Just heard a comment at halftime of ESPN’s broadcast of the Magic-Heat game. I don’t know which analyst offered this thought, but here is what was essentially said: “What Raymond Felton is doing this year is not that impressive because any point guard will put up numbers in the system of Mike D’Antoni.”
Okay, here is a quick check of the WP48 [Wins Produced per 48 minutes] of point guards who have played at least 500 minutes for the Knicks under D’Antoni:
2007-08
Chris Duhon [0.090 WP48]
Nate Robinson [0.125 WP48]
2008-09
Chris Duhon [0.069 WP48]
Nate Robinson [0.074 WP48]
Sergio Rodriguez [0.022 WP48]
So prior to this season, only Nate Robinson posted an above average mark (average WP48 is 0.100). And Robinson was below average last year.
Entering tonight’s game, Raymond Felton has played 550 minutes and posted a 0.208 WP48. Last season Felton played 2,470 minutes for the Charlotte Bobcats and posted a 0.159 WP48. Yes, Felton is somewhat improved (although it is early). But Felton was above average before he came to New York. So I am not sure we can argue that he is above average just because of D’Antoni. This argument is especially problematic given the performance of point guards in New York in 2007-08 and 2008-09.
Let me close by noting that the Knicks entered play tonight on pace to win about 39 games. About 13 of these projected wins can be traced to the play of Felton. And about 15 can be linked to Landry Fields (and only three to the play of Amare Stoudemire).
– DJ
P.S. Hope everyone has a Happy Thanksgiving!!
Dino Gunners
November 24, 2010
Hey DJ,
I have an unrelated questions but relevant to the wins produced metric.
Two general conclusions you have established with your measure is:
1. The stats stays fairly consistent over time (ie. it isnt too volatile)
2. It directly links players to wins produced
In summary the whole point of your wins produced measure is that it can directly link wins to individuals irrespective to their teammates and thus, could be a good predictor of future team performance
Last season the “big three” WP numbers of:
Lebron .433
Dwade: .299
Bosh: .254
This season these are the heat’s numbers (14 games played):
Lebron: .268
Dwade .207
Bosh .187
Which shows that all three have played significantly worse playing together.
It is also much worse than their performance last year and throughout their career (a sign of volatility). You cannot argue lack of synergy or lack of understanding because the “big three”, because the whole point of WP is to isolate wins to individuals regardless of teammates. Also, if they are each using less possessions, it should make them more efficient given they are not having to take unproductive shots because of the individual skill level of the ‘big three’ (they could just pass it to another who has a better shot opportunity).
Andres Alvarez predicted 65 wins based on their players and Arturo Galletti predicted 68.7 wins adjusted for 19660 minutes played.
Given that the Heat are 8-7 now (worse at this point than when only Dwade was on the team with a bunch of scrubs), you have to admit that your WP measure is not accurate in determining individual wins and severely underrepresents defense, coaching and “intangible” factors.
I remember a conversation in one of the comments where a poster said that synergy and teamwork matters and that a team of 5 he chooses would be better than the top WP of each position. He/She got laughed off and other posters basically implied it would be impossible to have a better team the the top “WPs” of any given year.
dberri
November 24, 2010
I have to “admit” something based on a sample of three players, across 15 games, where one player has been hurt?
I assume your serious. But really, seriously?
some dude
November 24, 2010
Not sure what a player being hurt has to do with it since WP48 is an individual stat.
But I am curious how many games need to be played before we can start trying to draw conclusions? Oh wait, Arturo addressed that.
“The second column tells us that once we have a 15 games sample the correlation is above 75% (which is good). 20 games is above 80%, 30 gives us 90% and 40 games is almost a lock at 94%. So at this point in the season player productivity for the full year can be predicted with about 70% accuracy (depending on sample size). In two more weeks this should be close to 80%.”
http://arturogalletti.wordpress.com/2010/11/18/a-guide-to-jumping-to-conclusions-in-the-nba-and-employee-of-the-year-for-2009-2010/
Italian Stallion
November 25, 2010
Personally, I think it’s fairly obvious that part of the problem in Miami is diminishing returns among 3 players who have all typically generated a lot of their value from very high usage scoring. There simply aren’t enough shots to go around and any “potential” offset in the form of higher overall efficiency hasn’t been realized yet partially because Wade has been hurt, James is shooting terribly from 3, and they are still figuring out how to maximize their value. The sample is small.
I still think the trick is going to be for each to realize that if he’s shooting less, he has to hit the offensive boards, make plays and do other things that add value elsewhere.
As far as wins are concerned, Wade has been hurt, Miller has been out, and now Haslem is out. That’s having an impact. IMO, as soon as they get healthy they are going to start playing a lot better.
Italian Stallion
November 25, 2010
A lot of people predicted that Felton would do a little better in NY because his specific skill set was a better match for D’Antoni’s open court running offense than Brown’s slow half court style. I had no opinion on that because I didn’t know his game well enough, but he does seem to playing better right now.
It’s not so much that any PG will play better for D’Antoni.
It’s that some PG’s fit some systems better than others.
How would Nash do in Phil Jackson’s Triangle?
My guess is not nearly as well.
Dino Gunners
November 25, 2010
DJ, I agree it is a relatvity small sample and maybe “admit” was too strong a word. But, given WP is an individual stat, how have these players not been able to produce from the start? Can you really say that 3 of the top 20 players in the league (by WP last year) need more than 2 months in the summer and 15 games together to really learn to gel? Haslem and others should have no effect on the WP of Lebron James or Chris Bosh because WP is an INDIVIDUAL stat, you are assigning specific wins to specific players.
IS, you’re theory of these players being good ‘high usage’ is probably correct and their productivity lowers with less usage, but WP doesnt or can’t take this into account, and thus the measure is flawed. It should take this into account because WP has been labelled as one of the best predictors of future team performances.
My qualitative observation is that the Heat are terrible at help defense . It has been clear that Lebron, Bosh and Dwade have no idea how to play in a team structure where they have to sacrifice for one another for the greater good (or just to have good team defense). WP also can’t take this into account.
I am not trying to be overly critical of your measure but I feel there is clearly some room for improvement, especially in the area of defense and usage.
some dude
November 25, 2010
IS – the issue for Wade and Lebron isn’t the lack of shots. Lebron is only 2FGA per 36 min less and Wade 3. But look at those percentages!
If you look at their shot chart breakdown, it’s easy to see why. They’re shooting less at the rim. Lebron’s dunks are cut in half. Their outside shot percentage is terrible and for good reason. Teams are doing what we said they would, pack the paint and go under screens.
As I always repeat, it’s not WP48 that matters alone, but how you get yoru WP48. Wade and Lebron are redundant. They get their WP48 the same way. Until one of them finds another way to do it (a part of it), the struggles will continue.
The other issue is their defense, while solid, doesn’t generate runouts for the offense. lebron and wade both lose lots of dunks, layups, and FTs because of this. They got them on the previous teams but are not getting them now. Lebron’s ability to be a roamer on D has also been somewhat compromised.
Not sure how they fix the D, but until one of Wade or Lebron learns how to play differently, meaning off the ball, the same issues will continue to haunt them against legit teams. You need players that do different things.
some dude
November 25, 2010
Dino and IS. Lebron’s usage is actually the same. Wade has dropped a bit and Bosh has had a massive drop.
Wade and Lebron’s tunrover rate has shot up and Wade’s assist rate has gone down. Both are shooting poorly.
Owen
November 25, 2010
It will be interesting to see how things turn out with Felton. If memory serves duhon got off to an obscenely hot start before slowly reverting to his long term mean. I wouldn’t be surprised if something similar happened with Felton.
Re the heat comment, let’s see how things are at midseason. The heat have been a lot better than their record indicates. It’s way too early to count them or the WOW out.
Gil Meriken
November 25, 2010
Yeah, but if you replace Kobe on the Lakers with either Wade or Lebron, the Lakers would be just as good, or better … (where’s the sarcasm indicator when you need it?)
stephanieg
November 25, 2010
Just from watching, the problem seems to be that Wade and LeBron don’t actually have very many offensive skills aside from dribbling the air out of the ball and driving and kicking. They lack fundamentals like a mid-range game, offball movement, post ups, pull up jumpers if the paint is packed, spot up shooting, passing on the go, etc. LeBron’s dribble is incredibly sketchy. I knew LeBron had serious holes in his game that probably held Cleveland back but I didn’t see it being a problem next to Wade and Bosh. But taken together, it’s all rather shocking to see how empty the big 2’s skillsets are.
Also, LeBron’s rebounding is way down despite the fact they get outrebounded by like 10 a game.
marparker
November 25, 2010
Dino,
You would have to compare the entire league’s numbers and compare them to last year to make an argument that the metric cannot explain play as well as the claims. Wade,Bosh,and Lebron’s numbers according to all metrics are down. Does that mean that no metrics are useful.
You can’t cherry pick 3 players who you know are having a bad season to disprove what the numbers say about the other 400 players in the league.
Scipio
November 25, 2010
“They lack fundamentals like a mid-range game, offball movement, post ups, pull up jumpers if the paint is packed, spot up shooting, passing on the go, etc.”
These are skills that NBA fans, the media, and coaches use to rate players, but skills that metrics do not catch because metrics do not care how you score your points. Most of the time I would agree that a point is a point — it does not matter how pretty you looked when you got it. But there is something to be said for versatility, and in the case of the Heat, perhaps we are seeing that versatility really is important? The Boston Celtics’ Big 3 gelled instantly because their skillset was complementary while the Miami Heat Big 3 often find themselves in adversarial roles with even Bosh using possessions in a similar way to Wade and Lebron.
RAM
November 25, 2010
@ dino gunners
The WOW metric is not accountable to critique because dberri has proven it 75% infallible.
Personally, of the big 3 I think only LeBron’s numbers represent an interesting test case. Wade has tended in the past to struggle with injuries and post lower WP48 as a result, so for him to have a wrist injury and a .200 is not uncharacteristic. Bosh put up .250 in a contract year, but .190 is pretty typical of his career.
On the topic of interesting test cases:
Ariza’s production will be interesting to follow this year. If he continues to post around .100 w/ the Hornets, where he has CPaul to initiate offense for him, it would corroborate dberri’s claim that his decline while with the Rockets was not attributable to usage. If he reverts to his pre-Rocket .200 level it would tend to corroborate the usage explanation.
John Giagnorio
November 25, 2010
Who are these people? I read this blog very frequently and recognize about half the names. Why only comment when it’s time to “admit” something? Don’t your “predictions” have more weight if you actually make them beforehand?
Scipio
November 25, 2010
John — you are right. marparker’s predictions carry a lot of weight:
“For those who think some kind of diminishing returns are going to happen these aren’t your typical high wow players who grab loads of rebounds and never shoot. These are guys who are superior in multiple categories.
Bosh was the leading rebounder on the 08′ olympic team. Wade and James were the two leading scorers. They shot 67 and 60 percent respectively. James was second in assists. These guys are not going to get in each other’s way, they are going to exacerbate each other’s best abilities. ”
It seems marparker enjoys analyzing a player’s skillset and making predictions about how well that player will mesh with others… until that player does not mesh with others as predicted. At this point, he stops talking about diminishing returns and skillset and turns to rigidly defending the model.
“Miami is going to win regardless. I am going to scream this until my head explodes
Noone can beat Dwade, Manu, Kobe(in that order) if they have a proper team assembled around them.
Lebron was going to lose to Miami once they signed Bosh unless he joined them. Its clear that Miami knows exactly what its doing.
And the hate will cease. Bosh and Wade are likeable guys and they are going to be having a great time playing with Lebron who is going to make them look ridiculous.”
https://dberri.wordpress.com/2010/07/08/does-a-dynamic-trio-guarantee-a-title/#comment-73516
I am just having a bit of fun. I think if Miller had been healthy and if Haslem had not been injured, this team could flirt with 70 wins. Picking up Dampier will help stop the leaking.
I fully expect all three to regain form, albeit at a slightly diminished level; the analysis that their playing styles are not complementary is legitimate. This model needs to allow for situational analysis (no, not of the marketing variety – you know what I am trying to say!).
Scipio
November 25, 2010
I will eat my hat if this team is hovering around .500 at 40 games into the season. But I have to admit… I sorta want that to happen. The unexpected is exciting. It is fun being wrong sometimes.
“I have the Fab Four with 69 wins assuming 36 mpg and no diminishing returns. Even knocking off 7 wins for diminishing returns has them in the champs grouping with 62 wins produced.”
Italian Stallion
November 25, 2010
Some Dude,
>>IS – the issue for Wade and Lebron isn’t the lack of shots. Lebron is only 2FGA per 36 min less and Wade 3. But look at those percentages! <<
James is down 3.2 per 36
Wade is down 3.2
Bosh is down 3.5
That's a lot of shots.
James is having a rough time from beyond the arc (probably random) and Wade has been playing hurt.
However, I agree that Wade and James are somewhat redundant and that can impact their results in other ways.
I don't see that as a problem with the system though.
It is generally conceded that you can't build a successful team with all Cs, all PG etc.. no matter how productive they are on any measurement. Constructing a successful team requires a wider range of skills.
Since I first came to this forum and began learning about the methodology one of the things I've frequently brought up is the idea of hybrid players. Not every player fits into the neat package of PG, SG, SF, PF and C. Skills overlap and some players don't play anything like the standard for the position. Hybrid players are simply an extension of the same thing.
Even though Wade and James play different positions, they are alike in many ways. So you can say this is a construction issue that needs to be resolved.
Again, I don't see this as a problem with the system.
I think the system measures the productivity of the players well, but it's still up to GM/coaches to build a team with a well balanced set of skills.
Italian Stallion
November 25, 2010
Owen,
I hope you are wrong and the Duhon/Felton scenario.
Felton did show signs of becoming a better player last year. Fingers crossed.
Dino Gunners
November 25, 2010
I have to cherrypick because its currently not possible to export the data in excel. Once it is possible, I don’t mind doing more statistical analysis
some dude
November 25, 2010
IS – well, my numbers were before the update. The shot decreases now are getting bigger. :D
Of course, that is to be expected. I think Miami, as a team, is taking less shots overall, but I don’t want to check the numbers. What’s more concerning is the high turnover rate of Lebron/Wade. Perhaps this is taking shots away from themselves?
I wasn’t critiquing the system, fwiw. I was just saying WP48 is a measurement of a total, but not explaining what someone did to get that WP48.
Wade or Lebron will have to find a different way to get a high WP48 compared to the past is what I’m getting at. What’s disheartening is how Lebron especially stands around doing nothing at all when Wade has the ball. And there’s so little movement in the halfcourt offense, overall. This is where Lebron’s lack of a post game, considering how massive he is, becomes an issue.
Michael
November 26, 2010
The Heat have been hurt. No way does Wins Produced fall under any scrutiny at this point in regards to their record.
Also, practically every outlet was enamored with this team, not just the WoW Network.
KnickfaninNJ
November 26, 2010
All of you arguing about why the Big three in Miami are showing less statistical value than last year are partially right.
It is true that a players WP tends to remain the same on different teams. But it’s also true that players get roles defined for them and then get traded for so that they can do the same role on another team. If you have a power forward who’s used as a true power forward and is traded to another team; it’s probably because the new team needs a player to play a true power forward role. I think this is what usually happens in trades. Then it’s not surprising that their WP is often similar (after accounting for age). So you should normally expect similar individual statistics before and after a trade. But I think that if the player is traded and asked to play a new position or the same position but a different role, then his WP could easily change a lot. This doesn’t happen so often in the NBA so you won’t see it in overall studies of the the NBA.
For example, consider Nate Robinson. At certain points in his career (and in all of his college career) he’s had the role of a scoring guard. He’s good at it (even in the NBA, where he’s incredibly small for this role). But he’s too small to defend opposing shooting guards well so coaches try to make him play point guard. I don’t know his advanced statistics at all, but my guess is his WP as scoring guard has been pretty constant, but his WP as a point guard started off pretty lousy and then improved, and could be significantly different than his WP as a scoring guard.
How does this relate to Miami? If LeBron and Wades roles have to change to accommodate each other then they won’t be doing the same things and therefore they won’t necessarily have the same WPs as before. I think this is happening. It’s true 15 games is a small sample, but remember that in statistics, the size of the sample needed to be sure there is a real difference depends on how big that real difference is. (This is an obviously true and basic property of statistics, but never mentioned in any basketball forum I have read. For example, put me on a basket ball court and you won’t need 15 games to see I am not at NBA level. The difference between me and an NBA level player is just too vast. It shows up right away.) The differences in the WPs for the big three this year to last seem pretty big, so I think 15 games is probably enough to say something is going on, even if one is injured.
On the other hand these are three incredibly talented, very competitive and probably very driven guys. I wouldn’t be surprised if they can do the other stuff they haven’t been doing in the past or will work hard enough at it to start doing it well in an NBA game. Bosh has already gotten better. I think they will find a way to make it work.
Italian Stallion
November 26, 2010
KnickfaninNJ
I think you are making a good point about why players tend to be consistent across seasons and trades.
Gil Meriken
November 27, 2010
KnickfaninNJ
So you’re saying WP can be influenced by a player’s role on the team (not necessarily magnitude, but what skills will be emphasized)?
ballin
November 28, 2010
KnickfaninNJ makes a very good point, I think. It dovetails nicely with a prior statement made by the guy that said something like “it’s generally conceded you need a variety of skills to win, and that you can’t have a team of all C’s or all PG’s win big”.
So while WP is a good predictor of a particular player’s productivity within a given role/position… 1) forcing a player to change their “role” can greatly alter their ability to produce (i.e., you wouldn’t play Chris Paul at center, would you?) and 2) you need players to fill a variety of necessary roles in order to be an effective team (i.e., if you have a team made entirely of PG’s, they won’t win no matter how good their WP’s are).
When you take those two factors into account, frankly, it becomes obvious that WP isn’t a be-all/end-all metric. It completely fails to account for either of these two factors which, when combined, can be properly titled “team synergy”. So while I think WP is a great, great way of comparing players across positions and even in terms of figuring out who the most valuable player on a given team is, it’s simply erroneous to conclude that “team synergy” (as I’ve termed the combination of the 2 previously mentioned factors) doesn’t have a strong impact on a team’s wins. Failure to heed this fact would lead to very, VERY illogical results in making predictions. I believe the Heat’s failure live up to its WP potential has been a result of a lack of team synergy, and cannot be explained away otherwise.
marparker
November 28, 2010
So let me get this straight:
20 games into this season because Wade and Lebron can’t figure out how to stop turning the ball over at career high rates, we have all discovered a new fault in wins produced.
Give me a break.
The Heat have an efficiency differential of 7. They have achieved that differential with a new team and 3 key injuries. Everyone is panicking about their record being 9-8.
If you break that record down a little further there are a couple of points that jump out.
1. The Heat are 4th in efficiency differential. One metric that combined schedule with efficiency differential that has been pretty reliable at predicting future performance ranks this beat up team 2nd in the NBA.
2. The Heat have played a brutal schedule. Of their first 17 games they have played 8 against teams with efficiency differentials consistent with 50 win seasons. That’s 47% of their games. Their record against these teams is 1-7. They have only 15 more games scheduled against these teams. Over their next 33 games(ending at game 50 on feb 4th) they have only 6 of these game scheduled. Its realistic to assume the heat can post a record of 36-14 after 50 games(consistent with a 60 win season).
3.Its realistic to assume that a team that good by efficiency differential will win 90% of its games against none 50 win teams. Its realistic to assume that same team will have a .500 record against 50 win teams. That low projection would leave the Heat with (45+7=52) more victories over the course of the season. (52+9 current wins= 61 wins).
4. The Heat have a 7 win differential with their 2 best players playing poorly. That differential should increase as Wade or Lebron or both began to play closer to the ability they have exhibited in the past. This is a scary team.
Gil Meriken
November 28, 2010
The Heat probably will look pretty dominant for the rest of the regular season as their schedule eases up, but their performance in a later-round playoff series will still be in question, no matter how much of a tear they go on from here out. Shades of Cleveland’s performance from last season. And if the Heat are destroying opponents until the second or third round of the playoffs, and then lose, I expect it will be chalked up to random outcomes/ and or injuries.
On the other hand, if the Heat win the title this season, I will definitely own up to it and eat some tasty humble pie.
marparker
November 28, 2010
Gil,
It won’t be me chalking it up to randomness.Unless they lose game 7 of the NBA championship on the road after leading by 13 points in the 4th quarter.
Gil Meriken
November 28, 2010
@marparker – The situation you described could never happen to a team in the Finals.
reservoirgod
November 29, 2010
Marparker makes excellent points. Miami is a scary team for all of the reasons that he cited. But it’s also a disappointing team because LeBron has been off his game much more than he’s been on it and Bosh got off to a slow start. This team was supposed to be dominant and haven’t looked it outside of the home opener against Orlando.
I think there’s going to be a “come-to-jesus” moment for LeBron & Spoelstra/Wade that will determine whether or not we’re going to see the dominant LeBron (i.e. 0.400+ WP48) this season. Wade has performed as expected when healthy. LeBron has not and that’s really the team’s issue.
todd2
December 1, 2010
Knicks aren’t getting help from their bench. What’s up with Mozgov?
Frank Kiah
December 14, 2010
very interesting point of view.
diehardNFFLbarnone
April 14, 2011
Shouldn’t 2007-08 and 2008-09 read 2008-09 and 2009-10?