NBA Week 7 in Review: The Warriors Are Mortal

The week ended in a bang with one of the greatest games of the decade — a double-OT thriller between the undefeated Warriors and the surprisingly strong Celtics — and it led to a defeat, presumably from fatigue, at the hands of an eastern conference cellar-dweller in the midst of a resurgence with the Bucks. The league is becoming stratified, but trans-tier movement is still possible —  who will join the Cavaliers at the top of the east? The east, in fact, is holding its own against the west for the first time since 2009, and it’s thanks to a large middle-class (lots of good teams and few great teams.) And with that, what can we learn from last week?

Whiteside and Miami’s Defense

As most people expected, the Heat are much better this season than last. The nature of the improvement is more unexpected: they’ve been worse on offense and their defense has turned around, as their rating is at elite status level. The cause here is unusual. The Heat have been at their best with a rookie, Justice Winslow, on the court, and for the second season in a row hAbe been better with Hassan “prodigious shot-blocker” Whiteside off the court.

The numbers here are pretty confusing. Whiteside has some of the best shot-blocking numbers of the modern era coupled with elite defensive rebounding stats, but his macroscopic advanced numbers are disappointing. But more specifically, there’s no clear pattern in opponent shooting stats. Referring to the table below, opponents took fewer shots at the rim and had a lower percentage when he was on the court in 2015, but something’s different for 2016: opponents are taking more shots at the rim with Whiteside on the floor, even though Miami’s defense has improved by a huge amount. With some further investigation, you can see how opponents are shooting higher three-point percentages when he’s on the court, which is likely not his fault, but it doesn’t explain everything, as Miami is somehow over 11 points per 100 possessions better on defense without him. One interesting common thread is that Miami forces more turnovers by a significant amount without him. Whiteside’s shot-blocking numbers are incredible, but don’t be fooled there — blocks only accoubt for a small portion of defensive possessions. Like many players at the extremes, he hunts for blocks too often and doesn’t perform as well in space away from the rim.

Table: Miami Heat opponent stats at the rim, 2015

Table: Miami Heat opponent stats at the rim, 2016

Francs and Dollars

Here’s a fun player comparison: 59.2 TS% versus 59.2%, 18.9 TRB% versus 19.6%, 2.5 STL% versus 1.7%, 4.7 BLK% versus 4.3%, 12.0 TOV% versus 20.0, 16.7 USG% versus 17.5%, and 0.9 BPM versus 0.3. If you don’t like reading a long line of numbers, it appears player 1 has a slightly advantage, thanks mostly to turnovers. But the most relevant numbers to compare? $1.2 million versus $21.4 million. If you haven’t guessed it yet, player 1 is Clint Capela and player 2 is Dwight Howard. The Rockets used a lot of their resources to snatch a star in free agency in Howard, but it appears that time has faded into the past for him and we’ll only see flashes of what he was in Orlando. The team has been better on the court with Capela too, and he provides hope and young blood for the future for a team in distress. But names have a large power over human minds, and Howard will likely stay entrenched in his role on the team as he ages and deteriorates more and more.

A Setting Star

With Kobe Bryant on a farewell tour, there’s a large segment of the NBA fanbase calling for, or expecting, Kobe’s inclusion in the All-Star Game. The reasoning here is that he’s a legend and well never see him again — but the All-Star Game is near the two-thirds mark of the season, not the end, and it’s about representing the best players in the league. I realize a diatribe about an exhibition game seems pointless, but All-Star selections are typically used to measure player success, not to mention historical legacy. There are very good players out there who have devoted their lives to basketball and deserve recognition too. This is especially pertinent in the western conference where guards like Mike Conley or Eric Bledsoe have never been All Stars, not to mention Kawhi Leonard and Draymond Green. There’s not enough room for everyone.

Many great players have had final All-Star Game appearance during their final season, but these players either retired near their peaks like Bill Russell or had careers that were a step above nearly everyone else in NBA history, like Kareem being modestly chosen as an injury replacement for Magic Johnson in 1989. But many of the greatest players with comparable careers who played into their late 30’s and lost their effectiveness were sent off without a goodbye all-star game, from Kobe’s former teammate Shaquille O’Neal to Hakeem Olajuwon and guards like Oscar Robertson and Steve Nash. I don’t think the argument of “Kobe’s still a Laker and that matters” is relevant because the game isn’t in LA. What’s amusing is that Kobe Bryant could steal a spot from Tim Duncan, who doesn’t have much time left[1. Unless he’s actually a cyborg.] either and deserves a selection[2. However, I’m assuming he’d rather stay home.]. But this probably just another case of Kobe Bryant mirroring his idol, Michael Jordan, who had a large role in the last all-star game of his career as everyone commented on and appreciated his career. But he was a coach’s selection, and his last season was more productive than what we’ve seen from Kobe so far. The all-star game has a history littered with bizarre selections, and we don’t need to add another one to the pile based on an unsubstantiated claim about how we treat legends. There are other ways to honor them.

Westbrook’s Inbound Self-assist

On Sunday, Russell Westbrook employed a crafty playground trick of throwing the ball off a sleeping defender on an impounds for a pass to himself, and he neatly drained the jumper before the buzzer sounded for the end of the second half. But as someone who uses and creates stats, I’m fascinated by how this play is coded. In the play-by-play logs and in the SportVU assist database, this is a mere unassisted shot. But doesn’t SportVU track the pass? Who does it list as the receiver? Did he get a potential assist or a secondary assist? And I suppose I should throw a joke in here about how Westbrook is so ball dominant that he naturally needed a way to assist himself.

The Aging of the High School Generation

Tim Duncan and Kobe Bryant were once both peers as elite NBA players at roughly the same time, but their paths have diverged where Kobe has has had trouble staying on the court or, when healthy, hitting a shot and Duncan has aged as well as anyone in NBA history. Kobe’s also one of the first high school players of his generation along with Kevin Garnett, who has settled comfortably as a limited minutes role player. Thus, it’s natural to see the beginning of an argument here that coming into the league early shortens your career, but it’s more complicated than that.

There’s a selection bias here with the subset of high school players. They’re more likely to be athletic, physical wonders who enter the league early, and those players generally peak earlier than other players. There’s also a confirmation bias that can come into play because people like to focus on the high school players who failed and forget the hundreds, or thousands, of players with college experience who failed just as magnificently. It’s tough to quantify how differently high school draft picks have aged, but in a few years we’ll have a full picture — we just can’t hastily jump to conclusions without smart adjustments.

Miles in Indiana

Indiana has been excellent this season and one of the most improved teams, but it’s not just about the return of Paul George; they’ve had a host of other useful contributions. For instance, the unheralded and underpaid CJ Miles has had his best season so far to date, thanks mainly to his shooting. Throughout his career, he’s been defined by his outside shot, which has provided a great deal of his value. He has a nice lefty release, and he’s shooting 43% so far on the season. That should fall back closer to his career average of 35% as the season progresses, but his outside shooting definitely has to be accounted for by opposing defenses. But he’s not purely a catch-and-shoot player either. He still has a lot of athleticism — here’s an example where he dunks high over Andre Drummond in transition. He’s not the best ballhandler or someone you want to run a lot of plays for, but he drives fairly well and he’s become a capable scorer for Indiana.

CJ’s athleticism can translate to defense too, like this giant block he had against Miami. He can accurately be described as a 3/D player, though he’s never been known for shut-down defense. His box score stats on defense are pedestrian, but his more advanced numbers, like RPM, are usually a little above average[3. I’d ignore his RPM stats on defense for this season somewhat because his raw on/off stats look terrible on defense, and that’s likely an anomaly. They were much better with him on the court last season, and he’s not a major negative on defense.]. Most stats and scouts agree here, and what’s important to note is that he’s a shooter who definitely doesn’t kill you on defense, which is so often the case with cheap shooters. Indiana has looked like one of the better teams in the league with this new generation of players, and Miles is a key component.

Twenty Game Win Streaks [4. I’m ignoring the Washington Capitols streak of 20 games back in the late 40’s because it was pre-shot clock. Plus it spanned two seasons, and using two separate seasons for a streak really irritates a few people for some reason — something about the purity of a superfluous stat.]

With the Golden State streak in the books, it’s naturally time for a little reflection and some thought about how the accomplishment sit in NBA history. While looking purely at wins is the only “right” way to do this for most people, it’s worth slicing up the data and looking more closely at the data to get a better picture at what happened. In the table below, I’ve included a summary of the major streaks with margin of victory and an estimated of SRS, which is roughly adjusted for opponents and homecourt advantage[5. I used a conservative value of 3.2 for HCA, which is most likely a gross under-estimate for 70’s basketball.] The odds are calculated using this page on a game-by-game basis.

Table: NBA 20 game win streak details

The first observation here is that the streak from the Bucks is underrated. They crushed teams on a per game basis more than even the Lakers, and their numbers fell less when adjusted for competition more than anyone’s. Total margin of victory is probably a good barometer of the power of a streak, and they’re second just behind the Lakers. The Bucks, by the way, won the title that year with a combination of a young Kareem Abdul-Jabbar and an older Oscar Robertson.

Houston is the oddest inclusion because they didn’t win a title and weren’t the favorites like the Warriors now. They lost Yao Ming mid-streak and even made a trade too, but they stretched it to 22 games for what’s possibly the most inexplicable streak of its magnitude. Tracy McGrady was running on fumes and they had a gaggle of odd role players, yet for one part of the season they went 31-3, like they were the ’96 Bulls at their best. Miami had smaller odds, but it’s more believable that roster would win 20 plus games than Houston’s.

If one includes the last four games of the 2015 season, then Golden State rises to second in the win streak rankings. People are loath to include games from another season, even though most lists from official sites do so. Nonetheless, it’s an impressive streak, and it doesn’t appear to be a “weak” one by different stats.

Miami was celebrated in 2013 for its ultra-long streak, but they were probably the weakest team on a per game basis. They were one of the “clutchest” teams ever, especially during the streak, which accounted for the long string in wins. They weren’t bad by any stretch of the imagination, but it’s worth noting they were about as good as the Warriors last season in terms of adjusted point differential.

The Lakers had a serene combination of Jerry West, Wilt Chamberlain, and Gail Goodrich. For a modern connection, the team also had Pat Riley, architect of the Miami super-friends team and participant in two of the longest streaks ever. The Lakers, and the Bucks the year before, were two of the greatest teams ever with absurdly large point differentials. The fact that these seasons were adjacent to one another says something about the era they were in — there was a lot of expansion. Interestingly, these streaks have come in clusters, and I’m not sure what it says about this current era. There have actually been three streaks of exactly 19 games since 2009, and two in the past couple years, so it’s not about an arbitrary cut-off.

There are few things that warm up the league in December like a win streak. They signify some combination of dominance and supreme luck/clutchness. It’s a great sign for the Warriors, obviously, but they’re firmly entrenched in history now and their doubters are silent. It was the greatest start to an NBA season, and it’s a sign of things to come.