Statistical Analysis. Humor. Knicks.

Wednesday, September 17, 2014

On Iman Shumpert’s Current Level of Production

(EDITOR’S NOTE: This is the second installment of a four part examination of what has gone wrong since the Knicks’ 18-5 start—a stretch in which they were beating teams by an average of over seven points a game—and whether or not the Knicks can return [in varying degrees] to their early-season form. You can find part one here.)

Despite the universal love that the fan base has for our 2nd year combo guard, I come to bury Iman, not to praise him. How much Iman Shumpert can and hopefully will improve as he continues to recover from last year’s horrific knee injury and hopefully, to grow as a player, remains an open question (and one for another day). Alas, as of this writing, he is far and away the biggest reason for the Knicks’ struggles.

If you look closely at the other four lineups posted in part one of this series, you may have been surprised by the fact that replacing Smith with Shumpert in our best lineup resulted in a net rating drop of nearly 30 points. Let’s take a closer look at what happens when Shumpert is substituted for anyone else in the grouping of Felton, Kidd, Smith, Anthony and Chandler (again looking at games from January 17th-March 3rd). In this chart, for the sake of clarity,  defensive rating is represented inversely so as to make it comparable to ORtg. In other words, a higher DRtg change in this chart indicates that the defense improved.

Shumpert is wrecking every lineup when he’s combined with our best players. Even if we limit our examination to recent games—let’s say March 3rd and beyond—Shumpert is still averaging a +/- per 48 of -15.4 and a net rating of -14.3. The next worst rotation player who has been healthy these last three weeks? That would be Jason Kidd, coming in with a net rating of 1.1 and a +/- of -1.8 per 48. The interesting thing is, Shumpert’s shooting was way up during that span: He shot 43.8% from the field and 45.5% from distance. Still, he remained the vinegar to this team’s oil.

If we go back to my automotive analogy, cars only work with the right parts. If the Knicks are a car—a team that runs well not because of the quality of its parts necessarily but more because of how those parts work together—then perhaps Shumpert is just the wrong part. He’s the truck nuts on a minivan or the loJack strapped to a 1988 Geo Metro. In other words, maybe Shumpert is being used incorrectly or improperly, or maybe there just isn’t a place for him in the style of the Knicks need to play in order to succeed.

So what exactly is happening when Shumpert enters the fray? Take a look at the substitution pattern that is kindest to Shumpert, the one where he replaces Kidd and joins our de facto starters on the floor. Remember, these units include Felton, Smith, Anthony and Chandler. The only difference is that in one, Kidd is the fifth man, while in the other, it’s Shumpert.

*Opponents’ points in the paint per 48 minutes

**Team Average (as well as other stats) are drawn only from that same span of games: 1/17-3/3.

The two data points that jump out, at least on the offensive end, are FTA rate and the 3FGM %AST; both indicate how critical wings that, once they receive a pass, are able to immediately find the flaw in a defense and/or connect with a teammate who can exploit it are to New York’s offensive agenda. Kidd isn’t drawing more fouls than Shumpert nor is he making more open threes. In fact, his shooting during this period was atrocious—23% from the three point line and 32% overall. What he is allowing the Knicks to do is use the space that defenses give when they double Anthony or collapse on a pick and roll, resulting in rhythm threes and advantageous drives from Anthony and Smith.

In short, Shumpert seems to be making bad decisions with the basketball and this year’s Knick model, built to play a more cerebral, half court style, needs guards that make great ones.

Right now, Shumpert just doesn’t have that ability (not many players do), but most talented guards who don’t have great vision (like J.R. Smith) make up for it with the ability to create plays for themselves. Here, Shumpert struggles just as badly. Once he steps inside the three point line, his offense collapses. From midrange, he shoots 33%; he’s at 15% in the paint, and a shockingly bad 36.7% in the restricted area. Those numbers have improved recently, but rather than thanks to improving health, I have another theory as to why—a theory that will be easier to understand after I make the following uncomfortable comparison.

There are some striking similarities between Shump and a post-Melo trade Landry Fields before the latter developed a bit of an off the dribble game/excess tinkering totally discombobulated his jump shot.  At a USG rate of 16.6%, Shumpert outpaces Fields’ rookie rate of 13.5% but he is mostly wasting those extra possessions and would be far more effective at this juncture of his career if he restricted his looks to the open spot ups that once were Fields’ bread and butter. Take a look:

(The Fields numbers are before/after instead of on/off because the absence of Felton after the trade would likely confound Fields’ numbers with Anthony on the bench.)

FGM %UAST %PTS off TOs TS% @Rim FG% AST% net rating
Shumpert w/o Anthony 27.3% 25.8% 51.6% 46.2% 12% -3.2
Shumpert w/ Anthony 35.9% 15.0% 44.4% 35.1% 20.6% -9.9
10/11 Fields pre-Anthony 28.8% 21.2% 61.5% 70.1% 8.5% 3.1
10/11 Fields post-Anthony 30.8% 17.7% 56.4% 63.6% 9% -1.4

Fields was often criticized after the trade for his inability to do anything other than shoot open threes. The cuts off ball movement and timely rebounding dropped dramatically as the offense pivoted to accommodate Melo’s skill set. While Shumpert appears to have a better handle, better passing skills than that iteration of Landry and (when healthy) is the superior athlete, the difference between the two is more one of will than skill. Shumpert is more willing to attempt to make plays/force the issue when necessary as illustrated by the jump he experiences in AST% with Anthony. But, when he puts his head down and bulls rim-ward on such ill-advised playmaking forays, Shumpert struggles a lot. By comparison, when given the opportunity, Fieds would usually kick the ball back to another player. Unfortunately, these clock-killing decisions likely had nearly as bad an impact on the Knicks’ offense as Shumpert’s low percentage drives.

The improvements for each at the rim without Anthony are easy to explain: these are more often than not assisted shots, shots where each catches the ball with very little to do to get a quality shot, or they are points off the break, as illustrated by the points off turnover increases.

When a player can neither see the floor well nor create off the dribble against set defenses, that player tends to struggle in an Anthony-led offense, one based on intelligence, patience, and execution more than the single-attack oriented, perimeter and pick and roll oriented offenses that the Knicks tilt towards while Anthony rests. These latter offenses more often create opportunities off the first pass, allowing players like Fields and Shumpert to simply catch passes and then shoot.

If you aren’t convinced (or are just looking for a reason to get more depressed) consider this fact: The Knicks’ current starting lineup shoots only 25% from three. Substitute in Smith, (again, just looking at games after Shumpert returned) and the Knicks’ three point percentage jumps to 42% despite the fact that, in that some period, none of the players in that lineup are shooting better than 36%. It’s relatively simple. Smith’s presence creates open shots and Shumpert’s doesn’t.

Without Señor Carmelo Isolacíon Anthony on the floor, the Knicks offense becomes less systematic (less like a machine, such as a car), and more free flowing, and that sort of offense is one Shumpert can much more ably contribute to.

Shumpert is not a point guard. We became all-too aware of this fun fact early last season. This year, the ridiculous -90.1 plunge to the team’s net rating when he replaces Felton only reinforces that notion. That said, he can be a valuable player if he’s used correctly. The question is, is it worth it to the Knicks to cater to the needs of one player (Shumpert) or would they be sacrificing more than they gained? We’ll save that question for a future installment, however.

Then, of course, there’s the defense—the thing we all love about Shumpert. Well, it’s not showing up in the stats. As our graph indicates, there’s only one lineup in which Shumpert helps the defense and that lineup (the four guard lineup with Shumpert replacing Anthony) and this unit has only played six minutes together all season. It seems that while Shumpert may still be a good man defender, his team defense still isn’t up to snuff, especially when he’s toiling among fellow Knickerbockers that lack athleticism/speed and, as such, is deeply reliant on smart rotations to maintain defensive integrity. Look at the Bockers’ opponents’ three point percentage when Shump’s on the floor versus when Kidd is present. Yes, the Shumpert lineup is better than the overall Knick average, but that number is dramatically inflated by Stoudemire’s presence (or lack thereof). That 12% drop when Shumpert plays makes it tough to argue that he is rotating and closing out with any degree of success.

This clearly isn’t Shumpert’s year.  If the Knicks have any hope of making it their year (which one could easily argue they shouldn’t), they need to dramatically reevaluate how they’re using their sole young player.

Up next: Amar’e Stoudemire…

10 comments on “On Iman Shumpert’s Current Level of Production

  1. max fisher-cohen Post author

    Thanks ruru.

    There’s one kind of confusing part here: “The Knicks’ current starting lineup shoots only 25% from three,” is incorrect. What I meant to write was the lineup of Kidd, Felton, Shumpert, Melo and Chandler.

  2. nicos

    Nice breakdown though it’s tough to get any kind of read on Shump this year- I agree that overall he’s been a net negative and his minutes have been part of the reason for the Knicks recent mediocrity. That said, I think it’s safe to say the minutes he’s getting are intended to get him in a position where he can be a contributor when he’s closer to 100% rather than for what he’s giving now (it’s not like he’s closing a lot of games).
    Also with those post March 3rd numbers Melo was either playing hurt or out entirely which really threw the whole team out of whack. Two, when he plays without Felton that means he’s playing with a point guard (either Kidd or Prigs, take your pick) who can’t (Kidd) or won’t (Prigs) penetrate and is unwilling to shoot the ball- both of those guys are heavily reliant on JR to break down defenses- something it’s a little much to expect from Shumpert at the moment. The Knicks already have too many low usage guys (Kidd, Prigs, Novak, Chandler) that they can’t afford another. And he’s been playing a lot of three defensively as opposed to on the ball where he’s at his best. On plus side, his three point shooting looks like it might be sustainable (hitting both from the corner and the wing) and he’s shown the ability to see the whole floor when driving which is a huge plus for a two. Unfortunately, while you can live with the poor midrange shooting as long as he keeps those shots to a minimum, his complete inability to score in the paint can’t continue if he wants to be anything other than a role player off of the bench.

  3. BigBlueAL

    It does seem like Shump has been playing alot better lately and the shooting numbers Max mentioned seem to bear that out but pretty surprising that his +/- numbers still are horrible even with his improved shooting lately.

    I do think though it is impossible to judge him on this season since he is returning from his injury, next season is the make or break season for Shump and his Knicks career I assume. Also will be interesting to see what type of player the Knicks draft this season, if they draft a SG/SF type similar to Shump will that mean they have soured on him and will be looking to possibly trade him??

  4. jon abbey

    Noah and Marc Gasol both out hurt now, the regular season is at least 12 games too long, probably more like 20.

  5. danvt

    Hey Max,
    I really understand this season much better because of this article and this series, thanks. So, it seems that we have a best 5 and a huge drop off after that, with the main culprit being Shumpert. The lack of decision making on offense and knowing where to be + lack of lateral movement have made him a liability despite flashes of quality play and decent shooting percentages.

    I just wonder if you’re drawing the right conclusion when you say this “isn’t his season”. I mean, he hasn’t played that much. Less than 20 minutes usually, right? Also, he hasn’t been there all season and the poor kid has never had a training camp in the NBA. Tonight against the raps he was a plus 16. Don’t you imagine he can improve from here?

    Amar’e, I have always thought, drags the team play down despite doing some nice things out on the court. I’ll be interested to see what the #s say about him. I didn’t think of Shumpert that way.

  6. max fisher-cohen Post author

    @danvt

    The problem is that 20 minutes is almost half the game. If you lose the minutes with Shumpert by 10 points, that leaves you only 28 minutes to make up that difference.

    But yeah, I’m sorry if the article came off as a condemnation of Shumpert for all eternity and that he should never play again. I don’t believe that. I was worried that it would come off badly, like I was picking on the kid who just got out of the hospital for dropping a pass in our recess football game, but I guess I see it as the team has made numerous bold moves over the past few years that indicate that they are in WIN NOW AT ANY COST mode. I can accept that as long as the team is consistent.

    I see Shumpert similar to how, now that emotional sheen has faded, I see the Lin choice. Like Lin, Shumpert has the potential to be a valuable part of this team, but saying a player has potential is just a nice way of saying that he is a risk. In retrospect, keeping Lin probably would have been smart, but no one really knew how Lin would play this season, and we still don’t know how things would have turned out had he stayed in New York.

    Especially in light of the lightning in a bottle the Knicks seem to have captured with their best lineup, like Lin, Shumpert may be a risk they do not need to take. Sure, let him play out the season at 20-30 mins a game, but come playoff time and the summer, if his chemistry with the team and overall play hasn’t improved dramatically, you have to consider making some changes.

    I’ll add that this is not a team building philosophy that I agree with. I don’t like the fact that this team has an expiration date and very little chance at a title, let alone a competitive ECF, before that date. I don’t like that after that expiration date, we have what will likely be a very arduous rebuilding process to look forward to. But at this point, I think we’re also too deep into the process to turn back.

  7. KnickfaninNJ

    Hi Max,

    I am really enjoying your analysis in this series of four posts. For me, they bring up two questions so far.

    Since Shumpert started the season hurt and not playing, does the Knicks’ win percentage correlate with his presence or absence in the line up? Your analysis suggests it will correlate.

    To my eye, the Knicks have done less passing and more iso ball as the season progressed. It’s almost as if they had a memory of d’Antoni as a coach and what he taught them, and that memory is fading as time progresses. Is there any statistic that can check this hypothesis? Passes per shot attempt would be nice, but I have no idea if anyone tracks that.

    Thanks again for the nice posts.

  8. max fisher-cohen Post author

    @KnickfanininNJ

    http://stats.nba.com/teamStats.html?TeamID=1610612752&MeasureType=Advanced&PerMode=Totals

    Assist ratio is the # of assists averaged per 100 possessions. If you scroll down a bit in that link, you can see the month by month breakdown. You’re right that there has been some decline in that number across the season. However, the Knicks have recovered some of their lost assisting this month.

    If someone with a Synergy account sees this, they might be able to give you exact numbers in terms of how often the Knicks isolate. I don’t think it’s just isolation though that is the problem. The drop in assists is probably as much a result in the drop in shooting from three (February was their worst 3pt% month by far), but that can be a chicken and egg sort of question — maybe the shooting was worse BECAUSE of the poorer passing.

Comments are closed.