NBA Week 11 in Review: Databall Blues

We’re deep into the NBA season and nearing the halfway mark for the regular season, as strange as that sounds. There are still a lot of unresolved storylines and surprises, however, even with all the tools we have available. The media’s focused on the circus in Cleveland and the condemned basketball empires in New York and Los Angeles, but, as always, there’s a lot of other stuff to analyze too.

The Whiteside of the Paint

In a depressing season for the Heat, their luck finally turned around. They unearthed a draft prospect from years ago, Hassan Whiteside, the prototypical super-long big man with raw skills. With an impressive 7′ 7″ wingspan and a 9′ 5″ standing reach, one of the best in the league, he had that tantalizing potential on defense and was pretty quick off his feet, blocking a ton of shots in college. But he bombed in the NBA and barely played in two seasons, wandering the globe going from team to team.

But last week he averaged a 14.9/9.8 per game line with 3.6 blocks a game and 76.7% from the field in four games. Those are the kind of eye-popping stats you don’t see from just any journey-man center in limited minutes. He’s been an active part of the offense too with a decent usage rate, relying on a plethora of hooks and dives to the rim. See this hook shot here and an alley-oop here. With his world-class wingspan, it’s tough to contest his hook and it’s easy to lob a pass to him near the rim. But shotblocking is his signature skill. You can see him denying Blake Griffin in the picture below (video is here.)

week 11 whiteside

Here’s what interesting about Miami’s poor start: I correctly pegged them as a sightly below average offense, estimating that they’d finish 19th in the league in offensive efficiency. Where are they now? 19th. Where they’ve under-performed is on defense, and Whiteside might be the elixir here for a team in shambles.

Booker’s Trick Shot

A few days ago, Trevor Booker had one of the most bizarre made shots in an NBA game. With the shot clock near zero, there was no time to square-up and shoot, so he caught the pass with his back to the basket and flicked it over his head … and somehow made it. It was essentially a H-O-R-S-E shot. What’s funny is wondering about how people recorded the attempt for statistics. The NBA play-by-play log called it a jump shot, but it certainly was not that. You can see the video here.

Kyle Korver the Real Life No-stats All-star?

Given Atlanta’s dominance in the first half of the season, there’s been a lot of discussion about which Hawk should be graced with an all-star appearance — to represent the team and not the game representing the best players in the league. Since Jeff Teague is the leading scorer, he’ll probably get in regardless of anything else he does. However, leading the team in scoring should not make you a lock for the all-star game, especially if it’s with a lukewarm 17 points per game average. Al Horford has made it in the past, but it hasn’t been his best season and his minutes are down. Paul Millsap is in all likelihood the best player on the team and he made it last season — but the media will surely ask for two Atlanta players.

But as Zach Lowe has argued recently, but what Kyle Korver? He’s having one of the best shooting seasons ever, and he receives a high amount of defensive attention because an open Korver shot is death incarnate. He’s shooting about 55% on open three’s, which is a step below the effectiveness of an open layup and not much else. ESPN’s fancy new stat RPM has him rated as the 13th player in the league, and it shouldn’t be so easily dismissed because he was highly rated last season too. Basically, he has all the characteristics of an offensive star and his team does better when he plays; that’s all the stat is saying. In fact, even basketball-reference’s new BPM loves him and sees Korver as the best player on the team, even though all it’s using is box score stats, the kind people traditionally rely on for all-star arguments. Naturally, Korver’s scalding-high shooting percentages are partly the cause here, but this is also the guy who owns the three-point percentage record — it’s no fluke.

The problem is that there are no precedents for a player of his ilk making the team, and the media will surely bring this up in arguments. But doing things the way they’ve been done before just because that it’s been done that way is terrible reasoning, and we don’t need to repeat the mistakes of the past without any real reasons. The all-star game is for the fans, sure, but fans aren’t clamoring to see Millsap and Teague, and people still depend on all-star appearances in career retrospectives and things like Hall of Fame discussions. It is important, not just “for the fans.” The obstacle here is that most all-stars are high scorers. If you look at guys under 15 points per game, which includes Korver, it’s a fairly long list, yet the majority of those are high assist point guards and big men who pick up a lot of rebounds and/or blocks. Parring down the list further, since the merger there are a small number of guys under 15 PPG, 8 RPG, and 7 APG. But there are two basic types on that list: reputation selections like Kareem at the end of his career and a bunch of injured guys like Kobe last year, and defensive aces like Dumars, Iguodala, and Hibbert. There is no precedent for a guy like Korver, and the road will be a tough one to beat to make it to the game, but man, what a shooter.

Starless Charlotte

Even though Charlotte brought in the highly valued Lance Stephenson, the team had a disastrous start to the season. However, this changed when they benched Lance — they were 6-19 with him and 9-5 during his injury. (He actually just came back from injury. The result? Another loss, ending their win streak.) Then Al Jefferson got injured, which should have sent the team into a tailspin, according to most NBA analysts who value his post scoring. But they’re 5-3 so far without him — it’s a soft schedule, sure, but they weren’t winning against anyone in the first couple months when he played.

The Hornets are doing this largely by playing better defense: their opponent eFG% allowed fell from 51.2% to 47.1%. Their offensive rebound rate also climbed by nearly 25 percent — that’s an enormous change. That’s largely personnel change though because Jefferson’s backup, Biyombo, has a rebound rate three times as high on offense. It’s also a distressing sign for Jefferson’s impact that Maxiell was the frontcourt player who picked up the most minutes in his absence and the Hornets improved.

The worst part of the Charlotte season so far also coincided with Kidd-Gilchrist’s absence — he’s one of the most overlooked players in the league — but there’s more value on defense in having a mobile big who can get up and challenge or block shots. You can see this when Al Jefferson defends a shot: in this video and the picture below, you can see his limited vertical-space defense leads to a Bledsoe layup. Here’s another example where Thomas drives right to the rim and over Jefferson pretty easily. Biyombo, however, is enough of a threat as a defender that even an aggressive point guard like Lowry hesitates and uses a floater when driving, like in this video (screenshot below.) Why wouldn’t Lowry get any any closer? Because Biyombo blocked him like this earlier, and it’s the type of help defense Jefferson has never been capable at executing.

week 11 jefferson

week 11 biyombo

The Clippers Bench: the Has-been’s

The Clippers, despite all their issues with Griffin shooting too many midrange shots and their problematic depth, are still one of the best teams in the league. However, it’s disappointing how the team’s management is restricting the potential of the team. The franchise is enamored with fringe stars and role players who were only relevant years ago when Doc won a title in Boston and was competing for another. But it’s not all on Doc because they’ve had this issue for a while. Of course, teams with a lot of cap space tied to superstars always need to target veterans because they’re cheaper, but it’s a bit absurd just how many “has-been’s” they’ve had and how little real help they’ve been able to find:

Bench help since 2012:
Hedo Turkoglu
Glen Davis
Danny Granger
Stephen Jackson
Antawn Jamison
Sasha Vujacic
Caron Butler
Lamar Odom
Grant Hill
Chauncey Billups
Mo Williams
Kenyon Martin

Grant Hill and Chauncey Billups aren’t fringe-level stars, but they certainly didn’t play like that with Los Angeles. Smart teams know how to filter out the Danny Granger’s from the Chris Kaman’s in the veteran market, knowing how to find guys who still have something left in the tank. The Clippers purged a great evil from their executive class, but they still have a lot of work to do in the front office — Doc Rivers should not be their GM.

Did Wiggins Have a Breakthrough?

There’s been a lot written about how much better Wiggins has been playing, but I’m not seeing it yet. Here Kevin Pelton argues that the change in the starting lineup with Shabazz Muhammad at small forward, forcing smaller guys on Wiggins. While his TS% has certainly had a vast improvement, going from 47% to 54%, what player are we seeing here? He still has a modest usage rage around 22-23 — that’s not the sign of a superstar wing. The comparisons to LeBron are bizarre because even right out of high school he had a usage rate of 28 his rookie season, and so did Carmelo, another highly valued prospect who came in after a year of college.

What’s happening here is that people are projecting his athleticism and big-name status from high school and assuming he’s just like the rest of the prospects. But we’ve been down this road before: sometimes a prospect will wash out of the league completely, sometimes they’ll turn into Gerald Green, and sometimes they develop like Paul George. It’s tough figuring out what will happen, and anyone peddling certainty is either a wizard or a liar. A lot of his potential is on the defensive end, but, again, it’s tough estimating when that will happen — the Wolves are a horrible defensive team now anyway. Wiggins doesn’t even have the high steal total that’s common among his ilk — young, athletic defenders — and it’s comparable to Carmelo’s steal rate per minute when he was a rookie. One might be ? to compare him to McGrady because their shooting stats were surprisingly similar: a low TS% of 50 and a usage around 20, but the peripheral stats are significantly different. McGrady’s rebounding was almost twice as high, his assist rate was twice as high, he had more steals, and he had nearly four times as many blocks. Some of those stats aren’t important on their own, but they’re rough guides on a player’s potential. Using NylonCalculus’s player comparison engine*, here’s the list of the most similar players: Alec Burks, Gerald Green, Rudy Gay, Richard Jefferson, Marvin Williams, Luol Deng, Richard Jefferson, Corey Benjamin, Larry Hughes, Javaris Crittenton, and Lamond Murray.

*The weights I chose: 0.5 for 2P attempts and free throws; 0.6 for 3P attempts; 0.4 for FT%, eFG%, defensive rebounds, assists, steals, and height; 0.35 for turnovers; 0.3 for blocks, 0.25 for offensive rebounds, and weight; 0.15 for fouls; and 1 for age.

Here’s a typical Wiggins possession where, fairly early in the shot clock, he settles for a long two-pointer. But those types of shots are the only reason he can even average 18 a game because he’s not adept at creating his own shot off the dribble. In this video, he’s isolated against Dragic, tries to drive, and then just loses the ball on the way to the basket. He had six turnovers that game. However, he also showed off why people are still so in love with him: he actually has a nice little sidestep in transition and jumps sky-high for a dunk in transition.

week 11 wiggins

Raw SportVU Tracking Data

There’s an old aphorism attributed to many different people over history. In one version, Oscar Wilde says, “In this world there are only two tragedies. One is not getting what one wants and the other is getting it.” It wasn’t too long ago that basketball fans only had access to box score stats, a simple summation of a game’s events that misses large swaths of what happens on the court. Now we have tracking information on every single player and the ball broken down not by the second but even further. We’ve gone from calculating something like points per game to convoluted stats based on thousands of lines of data … for every game.

For an example of the data I have access to, I included a graph below of a single event (which is what everything is organized into) from the first game of the season. What is it? What’s going on?

sportvu first possession example

The Y-axis there for height is showing the height of the ball throughout the play. The first big jump there is from the jumpball that starts the game, and following that are a series of dribbles going from the floor up to nearly five feet. Small jumps are from passes or a player holding the ball up. The big jump is the shot attempt, and you can see how as the ball descends it stops at ten feet and goes back upwards a couple more feet before finally coming down. That’s a missed shot and what’s known, of course, as a rebound. The ball is caught before it hits the ground by another player, and the player brings the ball up and down, possibly to secure it or jostling around looking for a teammate to pass it to.

Here is what the play looks like in reality. It’s an Anthony Davis missed shot from 20 feet rebounded by Evan Fournier. Height is actually only recorded for the ball and not for any players, so you get a two-dimensional map of players and a three-dimensional map of the ball’s journey.

For a two-dimensional map of the play, I provided something really simple below. This is only showing a two-second snapshot of the shot attempt and where every player stats (the red circle), their paths after that, and the ball’s path. You can see things like the pass from Evans, O’Quinn charging hard at Davis to close out, Asik and Vucevic going hard to the basket, the two Pelicans open on the right side, and Fournier caught in a weird place trying to cover Gordon at the last second.

sportvu first attempt example

That little graph is basic and shows a limited amount of information, but you can see functionally what the data is like and how quickly it’s tracking the court. Remember, that’s only two seconds, yet you can see the ball’s path in great detail.