Statistical Analysis. Humor. Knicks.

Tuesday, October 21, 2014

Carmelo Through the Years

Historically, statistics suggest Carmelo Anthony has not always impeded ball movement or prevented offensive flow. Over the course of his eight-year career, Anthony has primarily played with three point guards – Andre Miller, Allen Iverson, and Chauncey Billups. When Melo is paired with a point guard able to control the offense, the numbers prove he is not simply a stop-and-hold isolation player.

In Carmelo’s Rookie year (03-04), with Miller running the point, the Nuggets finished 9th in the league in assists per game. ‘Dre averaged 8.5 assists per 48 minutes – not a stellar number, but a solid one.  60% of the Nuggets FG’s were assisted, a number that held true regardless of whether or not Anthony was on the floor. In total, 55% of his baskets were assisted. The offense wasn’t just dumping it off and watching, but rather finding him in easy-to-score situations. Additionally, Anthony averaged 4asts/48 – an excellent number for a “shoot-first” rookie.

The Nuggets only improved in Anthony’s next two years, finishing second and third in the league in assists, respectively, with Miller averaging 10asts/48min. Even more impressive, 63% of Carmelo’s makes were off assists, including 64% of his jumpers. Clearly, he was not solely settling for contested shots. It helped tremendously, though, to have a pass-first guard orchestrating the offense, allowing Anthony to play to his strengths: getting position and scoring.

When “The Answer” (a true shoot-first guard) took over the reins on offense in 2006, Anthony only improved his contribution, averaging 5asts/48min and when he was on the floor. Meanwhile, 62% of the Nuggets FG’s were assisted – while off, this number dropped to 57%. Anthony, it should be noted, had a stellar offensive season, averaging 29ppg.

With Chauncey Billups at the helm in 2008-2010, the Nuggets fell to 18th in the league in assists. Billups averaged under 8 assists per 48 minutes (the lowest of any PG Anthony had played with, including Iverson), and thus Melo’s isolation habits began to show. 64% of his FGA’s were jumpers, as opposed to his usual number somewhere in the mid 50s.  Even more shocking, his scoring was only assisted 42% of the time, a far-cry from the 60% he was used to.

This year, playing without any semblance of a point guard thus far, Carmelo’s stats paint an ugly picture: Only 30% of his FG’s have been assisted, and his 42 eFG% similarly marks a career low. A whopping 77% of his shots have been jumpers – a 20% increase over years past — while a mere 1% have been dunks. Currently, the Knicks are 20th in the league in assists. Needless to say, however, it seems as though help has finally arrived.

In terms of guards, Anthony will soon be playing with arguably the best pure passer he’s ever called a teammate in Jeremy Lin, with the second year Harvard man averaging 14asts/48. As such, the perennial All-Star won’t be forced into point-guard duties – as he was for much of this season’s first stretch – and the ball movement won’t start with him. Now, Melo’s main concern will be moving without the ball, running the pick and roll, finding open spaces, and finishing at the rim.

Jeremy Lin has proven he will reward hustle and persistence.  Statistics prove Carmelo performed well in an up-tempo, fluid offense in Denver. Since his pairing with Billups, the isolations have drastically increased. Now, with a smart, young point guard, it’s time to prove everyone wrong.

24 comments on “Carmelo Through the Years

  1. flossy

    Are we really prepared to say with certainty that Jeremy Lin is a better setup man than Andre Miller? I mean, even if he’s just Andre Miller I’ll be happy, so whatever, but still.

  2. Zach Horst Post author

    Well, not with certainty. But, with confidence I think we can say that. Still has a lot to prove, I know. He’s well on his way, though.

  3. Grymm

    For some reason, I never really liked Chauncey as a point guard. He was a very good basketball player and leader, but not much of a creator. I wonder how much of Carmelo’s ball-stopperness might be derived from playing with a pg who wasn’t going to create him any shots.

  4. max fisher-cohen

    Good post, zach. As I’ve written here before, I am cautiously optimistic. Pierce had a similar knock on his game before he found three good buddies worth passing to (or receiving a pass from). The bottom line though is that the Billups thing is an excuse. Just like the “Douglas and Shumpert were atrocious” argument is an excuse. They may be good excuses, but if you prefer hard evidence as opposed conjecture (even if it’s well-reasoned conjecture), you should IMHO still hold some degree of skepticism.

    As far as the assists go, high usage isolation players who score by putting the ball on floor (as opposed to post ups) almost always have decent assist numbers. If you just look at guards and guard/forwards in the last 10 years who have had a usage rate over 26% and played more than 1800 minutes on the season, sorting by assists/36, you have a bunch of catch and shoot guys first — Kevin Martin, Michael Redd, Allan Houston, Rip Hamilton. The first ISO player who comes up is Vince Carter, who averaged 3.5 assists/36 in 04/05. Melo, for his career, averages 3.1 assists/36…

    This IMO is because when you almost always have the ball, and it’s always your responsibility to get your own shot, sometimes you’ll draw extra defenders and an easy pass for a score will present itself. This doesn’t mean you are trying to setup your teammates. You have the ball often enough, and you are a good enough ISO scorer, easy passes for scores are going to arise. This doesn’t make you a good or even a willing passer.

  5. massive

    I’m not sure if this has been stated here yet, but JR Smith was offered a 2nd year player option. We could have him for two years? Sweet.

  6. Brian Cronin

    I’m not sure if this has been stated here yet, but JR Smith was offered a 2nd year player option. We could have him for two years? Sweet.

    It’s there as an insurance policy in case he gets hurt or something like that. That’s the only way he’d pick up the second year option.

  7. Gideon Zaga

    I think you should have tagged Hoolahoop in the article lol. Only lazy people didn’t know this since they don’t do their Homework. Plus Melo’s secret lover Ruruland, who I am convinced is lala, has constantly pointed this out time and time again. Like they say Numbers don’t lie. Hope you emailed this to the Cock as well.

  8. hoolahoop

    “When Melo is paired with a point guard able to control the offense, the numbers prove he is not simply a stop-and-hold isolation player.”

    Yeah, just review the books. Don’t bother to watch the games.
    After that sentence I wouldn’t read another word.

  9. Zach Horst Post author

    I didn’t say he NEVER runs isolation, but clearly, a high percentage of his jumpers were assisted. I mean, you can’t debate the fact that isolation plays aren’t assisted.

  10. hoolahoop

    Another thing about stats. They are portended to be great objective measurements, but they are subjectively recorded by humans, subject to misjudgement and error – to a high degree.

  11. villainx

    Another thing about stats. They are portended to be great objective measurements, but they are subjectively recorded by humans, subject to misjudgement and error – to a high degree.

    That’s true, and if one doesn’t review the books and solely watches the games, does that lead to higher misjudgment and error, or lower?

  12. hoolahoop

    villainx:
    Another thing about stats. They are portended to be great objective measurements, but they are subjectively recorded by humans, subject to misjudgement and error – to a high degree.

    That’s true, and if one doesn’t review the books and solely watches the games, does that lead to higher misjudgment and error, or lower?

    I always like to look at the stats after/during the game. Surely, it clarifies the picture. But it isn’t the picture.
    And that’s what some people do here. They look at stats, then draw the picture (that they want to see).

  13. daJudge

    Nice piece Zach. I loved Bernard King almost as much as Clyde. I personally think we will see Melo really thrive with this team as currently configured. he brings something few in this league possess. If his skill are bridled, we are in for a real treat. Thanks for the stats to at least partially back up my growing optimism. The other thing involves a comparison between Lin’s turnovers and Melo’s FG%. I think the TO’s will go down and the FG% will go up pretty much for the same reason. Neither Lin nor Melo will be required or be inclined to force the issue with the other options/depth. If you are willing to cut Lin slack on the turnovers, to be reasonably consistent, I think you need to cut Melo some slack on the Iso’s. Also, I don’t know what went down, but I feel terrible for TD. One quick story—My wife canceled her beloved Saratoga Racino trip tonight to watch Lin, et al tonight. Freaking awesome–and I probably saved $300. Go Knicks.

  14. Doug

    hoolahoop:
    Another thing about stats. They are portended to be great objective measurements, but they are subjectively recorded by humans, subject to misjudgement and error – to a high degree.

    And the observations of the human eye are not?

  15. ruruland

    Grymm:
    For some reason, I never really liked Chauncey as a point guard.He was a very good basketball player and leader, but not much of a creator.I wonder how much of Carmelo’s ball-stopperness might be derived from playing with a pg who wasn’t going to create him any shots.

    he’s a combo guard, really. Drives to score, below average PnR.

  16. Juany8

    Hoolahoop Melo has his issues but at this point I’m pretty sure you’ve openly admitted you just don’t like Melo. Clearly he stops the ball and takes bad shots sometimes. So do Lebron, Wade, Kobe, and Rose, it’s called being a primary option on offense, while not having the benefit of being a one man fast break to increase both efficiency and total points. Carmelo has issues, some dealing with effort and focus, some dealing with the fact that he’s just not as fast as the aforementioned players, yet people still expect him to be Lebron and don’t seem to be ok with the possibility that Melo should probably be playing more like Dirk, who hasn’t been exactly praised for great defense and point guard vision. The Knicks don’t have a top 5 player and are unlikely to acquire one in the near future. Doesn’t mean we can’t challenge for a title with 3 top 25 players and some decent role players.

  17. Juany8

    I actually fully agree with your point about statistics by the way, most people who use them (THCJ being such an obvious example it almost feels like a parody) don’t seem to get that the statistics don’t hold any intrinsic meaning of their own. Just because you took an average of how many rebounds a guy has per game does not mean you have an absolute measure of how good he is at rebounding, unless you’ve never heard of boxing out. Yet all the time, you see people equating things like TS% to a player’s offensive quality, assuming that things like Usage rate don’t say anything about a player’s ability to play. That’s a pretty hefty, subjective assumption, not the empirical proof many act like they have. For statisticians, attempting to irrationally define player value by a single formula is more appealing than truly taking the time to watch video and analyze the game like every single scout has to

  18. ruruland

    hoolahoop:
    Another thing about stats. They are portended to be great objective measurements, but they are subjectively recorded by humans, subject to misjudgement and error – to a high degree.

    Ahahaha.

    “hoola hoop” trying her best to come off as learned. More of an an echo than an original thought.

    Poor time to use the argument. You’re suggesting that the 200 plus games Melo played with a passing point guard and was assisted on over 60 percent of his makes, compared with the over 400 games played with scoring points guards where he was assisted in the 40 percent range………you’re suggesting that’s all an accounting error?

    I like your sentiment about the flaws of stat-counting….,

  19. ruruland

    hoolahoop:
    “When Melo is paired with a point guard able to control the offense, the numbers prove he is not simply a stop-and-hold isolation player.”

    Yeah, just review the books. Don’t bother to watch the games.
    After that sentence I wouldn’t read another word.

    What are you trying to convey here? That you watched the Miller-Melo Nuggets and 60 percent of his shots weren’t assisted? Are you suggesting that he was isoing anywhere near as much then, just that the stats are wrong?

  20. hoolahoop

    Juany8:
    I actually fully agree with your point about statistics by the way, most people who use them (THCJ being such an obvious example it almost feels like a parody) don’t seem to get that the statistics don’t hold any intrinsic meaning of their own. Just because you took an average of how many rebounds a guy has per game does not mean you have an absolute measure of how good he is at rebounding, unless you’ve never heard of boxing out. Yet all the time, you see people equating things like TS% to a player’s offensive quality, assuming that things like Usage rate don’t say anything about a player’s ability to play. That’s a pretty hefty, subjective assumption, not the empirical proof many act like they have. For statisticians, attempting to irrationally define player value by a single formula is more appealing than truly taking the time to watch video and analyze the game like every single scout has to

    That’s right

  21. Duff Soviet Union

    When quoting assist numbers, I think it’s fair to point out that Denver’s hometown scorer has a rep for massively inflating the assist numbers of the hometown players. I haven’t checked the numbers for this year, but in recent history the difference between the percentage of Denver’s baskets paired with an assist at home versus on the road has been the largest difference in the league. Almost every team’s scorer does that to some degree, but Denver’s guy is the worst. In fact the only team to have a higher assisted % of made baskets on the road than at home is the Suns. Makes you wonder how many assists Steve Nash would have if he played half his games in Denver.

Comments are closed.