Fanservice: Followup notes on Melo, Rookies

“I, at any rate, am convinced that He does not throw dice.” -Albert Einstein talking about God

I’ve been online for only a few hours this weekend. The reason for this is that I was throwing a party at my house on Sunday.

I spent my weekend doing prep work. I scrubbed floors, I cleaned toilets and I even got down on my knees and  finished both with a razor (hardwater is a pain to get out). Our house gleamed by sunday (and a great time was had by all thanks for asking).

The funny part is that I logged off on Friday thinking “I’ll log on for a few hours this weekend and put up a quick post on a particular player, nothing major”. Life is funny sometimes (and God’s a funny guy).

Or say for example a Melo stinkbomb against the Spurs

I ended up taking on America’s Statistician in Chief, Nate Silver of FiveThirtyEight (and to my credit even though I wrote the piece at 2AM on a saturday I didn’t bring up anyone’s mother).

I know it's a repeat but some things bear repeating

Naturally I expected a bit of feedback and a few requests. Background first:

  • First off there’s this piece: The Nets and Knicks May Be Better Off Without ‘Melo.In it Jared Diamond of the Wall street Journal (with an able assist from our own Prof. Berri) debunks some of the hype surrounding an old favorite, Carmelo Anthony . The relevant points being that Melo to New Jersey (Derrick Favors and Devin Harris to Denver, The Nets get Chauncey Billups and Rip on top of Melo) would be a train wreck for the Nets. Melo to the New York (assuming it looks something like this: with Fields, Chandler, Curry and picks from a Randolph trade as the featured assets for the Nuggets) would be a Isiah-esque for the Knicks   as they would win roughly 29 games over a full year (and I bust a gut laughing at the ridiculous New York Knick management ).
Overrated, Overpayed player goes to the Knicks, stop me if you've heard this before

Here’s some food for thought:

Trading Fields might be a bad idea.

And as I finish typing that paragraph the ESPN ticker flashes with the news that the Nets are talking to Melo about an extension.


Nevermind, we’ll carry on.

I choose A

The second relevant piece was the post from the aforementioned Mr. Silver:

Why Carmelo Anthony Is the Ultimate Team Player (and What ‘Advanced’ Stats Miss About Him)

His claim is simple. Advanced stats such as Wins Produced (see here for the Basics) miss the true value of Carmelo Anthony as a player. To quote:

In taking all of those shots, however, Anthony has also done something else: he’s made his teammates much more efficient offensive players.

And because he makes his teammates more efficient by being inefficient himself, Melo is more valuable than he appears.

Now yesterday I pulled the data for every single game for the nuggets for the past two years (see here ). I then divided all the games into games where Carmelo played and games where he did not. Finally I worked out Points generated per shot for each scenario. The results are here:

You can see that Mr. Silver is right. Melo’s teammates are more efficient with him on the floor. But only slightly (.9% this year and .1% last year) and Melo ends up shooting that advantage away. At this point we go to the to the request.

A few readers asked me to redo the table using True Shooting %. Asked and answered:

The numbers tighten up with the shift. In 2009-2010, Melo’s teammates actually shot better in games he did not play in (but scored less points per shot). This year they’ve been slightly worse. As a team however they always shoot better (whoever you want to measure it) without the innefficient high volume shooter and there should be no doubt that that’s what Melo is.

For Melo, the Nuggets and apparently the Nets an old story is being repeated, that of the overrated scorer getting traded. The team getting him should remember the history of that sort of trade.

This is really not ending well


  1. some dude
    January 17

    I would prefer not to see this year’s numbers but rather 2 years ago. This season the team is being weird because of all the trade stuff.

    The other thing I’d do is move away from points per shot and to points per possession. One thing I noticed by looking at was fewer TOs when Melo was in the game.

    There’s another part of this analysis that doesn’t tell the whole story. Some of those “without Melo” minutes are garbage minutes. I’ve always believed in the garbage minutes don’t tell us much theory other than who stinks (if you can’t perform against scrubs, you stink). And there’s no way for you to really separate the minutes where Melo was hurt/resting and where it was garbage time, so the “true” points per shot without Melo is hard to ascertain.

    Anyway, the Knicks main issue is the lack of an interior defender/rebounder, not another scorer (inefficient or otherwise).

    Also, damn you Arturo for reminding me about that movie. Only the saddest movie of all time. Sure, it was great, but just so so sad. Why make me so sad?

    • January 17

      08-09? Fine. Same for points per possession. I won’t be doing another Melo post for a while (unless events dictate otherwise).

      The garbage time argument doesn’t really apply here actually. I’m looking at whole games with or without Melo. So Denver with and without Melo.

      I totally agree on the Knicks actually.

      Sad is what I was going for.

      • some dude
        January 17

        for the record, i think the results will be pretty similar. I just say it in the name of more accuracy. 😀

        For some reason I misread the games with and without part. Thought you were splitting it up differently.

        Of course, it would matter which players played those games too.

        A better way would be comparable lineups:

        Graham in place of Anthony with the other starters is a big drop in efficiency. At the same time it appears JR Smith in his place helps the offense.

        Just as an example of lineup approaching.

  2. EvanZ
    January 17

    With or without you, I can’t live…

    (sorry, couldn’t resist.)

  3. some dude
    January 17

    Arturo, I see that you said you want to do this for Lebron, too.

    The problem I see is that Lebron only missed 14 games the prior 3 seasons. One of which the team didn’t play any of the starters (final home game of ’09). I can’t recall if they threw any of last season’s final games. But if we go by the 2 season analysis you’ve been doing, we’re looking at 6 total games. How can you do this with such a small sample?

    Now, maybe you want to look at Cleveland this year, but that’s not going to work because the lineups have changed drastically. Moon has replaced Lebron with the Hickson-Varajeo for 111 minutes. Graham is terrible and makes that one skewed. Plus no Shaq or West, either. And then throw in that Cleveland is playing with no heart at all. Mo Williams is worse right now than he was on a bad Milwaukee team, and WP would predict he should be better without Lebron (from a WP standpoint). Then throw in a scheme/coaching change too

    So, I really don’t know how you’re going to pull off this analysis. he didn’t miss enough time with the team and the team out there now isn’t close to the same. The only way I can think of possibly doing it is looking at when another player comes in, say Shaq or Jamison or Williams, but also noting other players could be the cause, too (but doubtful).

    It’s actually why I have a hard time with the arguments both Silver and you have kind of laid out. There’s so many changing variables that it’s hard to isolate what is causing what. And with smaller samples, you have to then adjust for strength of schedule. And there are outliers (Manny harris just had a +/- of -57, for example). Etc etc.

    I am really unconvinced that any advanced stat approach will figure out whether “x player makes y teammate better,” because of the daunting task of controlling for the changing variable (we cannot isolate the OVB). The best I’ve seen so far is looking at 5 man line-ups and seeing their differential splits. But that still doesn’t do a good job.

    • Italian Stallion
      January 18

      I think there are always a lot of variables that are hard to control for, but if virtually all the high usage scorers that are generally believed to improve their teammates (Melo, Lebron Kobe, Amare for the Knicks etc…) look positive on this kind of study, it’s hard to conclude that they aren’t. From there it is mostly a debate about how positive.

    • some dude
      January 18

      i think this method is a start, but again you have issues of garbage time, strength of schedule, back to back games, etc that can skew results since often the samples are small.

      The efficiency splits is basically the best we seem to have at this point, as I mentioned earlier, but it doesn’t help prove anything (perhaps Kobe’s backups hurt his teammates?).

      I think with guys like Kobe, Lebron, Wade, CP3, and others the answer is obvious to our eyes. We can try to use stats to argue how much one helps over the other, but they all obviously improve the play of most teammates (not everyone, of course. sometimes styles of play clash).

      With a guy like Melo, it’s much harder to figure out. Do his drawn doubles do anything? The notion that his assists should be higher is a farce, most doubles create a hockey assist. But does this happen with Melo (does he hit the right guy?)? Does Melo cause more players to be in foul trouble giving teammates more foul shots from non-shooting fouls? Do Melo’s shots tend to be putback rebounds?

      These are questions guys need to analyze with stats AND watch the games (I admit I don’t watch enough of Denver to answer this). I do know Melo is a pretty good interior passer, not a common skill btw, and he is a streaky scorer and gets teams in foul trouble. But he’s also a mediocre defender, despite being a good rebounder.

      Another thing is Melo has mostly played with streaky jump shooters (Miller, Carter, Billups, Iverson) rather than more consistent ones. With a different cast of teammates, things could be different.

      tldr; I don’t think Melo’s impact is easily ascertained by the numbers in this case because it’s hard to control the variables and it’s not “obvious to the eye” in his case. I’d like to see him on a completely different team before casting personal judgment. I will say this. A straight swap of Landry Fields for Melo will not produce more wins for Denver.

    • Italian Stallion
      January 18

      OK, forgive the ignorance, but I have to ask this question. This is from the Pelton article on Kobe.

      “Bryant’s so-so efficiency makes it tempting to label him a ballhog. After all, his True Shooting Percentage of 54.1% is lower than fellow Lakers starters Lamar Odom (54.9%), Smush Parker (55.3%) and Chris Mihm (56.4%) as well as occasional starter Brian Cook (57.6%). It’s also scarcely above the league average of 53.2%.
      While I’m a big believer in the importance of True Shooting Percentage, it doesn’t tell a complete story. For one, it ignores the role of turnovers. Bryant’s turnover rate is in the NBA’s top 15. If we look at points scored per possessions used (FGA + (.44*FTA) + TO), Bryant — who averages 98.8 points per 100 possessions — shoots past every Lakers regular save Cook (103.8) in terms of efficiency. Still, this alone is not enough to explain Bryant’s dramatic impact on the Lakers offense. ”

      While I totally understand that high usage players tend to have more turnovers that tend to be taken away from their lower usage teammates, I don’t understand this paragraph. A player doesn’t get penalized in TS% by his turnovers. If you double TS% isn’t that telling you points per possession?

      • January 18

        No. TS% is Points/ (2*(FGA+FT*.44), need to add in turnovers to get to points per possesion.

        • Italian Stallion
          January 18

          OK. I understand that, but I don’t understand Kobe’s point per possessions seemingly improved when the turnovers added back because those are lost possessions.

  4. ilikeflowers
    January 17

    Speaking of rookies, if Wall can return to his pre-injury form we may be looking at another Rondo like PODD (possession-oriented distributor defender) in DC. Also, R. Lewis + return to SF + crappy teammates = above average productivity again.

  5. Italian Stallion
    January 18

    I think the turnover issue is one that doesn’t get enough discussion. Most high usage scorers (highly efficient or not) have the ball in their hands a lot and therefore turn the ball over more often than low usage scorers.

    The less efficient among them get very little credit for their extra scoring, but bare the full impact of the extra turnovers they commit when at least some portion of those turnovers are almost definitely coming out the teammates total.

  6. Mike
    January 18

    Grave of the Fireflies. Epic sadness.

    Nets fans will cry like I did at the end of that movie, if this Melo deal goes through.

  7. Man of Steele
    January 19

    Arvydas Sabonis was just a monster. As a 31 yr. old “rookie” he threw up a .389 WP48. For my money, the best center of our generation may have missed the NBA during his prime. Seriously, who puts up those kind of numbers at 31??

    Also, it’s interesting that one of the best comparisons for Griffin (productivity-wise) is Charles Barkley. They actually have similar skill sets (as young players) if you think about it: strength and relentlessness under the basket, extremely quick for their position (remember, young Barkley), explosive hops, etc.

Leave a Reply

Your email address will not be published. Required fields are marked *