Data Mining: A statistical breakdown of why the Nuggets have gotten worse

The Denver Nuggets are now 11 games into the season, and with just five wins under their belt they will be trying to get back up to .500 tomorrow against Dallas.

This is a small sample of games – certainly not significant enough to project what will happen in the latter half of the season, especially considering the fact that the team is adjusting to a new system and missing two of its best players in Danilo Gallinari and JaVale McGee. It is enough, however, to see a picture emerging of how things have gone for Denver so far, and what they might be able to do to improve.

Here we will take a statistical snapshot of where the Nuggets are at now, as well as what they’re doing differently (and – spoiler alert – mostly worse) than last season. This post will focus on Denver’s team offense, so look for more analysis on team defense and individual player performance in future Data Mining installments.

 

Overview

We begin with an overview of the Nuggets’ performance in several key statistical categories, and where they stack up against the rest of the league:

Net efficiency and simple rating system (SRS)

If you want to answer the question, “How good is this NBA team?” with a single statistic, net efficiency and SRS are probably the two leading candidates for providing a reasonably accurate answer. The Nuggets have taken a considerable dive in both, and their drop in league rankings (from 5th to 16th and 5th to 14th respectively) in these categories closely mirrors their drop in the NBA standings from 4th last season to 18th currently. Simply put, the numbers confirm what everybody watching the Nuggets already knows: The team has gotten substantially worse.

Offensive and defensive efficiency

Despite the fact that, somewhat surprisingly, Denver’s defensive efficiency has actually improved (if only so slightly as to amount to no meaningful difference), their standing relative to the league has dropped from 11th to 21st. This is likely due to the relatively fickle nature of league rankings at this early point in the season, or perhaps even the possibility that the league as a whole has improved defensively. Irrespective of the cause, it’s an interesting discrepancy and it will be worth tracking later in the season to see if it holds.

More significantly worrisome is Denver’s decline in offensive efficiency. It has taken a major tumble from 107.6 to 101.8, and what used to be their stalwart against becoming too bad despite poor defense has not, at least up to this point, been reliable this season. A big reason for this is that all four players the Nuggets brought in to address last season’s mediocre-to-bad shooting outside the paint – Randy Foye, J.J. Hickson, Nate Robinson and Darrell Arthur – have seen their shooting percentages drop markedly since joining their new team.

Shooting, assists and turnovers

The decline in Denver’s shooting percentages, as evidenced by the slippage in both true shooting percentage and effective field goal percentage, is the most obvious culprit to blame for their decreased offensive efficiency. Although their 3-point percentage has in fact improved this season from .343 to .378, this has been offset in part by the that a decrease in 2-point percentage (from .516 to .462) has offset the slightly larger share of total field goal attempts being taken from the arc this season (just over 23 percent of Denver’s field goal attempts have been from the 3-point line this season, compared to just under 22 percent last season). We will examine their shooting in greater detail below.

The next category where Denver has taken a hit this season (unfortunately they’re nearly all moving in the same downward direction) is assist rate, where they have slid from 17.9 to 15.8. This is not the fault of Ty Lawson, whose individual assist rate has leapt from 30.2 to 37.3 (his career average is 29.1), but fellow point guards Andre Miller and Nate Robinson, whose assist rates have dropped from 32.2 to 21.8 and from 31.4 to 21.8, respectively. The loss of Andre Iguodala who at 22.4 had the Nuggets’ third highest assist rate last season, certainly hurts as well. Aside from Lawson’s improvement, the only real consolation in the ball-handling department is that the team’s turnover rate hasn’t gotten any worse, and they’re still in the top third of the league.

Rebounding

The Nuggets were hoping to offset the loss of their second best rebounder last season in Kosta Koufos (rebound rate 17.3) with the addition of J.J. Hickson who, playing mostly at the center position for Portland last season, led his team with a rebound rate of 20.7. His drop to 18.9, Timofey Mozgov’s chronically poor rebounding (14.7 career rebound rate, 14.1 this season), the loss of JaVale McGee to injury, and the departure of Koufos have all combined to put a hurt on the Nuggets’ overall rebounding. What’s especially disappointing is that the addition of Hickson not only didn’t improve Denver’s defensive rebounding, they’ve gotten even worse and are now 29th in the league.

Pace

While none of the above is likely to be all that surprising to most Nuggets fans who have watched the team closely this season, pace stands out as the one category which defies expectations and the eye test. Despite Shaw’s (at least stated) intent to focus more on executing in the half court while retaining but somewhat de-emphasizing speed, the Nuggets are actually playing at a significantly faster pace this season than last. If, after watching a good amount of sluggish play from the team this season, this doesn’t quite compute with you, then join the club. But the numbers don’t lie, and Denver is still running as much as ever.

 

Shot locations

By taking a look at where on the court the Nuggets are shooting from in comparison with last season, we can quickly get a better idea of why their offensive efficiency has dropped so far:

And just to provide you with a graphic view of the above stats, here are shot charts from 2012-13 and 2013-14:

Last season under George Karl’s rim-attacking dribble-drive-motion system, the Nuggets took nearly half their shots at the rim, a whopping 45.9 percent. Denver led the league with 3194 attempts in the restricted area, more than 500 more than the next in line (Detroit, with 2670). As most of you already know, the two most efficient shots in basketball (yielding the most points per possession) are corner 3-pointers and shots at the rim. Leaving aside for now whether it was well-suited for playoff success, Karl’s approach was designed to maximize the frequency of these high value shots. And at least in the regular season, it was highly successful.

The most dramatic change we can see in Shaw’s approach in terms of shot selection is the enormous drop-off in at-rim shot attempts. As a share of Denver’s total shot attempts, they have plummeted down to 34 percent, a full 12 percent decline which effectively means that instead of taking close to half their shots in the restricted area, the Nuggets are now taking just a third. Compounding the damage is their decline it at-rim field goal percentage, which has lipped from .631 to .593. In Shaw’s defense, it would be either flatly wrong, or at least premature, to suggest that this is entirely his fault, given that he doesn’t have the benefit of deploying four of Denver’s top six players last season in made at-rim shots (Koufos, Iguodala, Brewer and now McGee). But we have also seen a hesitancy from everyone not named Ty Lawson to attack the rim with the same fervor they did last season, and that appears to at least be partly attributable to their efforts to adhere to Shaw’s system of running the offense through the low post and elbow.

This might not be as much of a problem if they were replacing one efficient shot (in the restricted area) with another (corner 3-pointers and above-the-break 3-pointers), but in fact nearly all the displacement of at-rim shots has been diverted to mid- and long-range 2-pointers. To a certain extent this is to be expected since Hickson and Arthur do most of their shooting from these areas, but it seems clear at this point that the pendulum has swung too far in the direction of inefficient, low-value shots, and that Shaw will need to find a way to restore better balance to Denver’s overall distribution.

Denver’s 3-point shooting has also taken a shift, albeit a more subtle one, towards less efficiency. The good news is that their corner 3-point percentage has increased dramatically from .351 to .500. But the bad news is that they’re taking fewer corner 3-pointers and more above the break, and also the probability that the .500 mark can’t stand and will regress to the mean in due time. But the Nuggets do have, at least on paper, a more capable crew of long-range shooters this season, and focusing on getting them more good looks from the arc should be a point of emphasis for Shaw.

 

Play type

A look at how much of the time the Nuggets are using different types of plays provides some detail to the breakdown of their shot distribution:

From left to right, the above play types are ordered by their frequency last season. And immediately we can see a partial explanation for this season’s lack of at-rim shots: fewer baskets in transition. This is another case in which Shaw can at least be partly absolved of blame, as he doesn’t have the luxury of relying on Brewer and Iguodala on leak-outs and fast breaks off steals. But given that the Nuggets are still, in fact, playing fast, this is an area he really needs to cultivate.

In terms of play type, the strongest expression of Shaw’s new system can be seen in the increase in pick-and-rolls and post plays. Taken collectively, these plays have jumped from 22.1 percent of Denver’s offense to 30.9 percent, a clear reflection of his halfcourt emphasis. The Nuggets’ PPP hasn’t changed dramatically for better or worse in any of these areas, which is at least not bad news. But taken with the decline in transition plays and cuts (ie. At-rim shots) and spot-ups (most 3-pointers fall in this category, though it also includes spot-up 2-pointers), the result, as mentioned above, is that even if none of the shooting percentages drop, the overall shift in the shot distribution from more efficient to less efficient shots will result in lower overall offensive efficiency.

 

In conclusion

The Nuggets are going through a difficult transitional period, further complicated by injuries which are setting back the team’s learning curve. But some of the problems described above might be alleviated, at least to a certain extent, if Shaw took one or more of the following steps:

  • Stop posting up Faried. Let him just play at (and above) the rim, and cut to the basket. Faried will never be a good post player. I repeat, Faried will never be a good post player. Stop trying to make him something he’s not, and you will be rewarded with better production and more efficiency, and probably a more spirited defensive effort as well (since he’ll be having more fun).
  • Give more minutes to Jordan Hamilton and Evan Fournier, and fewer to Nate Robinson and Randy Foye. Through his entire short career, Hamilton has quietly been an exceedingly good rebounder for his position. This season, his rebound rate is as good as or better than that of the likes of Randolph, McGee and Arthur. Fournier isn’t a superb playmaker, but he’s a competent one (at least as good as Foye). Both J-Ham and Evan have a more diversified offensive arsenal which includes the ability to attack the rim and hit from long range, in addition to other attributes (rebounding in Hamilton’s case, assisting in Fournier’s) which make them better all-around players than their teammates who (combined) are getting nearly ten more minutes of playing time than them. It will be much more valuable in the long run to cultivate consistency in Hamilton and Fournier than it will be to lean on the occasional hot shooting night (and not much else) from Robinson and Foye.
  • Run more plays for the 3-point shooters. I’m fine, in principle, with wanting to build a team better designed to weather the playoff grind. But one need look no further than the San Antonio Spurs to know that it’s quite possible to run your fair share of halfcourt offense through the low post and elbow without having to abandon high value shots from the corner and at the rim. And in fact, doing both simultaneously should help to spread the floor, keep defenses honest, and in turn open up better looks from the post.
  • Get Wilson Chandler more involved in just about everything. With Iguodala gone, Chandler remains as the most Iguodala-esque, two-way jack-of-all-trades player on the roster. It may have been just a fluke, but the fact that he had only two shot attempts in 27 minutes against the Bulls is a mistake that should never happen again.

Statistics courtesy of NBA.com/stats, MySynergySports.com and Basketball-Reference.com

You can follow me on twitter here: @denbutsu

  • Makkon Farstriders

    Great article- lots of good stats here. I can tell a lot of work went into this thing. I’m looking forward to the defense and individual breakdown articles.

  • Charliemyboy

    Good analysis. Do not want to wait too much longer to know this will be the team, no longer in transition. Playing much better. Indications of CQI. BIG test today; if they beat Dallas, they will move from 27th to 22nd to 16th to 12th on Hollinger. Worried about Gallo, after seeing what happened to Rose. Poor Rose. Very sad for the NBA. Gallo drives w/o abandon. He may have to change his style to last long. Better let Ty rest. He rested on his own last game. Agree on Fours and Ham; keep giving them time. JaVale needs to pump iron, get into a good fight and be serious an yelled at to grow up. Now, everybody happy?: Iggy was injured. Looks like Karl might end up in Toranto or NY or NJ?

  • Ernie

    Thanks for the info. There is almost too much analysis here, I kind of wish it was done in two or three pieces to digest it all.

    It looks to me as an observer that the problem is too often the guards dribble around for a while and then pass to forwards taking 15-18 foot jumpers. And those forwards do not have Gallo’s skill at shooting. Seems to be less of an emphasis on passing and moving the ball, regardless of pace. Which unfortunately is neither efficient nor fun to watch.

  • Poz303

    Ah, Denbutsu, been a while since I have seen your statistical analysis. Great work and some interesting findings. I am surprised the pace is higher, might be because Andre Miller is getting less playing time this year.

    I agree with your conclusions. I have been saying Fournier should start ahead of Foye. Shaw did make me shake my head when he didnt put Fournier back in the game in the second half after having played a great first half against Chicago. I wouldnt have thought to play JHam more but last game he (again) teased with his potential. His rebounding has always been very good for his position.

    I am surprised they dont take more corner 3′s especially after seeing their shooting percentage from that location.

    Thanks for the article.

  • trank

    thanks, good stuff, and suggestions. go nugs.

  • NiceGuyHickson

    Will Denver make the playoffs !

  • gimpcom187

    very well done article. Assuming you take a stats in depth look again id break it into 2 articles: offense and defense and look at their defense a little more thoroughly. Either way thanks for the data article.

    PS maybe I missed it, but I would have enjoyed seeing a stats heavy article from you concerning last year’s playoff series after the Warriors Spurs series (where it became clear the Warriors were playing on another level than their regular season showing). All I saw here was hatchet jobs on Karl… Maybe I missed it though.

    In terms of the pace they have increased after the 3rd game or so. shaw did adapt quickly which is a good sign.

    • http://www.roundballminingcompany.com/ JoelRMC

      Yeah, we will definitely break down the defense as well at some point. That would have been just too much (= too long) for this post, though.

      As for “hatchet jobs” on Karl — While I don’t think any of our writers have pulled any punches or been shy about being critical when (from our perspective) criticism was merited, I certainly don’t recall writing anything or reading anything from the other writers that could fairly be called a hatchet job. Harsh? At times, definitely. Cynical and snarky? Probably. But if you have any examples of any of us really going overboard and venturing into truly unfair/inappropriate territory, and can link to them, I’d like to see, because that really doesn’t fit with my understanding of how we covered Karl here.

      • gimpcom187

        I only have 1 writer I am referencing, but hatchet job may be an exaggeration. “Clearly negatively biased against Karl” is just so cumbersome. Im only interested in calling him out when he writes about Karl so I’ll refrain in this comment section. Nonetheless, I would say the point was there was some specific statistical analysis that could have been made as to why the nugs struggled in that series outside of blaming Karl and it wasn’t mostly glossed over.

        . The Warriors D became much better with bogut. In addition to the ATROCIOUS shooting of Chandler and Brewer in that series. There was little reference to that on the site after the series was done (some in the game by game recaps, but when summing up the season and series it mostly veered toward: Karl is a get us to the 1st round coach he can’t do more than that, blah blah blah). As the warriors showed in round 2 with the spurs (who were essentially co-champions and have the indisputably best coach in the league) They were a VERY difficult team to beat. Im almost sure I saw ZERO on the site by the writers referencing this development.

  • Ckwizard

    Lol just wait like what you did here and the subject but come on man this team is about on par with last years team THROUGH THE FIRST 12 GAMES… Your analysis has a year with a partial Gallinari vs 12 games no Gallinari. Actually for the number of new rotation players this team is better than last years team and is showing better chemistry and is improving defensivly. It is also nice to see some attention to things like “boxing out” and “Free throw shooting”

    Keep the template and see how they stack up at mid season and end of season. I will wager this team wins more “road” games than last years team and has a respectable home record again. But like I said last year time and time again this team is going to get better and still has a lot of room to grow! This team will finish Fourth or Fifth in the conference! But hey I was wrong last year when they were sitting at Less than five hundred and I held my optimism… Oh wait no I wasn’t.

    • http://www.roundballminingcompany.com/ JoelRMC

      I actually agree with your first point and I thought I made that clear when I wrote — in the second sentence of my post — that “This is a small sample of games – certainly not significant enough to
      project what will happen in the latter half of the season, especially
      considering the fact that the team is adjusting to a new system and
      missing two of its best players in Danilo Gallinari and JaVale McGee.”

      That said, while some of the differences from last year might wash out with regression to the mean or simply returning to full health, there are some changes showing up in the stats that don’t just reveal what the team *is* doing differently but also mirror what Shaw is *trying* to do differently, most especially running more post-ups and P&Rs. So those aspects in particular I think can give us, if not an accurate picture of how this team will end up, at least a relatively clear idea of which general direction it’s headed in.