NBA Week 4 in Review: Seasonal Changes

Nov 21, 2015; Houston, TX, USA; New York Knicks forward Kristaps Porzingis (6) drives the ball as Houston Rockets forward Marcus Thornton (10) defends during the fourth quarter at Toyota Center. The Knicks defeated the Rockets 107-102. Mandatory Credit: Troy Taormina-USA TODAY Sports
Nov 21, 2015; Houston, TX, USA; New York Knicks forward Kristaps Porzingis (6) drives the ball as Houston Rockets forward Marcus Thornton (10) defends during the fourth quarter at Toyota Center. The Knicks defeated the Rockets 107-102. Mandatory Credit: Troy Taormina-USA TODAY Sports /
facebooktwitterreddit
Kristaps Porzingis
Mandatory Credit: Troy Taormina-USA TODAY Sports /

After four weeks, the NBA has lost that new car smell and several teams, like the Los Angeles Clippers and Houston Rockets, are wondering if it’s time to panic. There’s enough of the season left where possibilities are expansive, but it’s tough to make up ground if you’re far behind and dreams are already being awarded or crushed. With a keen eye towards both stats and video (yes, you can do both), let’s take a look at a few random topics from last week in the league.

The Big Apple with the Big Improvement

After a season where the New York Knicks finished with the second worst record in the league, they are admirably above 0.500 after nearly a month’s worth of games. It’s not just because of Carmelo Anthony either, because even when he played last season they were just 10-30. He hasn’t shot well this season so far, but he’s at least kept his turnovers down, which has been a strength of the team so far. But he hasn’t been their most exciting player.

Enter Kristaps Porzingis — much has been written about him already, so I’ll try to find new angles. What’s remarkable is that he’s a rookie who’s a large part of a decent offense,  as the list of large players his age with a usage rate as high as his or better is quite short and filled with future all-stars. Thankfully, he doesn’t appear to be a defensive liability like the few duds on that list — Eddy Curry, Enes Kanter, and Andrea Bargnani. But the improvement is with their entire frontcourt. Robin Lopez is a steady force of production at center and Kyle O’Quinn is too, adding some much needed defense.

Unfortunately, there’s one sign the Knicks are shining with some fool’s gold right now: they’re number one in three-point percentage defense at 29.6 percent, which essentially just means they’re lucky because teams have little control over percentages. If opposing teams had been hitting those shots at the league average percentage, it’d translate to minus-3.2 points per game or, over the full course of a season, roughly 8 losses[1. Adjustments are needed with that number, however, as the Knicks would not recover every single three-point miss, but it gives an idea of what sort of magnitude this stat has. With a fairly conservative adjustment, the transition is still around -2.5 points per game.]. If you’re assuming the opponent three-point percentage will hold, I’ll also point out that last season the Knicks were last in the league in that defensive stat.

Mandatory Credit: Jeremy Brevard-USA TODAY Sports
Mandatory Credit: Jeremy Brevard-USA TODAY Sports /

Charlotte’s Modern Lineup Tweak

Due to an injury to Cody Zeller, the Charlotte Hornets had to create a new lineup and, lo and behold, the team with Al Jefferson and his supposed throwback game was playing small ball with Marvin Williams at power forward and two point guards in Kemba Walker and Jeremy Lin. As I wrote last week, the team has really been struggling with Zeller in the lineup and the opposite with Williams. The cause here might be how Zeller is one of their only rotation players, besides Jefferson, without three-point range and the team has been shooting better without him, per NBAWOWY.com. Zeller will be back soon, but this is something to track — will Charlotte follow league-wide trends and their own stats and bench Zeller, or will they continue with the status quo?

Mandatory Credit: Bill Streicher-USA TODAY Sports
Mandatory Credit: Bill Streicher-USA TODAY Sports /

Revenge of the Doug

Last season, Doug McDermott was unequivocally one of the worst players in the league in limited minutes, and this was after an interesting statistical discussion during the draft because he had some of the lowest steal and block rates you’ll ever see from a college prospect. Poor performance in those categories is consistently a terrible sign for future prospects. His advanced stats from 2015 were from the bowels of the earth. Doug’s Box Plus-Minus, was so poor that a team of only Dougies would be expected to be outscored by over 33 points per game over a season. John Hollinger’s PER is usually the kindest metric for shooters, but his PER was 6.1. Since league average is 15 and end of the bench guys are usually around 11, that’s atrocious. Only seven players in NBA history with at least 1500 minutes have ever equaled that mark or worse. The problem with those metrics is that they’re blind on large segments of the game, but that’s mostly defense — don’t judge Bruce Bowen or Jason Collins using PER — and RPM, which looks at a player’s impact, wasn’t impressed either, since he ranked 458 out of 474 players last season. But he played just a little over 300 minutes and he was a rookie, so a final judgement is premature.

So far, McDermott has played very well, and predictably it’s been because of his shooting, headlined by a 64 true shooting percentage. But you can’t just be a stand-still shooter and excel in the league; it’s too fast and the NBA’s ultra long and athletic players can close out quickly. Here’s a play where he’s fighting hard for an opening by moving without the ball and cutting hard, managing to lose three Phoenix Suns players before finishing inside. Even if you’re slower than everyone else, you can still score inside on plays like a backdoor cut. He’s using intelligent movement to get open on the perimeter too, like this play where he fakes out his defender and cuts the other way for daylight, then nails a three-pointer. His defense will always be poor, so he needs to score often and efficiently enough to outweight it. Currently, it’s actually been a success.

Mandatory Credit: Kyle Terada-USA TODAY Sports
Mandatory Credit: Kyle Terada-USA TODAY Sports /

Weekly Check on Transcendence

The Golden State Warriors still have not lost, and it’s hard to imagine how they can. Through the games on Sunday, if the Warriors started every game by spotting their opponents ten points, they’d still have a good enough record for fourth in the Western Conference. They’re not pummeling the league into submission through sheer brawn and grit, thankfully; it’s more like an art performance. They’ll have beautiful plays like this one, and they’ll push the pace and score quickly and attempt wild three-pointers. How great is it to have a dominating team that doubles as the league’s most aesthetically enjoyable?

Mandatory Credit: Mark D. Smith-USA TODAY Sports
Mandatory Credit: Mark D. Smith-USA TODAY Sports /

Old Germany

After a rough season last year, Dirk Nowitzki has looked vigorous and he’s been shooting the dimples off the ball. He’s actually shooting less often than he was last season and it’s his lowest usage rate since he was 21 years-old, which is a concern, but he’s made more than half his threes and he’s well above average everywhere. He’s actually 13th in the league in RPM, which is the best public barometer of offense and defense. He’s still a focal point for opposing defenses and when he’s nailing his shots, he’s a huge boost to a team’s efficiency because he rarely turns the ball over and takes a number of tough shots most other players convert at low rates.

Nowitzki’s skill and size are a combination that terrorizes opposing defenses because if a smaller player gets switched onto him — and this is something that happens with great frequency in the modern NBA — he can bury that smaller player with a turnaround shot. Here’s and example where a pick-and-roll is reset several times just to create the right match-up. Bigger defenders aren’t comfortable guarding him either, because he’s a dangerous catch-and-shoot player. Most stretch 4’s these days are pretty shaky shooters who can be given a lot of breathing room, like Favors did to Nowitzki here, which ended poorly. Dirk, however, is one of the greatest shooters ever. It’s tough to tell how a player will age but he’s going to be able to nail those jumpers until the oceans dry up.

Mandatory Credit: Gary A. Vasquez-USA TODAY Sports
Mandatory Credit: Gary A. Vasquez-USA TODAY Sports /

Dunk of the Week

The first one, and the front-runner, was a DeMar DeRozan slam where he catches as much air as you’ll ever see in an NBA game, and for an added level of difficulty it was over the Stifle Tower, Rudy Gobert. Andre Drummond, who’s currently enjoying a breakout season of sorts, is a wrecking ball inside, and he hammered down on the rim pretty hard in a game versus Washington. It’s like a highlight stolen from Amare Stoudemire ten years ago. Lastly, after a strange start to the season, Gerald Green displayed his still elite vertical skills for a pretty reverse dunk. Alas, it was after the whistle, but Green, when given playing time, is always in the running for dunk of the week. This leads me to an important question: what should those dunks be called, the ones that do not count and do nothing but entertain the crowd and annoy the coaches?

Mandatory Credit: Matthew Emmons-USA TODAY Sports
Mandatory Credit: Matthew Emmons-USA TODAY Sports /

Free Throw Defense

Defense seems pretty simple to measure for teams. Look at how many points they’ve allowed, maybe adjust for who they played, and then put it on a different scale, like per possession. But when you break it down by its components, you see issues with measuring defense. There’s been a lot of discussion about three-point defense, like in this very week in review concerning the Knicks, but free throw defense is even more inexplicable: it’s the only action where you score and the defenders can do nothing but stand at the sidelines. Perhaps there’s something to free throw defense, like being able to foul the right players or hitting them hard enough they’re seeing stars as they walk to the charity stripe.

To test the veracity of this facet of defense, I separated every season from 1997 to 2015, except for the short 1999 season, into roughly equal halves using January 20th as the cut-off date[2. For the 2012 season, I used February 24th.]. Then I simply compared a team’s opponent free throw percentage from one half to the other. Presumably, if free throw defense is actually real, if a team defends well during one half of the season, surely they’d do the same for the second half.

The results are intriguing because there is a bit of a positive relationship. The correlation coefficient is 0.275, which indicates a little over a quarter of a team’s free throw defense is “retained” from one half of the season to the second. For comparison, coefficients are shown for three-point percentage and two-point percentage. Three-point defense actually has less of a correlation, and this is important because teams are taking so many three-pointers now it has a significant effect on defense. Conversely, two-point percentage holds up very well from one part of the season to the other. This is likely because those shots aren’t usually open and depend on how well the defense contests. Three-pointers are usually taken standing still when a shooter has enough daylight, and free throws, of course, are not even guarded.

Table: Team opponent stats, first half of season to second

Team statsCorrelation (0-1)
FT%0.275
3PT%0.156
2PT%0.715

What’s surprising is that free throw defense apparently exists, though it’s muted. There are a couple of reasonable explanations. One is that teams do have some control over who gets fouled and who goes to the line. Smart defenses will not foul good shooters on the perimeter while they’re shooting and they’re eager to hack poor shooters when they have a decent shot near the rim. For a quick look at this, I calculated the expected opponent FT% based on a player’s season FT% for a few teams at the extremes. The other mechanism has to do with how free throws are chained together. Players shoot worse when they take only one at a time versus two, and they’re even better when taking that third free throw when getting fouled on a three-point attempt.

The expected percentages were a lot closer to the league average, suggesting that teams don’t “choose wisely” with their foul targets. But the expected percentages weren’t exactly league average, and the Pistons in particular looked poor in an expected percentage that was a whole point above the average. But there’s a clever explanation here: because of Andre Drummond, Detroit was the second worst free-throw shooting team in the league, and since they can’t play themselves their schedule had better free-throw shooting teams than most.

While a difference of one or two percent is marginal at best, one month into the season the range is much wider. For fans of, say, the Pelicans, who are arguably the league’s unluckiest defense since their opponents have the highest three-point percentage and the second-highest free throw percentage, this brings hope. Teams can’t control everything that happens on the court, including a few components of defense, and understanding that leads to a more accurate perception of the league.

As a bonus, below is a table showing defensive ratings from basketball-reference adjusted for opponent free throw percentage and three-point percentage[3. The formula is: FT% = 0.13382*TeamOppFT% +0.86618*LAvgFT%, where the LAvg is calculated without the corresponding team. For three-pointers, the formula is 0.12387*TeamOpp3PT% +0.87613*LAvgFT%. Then the disparity of actual free throw, or three-pointer, points surrendered versus predicted is calculated per 100 possessions. After that, offensive rebounds are estimated with a super simple formula: 0.13*(1 – FT%) for free throws and 0.23*(1 – 3PT%) for three-pointers. Both factors are multiplied by the number of free throws per game (or per 100 possessions) and the difference between opponent % and the estimated %. Finally, a conservative factor of 0.9 is applied to each component, and then they’re subtracted from the team’s defensive rating.]. This is an experimental design, and it’s still early in the season, but a few things stand out and make sense. Miami’s defense doesn’t look truly elite, and Golden State looks a bit more mortal — they have been getting lucky, which happens to every great team on a win streak. The Wizards and 76ers look better, which makes sense; they were better last season on defense. The Knicks tumble down the rankings, and that should illuminate something about their nice start. But the biggest change is New Orleans, who go from absolutely awful to something more palatable with someone like Anthony Davis on the team. These numbers should be adjusted for strength of schedule too, but further into the season schedules are mixed enough that this method alone could provide some smart adjustments.

Table: defensive ratings for 2016 adjusted for opponent FT% and 3PT%[4. These ratings were changed after initially posted due to a change in the formula.]

TeamDrtgAdj Drtg
San Antonio Spurs8.07.8
Miami Heat6.84.3
Boston Celtics4.94.2
Indiana Pacers5.93.5
Detroit Pistons2.53.0
Chicago Bulls3.72.7
Utah Jazz2.42.5
Golden State Warriors4.22.4
Toronto Raptors2.02.3
Cleveland Cavaliers1.52.2
Atlanta Hawks0.21.1
Charlotte Hornets0.50.9
Phoenix Suns1.80.7
Oklahoma City Thunder-0.90.4
Washington Wizards-1.70.1
Orlando Magic0.8-0.2
Minnesota Timberwolves0.6-0.3
Philadelphia 76ers-3.1-0.8
Denver Nuggets-4.0-0.9
Dallas Mavericks1.2-1.1
Memphis Grizzlies-2.4-1.1
Brooklyn Nets-4.3-1.9
New Orleans Pelicans-7.0-2.5
New York Knicks-0.1-2.9
Los Angeles Clippers-3.4-3.1
Houston Rockets-4.1-3.4
Portland Trail Blazers-2.1-3.9
Sacramento Kings-2.7-4.0
Los Angeles Lakers-4.6-5.4
Milwaukee Bucks-7.4-6.4