.
Davis has been playing remarkably. In fact, he’s been doing everything in his power to make his teammates better. And for some reason, the team has just been able to ride emotion and Davis’ brilliance to make a historic run in the playoffs.
But the reason no one will blame davis next year is that we ultimately know that warriors are not a very good team. as brilliant as davis is, there’s not much he can do to change that over the course of an entire season. That’s what is so exciting about the warriors run right now. it shouldn’t be happening and it may not last that long.
Davis is playing out of his mind. And everyone has the sense that this just cannot last. You can only do so much for your team. At some point you need a solid surrounding cast.
And I think the same thing applies with Garnett. Duncan has a really wonderful surrounding cast. Who in this league would seriously take jrich over either parker of ginobili. both players would fetch a lot in the open market. they are really great, underrated players.
and a final point, the one year when garnett was surrounded by some talent, his team took the lakers to six in the western conference finals. and they didn’t even have cassell in that series.
Charles
May 15, 2007
Keep in mind that it’s not entirely fair to compare supporting casts, as those casts were constructed around their respective superstars, with the strengths/weaknesses of those players in mind. It seems that, in order to settle this, you would not only need to know how each player in both TD’s and KG’s supporting casts played when their respective superstars were not in the game, but also how they played with other teams (controlling for their other sets of teammates, as well as the point they are at in their career, etc.) In fact, the former set of information could be problematic, as a “supporting cast” is, by definition, set up to compliment the skills of a particular player, and might be especially likely to be “incomplete” without that player. Fortunately, the latter information would be easier to get, anyway.
Ken
May 15, 2007
Stephen, you completely ignored Will’s point. To illustrate:
Duncan tends to stand closer to the basket while his teammates have the ball. Garnett usually stands farther away.
Thus, when a teammate drives for a lay-up, Garnett’s defender is better able to help defend against the driving offensive player. This makes Garnett’s teammates more likely to be going 1-on-2, while Duncan’s teammates are going 1-on-1. Duncan’s defender is less likely to help, since Duncan is just a short pass away from a high-percentage dunk attempt.
Win Score, PER, and other systems do not observe where a player stands when other players attempt to shoot. Thus they may misattribute performance among individual players.
There are many other unobserved aspects of play that could produce similar misattributions. For example: willingness to set a pick, deliver a hard foul, move without the ball, and many aspects of team defense. People who know basketball would generally agree that Duncan is better than Garnett in many of these areas.
Some suggestive evidence is available from changes in teammate performance when they join or leave Garnett’s or Duncan’s teams. Players’ roles and abilities change over time, so take this with a grain of salt. But you can see some evidence for Simmons’ argument.
Jaric:
.414 FG% w/ Clippers in 04-05
.399 FG% w/ T-wolves in 05-06
Mike James:
.469 FG% w/ Raptors in 05-06
.432 FG% w/ T-wolves in 06-07
Ricky Davis:
.464 FG% w/ Celtics in 05-06
.429 FG% w/ T-Wolves in 05-06
All three are “slasher” types who are potentially affected by the unobserved positioning variable discussed above.
The larger point is that Win Score, PER, and the rest of current-generation basketball statistics should be seen as a complement to informed basketball observation, but never as a substitute. It is possible to correct the deficiencies described above, but doing so will require much better data and a more sophisticated class of models.
Brian
May 15, 2007
I think that stats miss the most important difference between Garnett and Duncan. Duncan has always made it his mission to put other players in a position where they can make plays and develop their own games. His style of play allows others to get higher WP. I think an essential addition to WP would be a somewhat Lenovo stat about how one player’s presence in a game raises others’ WP. I also think Garnett’s very stat oriented play has hindered the development of young guys. Szcerbiak, had he played with Dunacn, probably would have become twice the player he is today.
Jason
May 15, 2007
Typically with sports in regards to history (how good athletes are perceived), the judgement is made on 3 items: stats, how one remembers that athlete (viewing), and postseason performance (or more likely championships).
Stats, as crunched above (plus traditional stats) could support either player but only at a slight edge. How one views the players could also support either player, though I would also contend only at a slight advantage. What will ultimately break the tie will be postseason performance and rings. To this the edge is Duncan.
Therefore, given now (2007), Duncan would be viewed, historically, as the better player. Perhaps only slightly. Obviously, neither is finished with their career so anything can change – KG becomes a champion, either player’s stats corrode at a higher rate, etc .
Paul Swanson
May 15, 2007
With cherry-picked shooting percentages being introduced to the thread, it should be noted that Ricky Davis bounced back to hit 46.5% in 2006-07 — with career bests in both three-point and free-throw accuracy — while Marko Jaric posted career-high percentages in all three categories this season. Make of it what you will…
Jason M
May 15, 2007
One other possible explanation in defense of Simmons’ piece might be that, given the superiority of Duncan’s teammates to Garnett’s, Duncan is better served by sharing the ball with these superior teammates more often than Garnett is by sharing with his. Thus, Duncan’s stats won’t (can’t) be as high as Garnett’s, but Duncan may still be a better player. I believe this notion has been brought up in other posts on WoW – namely that having a great supporting cast leaves less weight for the superstar to carry and thus gives him less opportunity to take the team on his own shoulders.
Chip
May 15, 2007
One thing to possibly look at is, how many more wins could Minnesota’s front office add in supporting players if they got to use the extra four million a year that Garnett makes over Duncan
Adam Hoff
May 15, 2007
You can round and round forever on stuff like this. Sure, the $4MM extra could help, but so would the GM not cheating to sign Joe Smith and costing his team almost every pick this decade. You can’t really compare these guys, because KG plays for a poorly-run franchise and Duncan plays for the flawlessly managed Spurs. Apples and oranges already.
The one thing that made me laugh is the comment that KG is so stat hungry that he makes his teammates worse. Have you ever watched Garnett play? I can’t think of a less accurate description of him as a player and a person. I guess he is a little greedy though when it comes to rebounds. That bastard! He’s so selfish, snatching all those rebounds away from the other players. He should let them go and hope Ricky Davis grabs them. I mean, let’s be serious here.
I will say that the whole “standing closer to the basket” thing brings up a good point; that stats should be used to augment, not replace other evaluation methods. Another example of how traditional stats (that are then used to produce more complex stats like WP48 or PER) don’t tell the whole story is when a dangerous guard guts the defense, draws four players, and then misses the shot, allowing a teammate the world’s easiest dunk opportunity. Barbosa (when he’s not intimidated, as he has been lately) does this. WofW Public Enemy #1 Allen Iverson used to the be the king. He made Tyrone Hill a whole lot of cash on these plays.
I wrote a column about this that you are all welcome to check out if you are tremendously bored: http://www.whatifsports.com/insider/default.asp?article=20060530
chris
May 15, 2007
Garnett is not the player that Duncan is. I dont care what stats say. Duncan is a better post player(the most important part of playing inside) he is a better on the ball defender and a better defensive player overall. The big ticket is a nice player, but Duncan is head and shoulder above him. I dont care what stats say, this isn’t baseball. Wins produced? Come on, David Robinson was a big soft underachiever his whole career until Duncan showed up. WATCH THE GAMES, don’t just analyze stats and you would realize there isn’t even an argument.
Will
May 15, 2007
I’ve read through these comments and the only one that seems to have any validity (in my opinion) is what Jason M wrote. He’s right, it has been shown on WOW that having superior players does negatively impact one’s stats, albeit marginally. So it is possible, probable even, that Duncan is slightly better than his stats indicate, and Garnett is slightly worse. I’m writing from memory here, but I really don’t think the difference is enough to make a big difference in their respective Win Scores – though it would certainly help if Dave could write an addendum to the post that factors in the average teammate skill difference.
Ken suggested that Stephan was missing my point, but Ken, I think you missed mine. You went on to talk about positioning, and how Garnett’s supposed tendency to stand away from the basket led to 2 on 1 defensive situations when Garnett’s teammates drove to the basket, while Duncan’s defenders were more apt to stay near the basket and near Duncan. But I think this actually would prove the opposite, since as you point out, a near-basket shot is a higher percentage shot. Why couldn’t Garnett’s teammates simply pass the ball into the post? It seems as though, according to your reasoning, Duncan’s positioning would ‘keep’ his defender in an area that is more important to defend.
Ken then went on to talk about how positioning and other ‘unobserved’ (?) aspects of play could impact a player’s stats/contributions, and that basketball experts agree that Duncan is better than Garnett in these areas. First, some evidence should be offered to prove this point. But second, these ideas ignore the whole point of WOW. Basketball is incredibly complex. WOW attempts to pare down the complexity to only the statistics used in the calculation of Win Score, all of which can be found in the box score. And as the writers have often posted, the Win Score and Wins Produced of a team’s players can predict the actual wins of a team extremely accurately – I don’t feel like looking this up, but I think their accuracy is somewhere over 90%, 95% even. So if their metric is accurate, then it follows that the scores used to predict a team’s wins must also be an accurate reflection of how ‘good’ each player is. Therefore, there is at least a 95% chance that Garnett is better than Duncan in terms of helping his team win games.
Chris, with all due respect, how did you find WOW and what made you decide to post something? Have you actually read any of the articles?
Adam Hoff
May 15, 2007
I am generally a fan of efforts like WofW, but there is danger in creating a system – regardless of how simple or complex it is – that comes from box scores. So many basketball stats are flawed. Just take assists. If you throw the perfect pass and your teammate lays the ball in as the defender opts not to pick up a foul, you get an assist. If you throw the same pass, but the teammate gets hacked and then makes both free throws, you don’t get an assist. Both passes led to two points. Both passes were valuable. Both are plays that lead to wins. But one counts as an assist and then counts toward things like PER and then, eventually, is one of the building blocks that is used to actually say how many wins that player is worth.
Again, I’m not hating here. Just noting that while this formula is impressive and valuable given the materials (basketball stats) with which we have to work, we should be cognizant of the fact that a lot of those materials are flawed from the outset. Nobody’s fault; just the way it is. So you sort of have to take the conclusions with a grain of salt.
It doesn’t mean that saying “KG is better than Duncan” is an incorrect statement; it is just a statement that I don’t believe can be proved purely through statistical analysis.
That said, since I will never forgive Duncan for his whining, moping, dumb-foul-committing performance in the ’04 Games, I will take Garnett. I’m patriotic like that.
Tommy
May 16, 2007
Come on don’t you see!! Duncan is obviously better. Hi try every off season to actually lower his FT%. In 01-02 .822FT% but in 06-07 .629FT%. That’s a Great Achievement!! Hi try so hard to get his teammates involve that actually lower his APG also, Great Achievement!!
In the other hand KG the stat junkie who never gets his teammates involves, never try to make them better actually lead to the only All-star year in Sam Cassell carrier!! What a Mystery!!
And for those who say because Duncan teams win more than KG hi is a better player. Follow that reasoning, I think Darko Milicic (who have a champion ring) is a better player compare to Karl Malone or Charles Barkley!! Really it is!! Milicic the next best thing!!
Phil
May 16, 2007
re: Adam Hoff
Really liked your whatif article, i think a similar thing happens for situations where a defensive player tips a ball away from the player in posession, but another teammate picks up the ball and the steal. This also applies to saved balls, tipped rebounds etc…. I also seem to remember an article somewhere about making stats for this and things like charges taken etc….
Will
May 16, 2007
Adam, if ‘so many basketball stats are flawed’ then why is Win Score able to predict wins so accurately?
I agree that the hypothetical situation you describe can and does happen. But KG and Duncan have both been in the league for at least a decade – are you suggesting that over 10+ years one has suffered from far more bad luck than another? They both play against the same teams so both are subject to the same random chances that you describe. It seems that you are taking a sample size of a single play and extrapolating it to apply to multiple games/seasons.
Ryan
May 16, 2007
Bottom line, is that Tim Duncan has always played with a better GM, Coach, and team than KG. Head to head, KG’s stats are better than TD’s. But TD has helped his teammates improve, not because of the type of player he is, but because of the system he is in. If they switch places, this argument is reversed. If KG is put in a system that has creates an opportunity for success, he gets credit for making his team better. The lack of quality coaching and managing is responsible for no playoffs in three years, not Garnett’s inadequacies.
Will
May 16, 2007
Well said Ryan, I almost totally agree.
The most fascinating – and wide-reaching – implication of Win Score is that coaches don’t matter much. A team’s success is almost totally dependent on the GM (barring injuries). Give Phil Jackson Jordan/Pippen or Shaq/Kobe and he’s got 9 rings. Give him Kobe and Odom and you’ve got two first round exits.
steve
May 16, 2007
Hard to compare…duncan would have better career stats if he never played with robinson, however he would not have had the rings. Garnett is definitely a great player, but theres just something that duncan posseses that garnett is missing…i dont think duncan would have ever missed the playoffs regardless of what team he was on or or he played with. I think he is a rare type of player who finds ways to win and make everyone around him better. I dont see garnett as someone who does that.
Andrew
May 16, 2007
It is worthless to try to start any sort of discussion about player performance here because David Berri has made it clear he has no interest in speaking up about NBA player performance outside of what wins produced spits out. This blog isn’t so much about basketball but about a single metric. Everyone, including Berri himself I imagine, knows that WP is not the be-all-end all of player evaluation. The fact that Berri refuses to engage in questions like what is WP not measuring and whether Duncan improves his teammates WP or is he just blessed with better teammates is what makes this blog somewhere to glance at to find one opinion but no real analysis. One number that attempts to quantify all a player brings to the court without further analysis or exploration is of limited interest. Analysis like Ken’s is interesting, it provides something to watch for in game situations. Where does Duncan typically position himself on offense? Where does Garnett? Bill Simmons offers interesting anecdotes such as how Duncan aides on court communication. Just providing WP/48 numbers and refusing to engage in further discussion does not make for a compelling blog.
Adam Hoff
May 16, 2007
Will, I’m not saying anything is wrong with the results. I was just pointing to one particular play that shows how stats can be dicey. I have no way to prove this, but I bet LeBron leads the league in “near assists” – passes that lead to two points at the line. This seems to happen 3-4 times a game with the Cavs. If he somehow got credit for these, it would impact his Win Score stats, no? I just think there are elements of basketball that go beyond the box score.
Will
May 16, 2007
Adam, I think we’re arguing past each other to an extent. I agree that stats don’t tell the whole story. But what they don’t tell doesn’t have much of an impact on games in the long haul. If these other elements did have a big influence on game outcomes then the results based on Win Score would not be nearly as accurate, right? You’re saying nothing is wrong with the results, but you’re also saying something is wrong with the data used to calculate the results.
I think the main thing people are arguing about in these comments is a matter of scale. People see things in individual games that contradict Win Score. Hey, the Warriors just beat Dallas, right? – But Win Score said Dallas is a much better team. Well, I think the 82-game season proved Dallas was better, but since the 7-game series was a small enough sample the outcome was able to be influenced by these ‘other elements’ you mentioned. So I still think the large sample size of Duncan and KG’s careers is enough to base solid conclusions on.
Also, having a 7’3 center probably helps LeBron get more assists than he would have otherwise.
Adam Hoff
May 16, 2007
Yeah, I suppose you are probably right. The predictive ability of the system is what is winning me over. I didn’t realize at first, but I had gone more in depth on the LeBron thing a few weeks ago: http://wisinsider.blogspot.com/2007/05/case-study-cant-trust-assists.html. Now I probably wouldn’t bother to write that piece, because I would see it as more unique to that game, and a statistical abnormality that would be absorbed over a bigger sample size.
I’m going to pull the book back off the shelf and read some of this again, more carefully. I’m coming around …
Moni
May 16, 2007
I’m a big time Garnett fan, but i can confidently say that Duncan is a superior player. Garnet doesn’t really have a low-post game other than his baseline fade-away. Duncan passes out of double teams better as well. KG is a better perimeter defender, but when it comes to controlling the middle, duncan is superior. KG has better perimeter skills but his greatest strength is also his greatest weakness. He’s good at everything, but he’s not GREAT at anything. I would want Mike James to take the game winning shot over KG.
Pacifist Viking
May 16, 2007
Garnett has led the league in rebounds per game the past four seasons, so I’d say he’s at least great at rebounding. Considering the Wolves have given him no help on the boards, he’s almost required to lead the league in rebounding to give the Wolves any shot at all.
Will
May 16, 2007
Ok Moni, I’ll take your word for it. I’m convinced because you’re ‘confident’.
(not the) Jason (who posted above)
May 17, 2007
The quality of Duncan’s teammates does seem to be better. Perhaps he has had something to do with their performance, perhaps not.
It’s pretty clear that statistically, both are big-time contributors. While Garnett appears to have a slight edge with WP, it’s close and goes back and forth. If it’s true, as Dave has commented, that there’s a diminishing returns at some point, that Duncan has seen more production from his teammates may deflate his score and Garnett may be allowed to show somewhat better. Statistically, judging between the two seems to be splitting hairs.
But Simmons contention that Duncan hasn’t played with *substantially* better teammates is absurd, and this without resorting to WP. I’m not sure what a “top 15 player” means, or who Simmons gets to rank this, or whether there’s something significant about “top 15” but by the general hype value of NBA teammates, Duncan has had the better supporting cast. I don’t think that anyone can honestly say otherwise. Certainly those who put people on the all-star team don’t seem to think the non-Garnett T-wolves are as good as the non-Duncan Spurs.
Adam Hoff
May 17, 2007
I would add that Ginobili arguably carried Duncan to a title in 2005, not the other way around. We have short memories, so nobody really remembers that Manu was sick that year. Maybe they will now that he had a big game last night.
dberri
May 17, 2007
Wow, 33 comments. I think [(not the) Jason (who posted above)] captures the essence of the post. The point I was making was not that Garnett is clearly “better” than Duncan. The point is that Duncan has been able to play with much better teammates than Garnett. And if Duncan spent the last ten years playing with Minnesota’s collection of players, Duncan would not be thought of as the best player of the last decade by Simmons.
I would add that I find it hard to believe that it is because of Duncan that the all the Spurs not named Duncan are twice as productive as all the T-wolves not named Garnett. Perhaps I should post on this later.
Nils
June 15, 2007
The formula used in WoW rewards teams for defensive rebounds, because getting one signifies that your team has both stopped the opponent from scoring and has regained possession of the ball. Fair enough, in predicting team wins.
But most defensive rebounds are essentially uncontested; getting one does not mean that the rebounder is a great player, but plays for a team that has accomplished its defensive objective. A 2-guard who plays in-your-shorts D and gets a hand in the shooters face (think Trenton Hassell or Lindsey Hunter) is rewarded by Berri’s system with the statistical honor of playing with a numerically great rebounder (e.g., Garnett or Ben Wallace).
Under Berri’s system, the ideal player would spend all of his time underneath the relevant basket and call timeout whenever he did not have an opportunity for an uncontested dunk.
Prescott Home Auto Insurance
June 26, 2007
Garnett is good, very good. But some players just don’t rise to superstardom, like Tim Duncan. Garnett has the talent but maybe after playing on a team that never competes for the top of the Western Conference each year takes its toll.
I really don’t think anyone will confuse Duncan and Garnett given the titles San Antonio has won.
bayesk
July 4, 2007
How about this test, one that Simmons likes to use himself: one pick-up game with your life on the line, you get to pick first from current NBA players. Tim or Kevin? How ’bout all time?
steve
August 4, 2007
this is extremely fascinating;I’m a long time Knick fan, and Simmons for one Skewers Ewing on a regular basis. Could I possibly get a hold of a similar analysis of Ewing and his teammates W per 48 for his Career. Now that Garnett is a Celtic, we’ll see if his tune changes.
mrparker
August 11, 2007
Nils,
You are missing the significance of a
defensive rebound. If I play in your
face defense for 24 seconds and contest
a shot from the best 3 point shooter
in the league that causes a miss only
to have the shot offensively rebounded
then my great defense did not matter. Now I
have to run around picks for 20 more seconds
or worse whoever recorded the offensive
rebound got fouled and possibly made a put
back which is a situation that happens more
often than not.
After playing defense for 40 or 60 seconds
at a time, my ability to contest shots
is greatly diminished as almost noone is
in good enough shape to crouch and slide
for minutes at a time without becoming
fatigued.
However, if I have Tim Duncan or Kevin
Garnett on my team this never becomes a
problem. So while I may be a good on
the ball defender it won’t matter unless
I am paired with the right teammate.
For this reason I recieve no credit for
being a good on the ball defender, because
I need the great defensive rebounder
to be successful. Instead, the credit is
given to the defensive rebounder.
Their is only one reason why the Bulls
were a 72 win team in 96 and he was a
blonde dude who scored less than 10 points
a game.
bobby
August 18, 2007
simmons is mainly saying that their wins production totals and productivity aside- he is saying that duncan makes them look better. those players placed on the 2007 celtics if you put ginobili in paul pierces place he would be nothing. and if you put pierce in ginobili’s place he would look like jordan resurrected……….. he’s saying those players are a product of the system, the coach, the team and most importantly A PRODUCT OF DUNCAN.
bobby
August 18, 2007
theres only one way to find out. for duncan and garnett to go head to head in the finals this year with equal supporting casts. garnett has ray allen and pierce on his side which is a better top 3 star players, and duncan has the depth, the chemistry, plus his own 2 that arent too shabby in parker and ginobili………… both hwill win around 60 games this year and make it thru the playoffs to the finals and go to war with each other as two powerhouses. once and for all we find out whos better (hakeem/robinson style!!!) between duncan and garnett! i really think this is gonna happen by the way, the celtics with paul freakin’ pierce and ray freakin’ allen with kevin freakin’ garnett those 3 superstars are gonna STOMP thru the east…. and the spurs will be there waiting for em!
Peter
August 19, 2007
You missed Simmons’ point. His argument was that what Duncan provides can’t be measured by statistics. That his presence inflates the effectiveness of Ganobli, Parker, and Robinson. You are the statistician who crunched the numbers, you have concluded that Garnet is just as good, and you are still wrong. Statistics can not describe the effectiveness of Tim Duncan.
Mr Parker
September 2, 2007
Dave,
Why’d you even bother publishing your findings. Noone is willing to listen as they are all stuck in their ways. You must get sick of the snarky comments from the conventional side of the argument.
dberri
September 2, 2007
Mr. Parker,
Don’t take things too seriously. It takes time for new ideas to gain acceptance. One thing to keep in mind…many baseball fans still don’t embrace statistics.
Kevin
September 8, 2007
Timmy makes his teammates better. This article doesn’t consider that Duncan’s inside presence makes it easier for his teammates to get open looks and driving lanes. Kg just shoots fade aways all game.
William
October 27, 2007
Suppose Jason Kidd, Kobe, Lebron James, KG and Duncan play on one team (with other quality backups as well), and win 70 games. Then each of them will have a win production of something close to 12-15. So, playing on a good team (e.g. having good teammates) will have a serious negative impact on the star player’s “win production”. Therefore, your analysis has serious bias that favour those who play with bad teammates (e.g. KG) but underestimate the importance of star player (e.g. Duncan) who have good teammates.
Marshall Jevons
December 4, 2007
What about practice? If you look at all the mind-boggling moves Tony Parker and Ginobili make around the basket– and their respective abilities to penetrate and create offensive opportunities for their teammates, you must consider how they honed this skill set. Could it be practicing every day against two of the absolute best low post defenders in history to start their careers? Could Duncan’s (and Robinson’s, though to a lesser direct degree) defensive prowess make Parker and Ginobili great, not only on the game floor, but also on the practice floor?
How do you quantify practice? It’s pretty clear that Tim Duncan is the most coachable superstar in the NBA (see the Spurs preseason video by clicking the link provided on my name). Maybe ever. He takes direction from Pop like a marine, and if you ever watch the difference between a Spurs practice and other teams’ practices, you see one thing. The head coach really rides the Spurs’ stars and the other coaches (especially George Karl with Carmelo, Garnett with Flip Saunders, and Bob Hill with Ray Allen) don’t really give their star players much direction, seemingly because their advice is unwelcome.
[Email me if you’re interested in the feel of how coaches run their workouts the day before games, nearly all the Spurs’ opponents practice in Trinity University’s athletic complex the day before games, and I’ve been able to see quite a few over the years].
If Garnett doesn’t make his teammates better in practice (his legendary work ethic may not necessarily translate into team unity or crisp passing or players actually taking heed of the coach), and I submit that he does not, his value on any team is muted.
SHC
December 27, 2007
How about comparing playoff performances of both players? Isn’t that a better indicator of who can perform under the greatest pressure against the best competition?
Using the NBA.com’s stats, I got these win scores per 48 mins:
Duncan : 16.3511 over 138 games
Garnett : 14.9070 over 47 games
Granted, Duncan played in almost 3 times as many playoff games as Garnett, but 47 games are still more than half a season’s worth of games.
One interesting thing is that Garnett’s shooting percentage drops way below his career percentage. This might be indicative of his dependence on his jump shot and his lack of an ‘unstoppable’ offensive move.
magicmerl
December 29, 2007
Great point SHC.
One factor that might not have been considered much by others is the different contracts.
Garnett too the fattest contract in history, and more power to him, because he was worth that much. But because of the salary cap that meant that the wolves *couldn’t* put talent around him to allow him to contend for a title.
Timmy on the other hand settled for a modest 16mil (he left about 6mil per year on the table when he signed his first post-rookie contract, money that essentially equals having Manu Ginobli on his team), and every time he resigns an extension he takes a pay cut back to 16 mil. (p.s. Bruce Bowen opted out of his contract to resign for a lesser amount a few years ago, also so that the spurs could resign Manu. now that’s a championship play).
Garnett is doing the same this time, which could bode well for his Celtics era. But he was a salary cap hog in minnesota, which is why he never had good teammates.
Does this make sense to anyone?
Jeremy
June 25, 2008
Looks like KG is better eh
penbeast
March 6, 2009
Two minor comments.
(1) The idea that Duncan standing closer to the basket opens up driving lanes for dunks seems counter to everything I learned playing. A player who can take his defender AWAY from the basket opens up the lanes. There is an interesting article on apbr.com about how big men who shoot more from outside have an impact on team efficiency greater than their individual contribution (which tends to be less efficient than inside players) . . .
The comment about salary may be significant though; it’s at least an interesting point as to whether a player’s willingness to leave cap room for his team is connected to the team’s ability to win in the modern game. Seems likely to be correct.