NBA Week 2 in Review: The Rising Dawn
By Justin
The second week of the NBA brought dizzying highs like the Warriors storming through the league and terrifying lows like the sorry state of the Nets and a string of losses. The NBA season is fully engaged and storylines are already forming and worrying fanbases. With a quick look through the week, let’s tackle a few random topics and investigate some early trends and observations.
Memphis Grizzlies in Ruins
The Grizzlies were famously poached in California, losing by an incredible 50 points to the Warriors. This early in the season, everyone has the “small sample size” alert sign on, but this is different than some hand-wringing over a 1-3 start: losing by 50 points is a truly awful sign, even if the Warriors are playing at an all-time level. The power of point differential is that it gives a sense of magnitude between two teams. Blow-outs by that large a margin are either rare and bizarre or an indication that the two teams are on vastly different levels. There are only a small number of 50 point wins in NBA history, and note that teams like last year’s 76ers and the 11-win Mavericks from 1993 are on the list. The only playoff teams on that list are generally weak ones — the most successful one recently, for example, was probably a Boston team with 44 wins and a negative point differential that reached the second round due to a weak Eastern Conference in 2003; the most successful in history was a Bullets team that reached the Finals with 42 wins, but that was back in 1971 and the Western Conference right now is a lot more competitive.
Memphis was blown out by Portland too a few days later, though it was by a more typical margin. But they were destroyed by Cleveland by 40 points in their first game, and that was at home — that’s a very distressing signal because good teams rarely get beaten like that and Memphis is pretty healthy right now. For a hopeful note, the team has been torn apart by shooting percentages, and those usually regress to the mean pretty hard after a long period of time. For example, after the first two weeks opponents shot 42 percent from behind the three-point line … and the Grizzlies shot a ghastly 27 percent. They aren’t quite that bad, but the team might be slowly grinding down and a watchful eye is needed here because the west is loaded and playoff seeds are reserved swiftly.
Willie Cauley-Stein
As a mid-lottery pick, there wasn’t too much fuss made about Willie Cauley-Stein, but through week two he’s looked like the best rookie, according to a number of stats. Most obviously, Willie already looks like a plus defender with his huge seven-foot frame and agility. He stalks the court and can block a shot from almost any angle, like this block where he swats Courtney Lee from behind on a fast break. He’s got great timing, but partially due to the presence of DeMarcus Cousins he’s pushed out to the perimeter and he actually has more steals than blocks. That, however, also says a lot about his quick hands. But most surprisingly, his offense hasn’t been a liability because he’s shot such a high-percentage from the court. The talk of the NBA now is all about spacing and three-point shooting, but that’s just one way to attract, or distract, defensive attention; another is being a dangerous target for catching it near the rim, like a Tyson Chandler-esque player. Cauley-Stein has that potential, and so far Sacramento has played quite well when he’s on the court.
Boston’s Subtle Start
The Celtics got some attention before the season for some optimistic win projections from a few analytical models, and so far I imagine critics are using a heavy dose of confirmation bias to write them off before Thanksgiving. But the results so far have been actually positive. They’ve been hanging around 0.500 ball, but their margin of victory so far is +3.1, which is a more stable measure of team strength. For example, they recently beat the Wizards by 20 points, and most traditional analysts would rank the Wizards above them.
Looking closer at the team, the Celtics are slumping, and Isaiah Thomas in particular has had issues with scoring. But these problems are ones that usually recover quickly: the team is missing a lot of its shots, especially outside ones. They’re shooting under 30 percent from behind the arc, and since this isn’t the 1980’s I would expect that to recover. More specifically, Thomas is far below average on shooting efficiency and historically that aspect of his game has never waned, even after being shipped around the league like scrap metal. Noting that the star-less environment in Boston is too tough isn’t correct, by the way — he scored at one of the league’s highest rates last year in Boston and succeeded in a tough situation in Sacramento. Some of his key stats, like free-throw rate, steal rate, and assist rate, are on track and actually a little above his career averages. The team has also been without Marcus Smart for a few of their games, and he’s a vital part of their power. The team should be a little bit better the rest of the season based on past results, and that’s enough to snag a decent seeding in the east. Don’t count them out in November.
Longest NBA Shot Ever
According to basketball-reference, the longest recorded shot since 2001 was 88 feet from Baron Davis[1. The linked article only goes up to 2013, but updating the data no one has usurped his crown.]. But this has been surpassed: Jae Crowder just nailed a 94 foot shot last week. Unfortunately, he was standing out of bounds and thus it did not count. In light of this recent news, I propose these shots should count and they should be worth, say, ten points to make it interesting. We don’t want to miss out on a play like Crowder’s again. But that’s what I would term a rare event fallacy. In viewing the league like that, you would forever be chasing past events while new ones elude you because the ones you’re targeting happen so infrequently. Instead we should enjoy the NBA for all its quirkiness and await the inevitable 90 foot, falling out of bounds Stephen Curry three-pointer.
Justise Winslow
The Miami Heat have been a fortunate team the last few years, and that string of luck continues even after LeBron’s leave. Justise Winslow looks like a great tenth pick in the draft, and at 19 years-old he’s already contributing. He’s known for his defense first and he definitely has an NBA-ready body on the wing. At the beginning of this clip, you can see his quick hands in creating a turnover, and in another clip you can see him breaking up an alley-oop. On offense, he’s an okay shooter who needs a better touch from outside, but he’s strong enough to post-up inside and score and drive to the basket and finish with contact.
Basketball Death Tandem
Chemistry is a term applied to basketball constantly, and as a scientific reference this makes sense. Certain combinations of players are more powerful than others. Conversely, some combinations are deadly, and it makes basketball more interesting to analyze and watch. One duo I worried about was Lance Stephenson with Austin Rivers on the court too. They’re surprisingly similar players in that they’re slashers and combo guards who can’t really shoot and can, theoretically, play sound defense. But they’re not great off-ball players, and Austin Rivers is usually a poor basketball player even with the ball in his hands. Playing them together makes little basketball sense — forget about any advanced measures.
Of course, so far the duo has been an anchor made of depleted uranium in limited minutes. Austin Rivers is infamously known for having one of the worst rookie seasons ever and being a player who found the right GM/coach. Lance Stephenson, meanwhile, is continuing a dreadful shooting slump he had last season where he’s not only missing the net, the rim, but now he’s missing the entire backboard. Hopefully, the dubious duo will not find much time on the court together this season, but they’re both ballhandlers who could see a lot more action when Chris Paul rests, which makes for a crowded bullpen with Jamaal Crawford still there and Josh Smith in the frontcourt, another wayward shooter with no shyness. Perhaps the bench will coalesce into something useful and talent will win out, or perhaps the four aforementioned players will jack ill-advised shots until the court sinks into the ocean and that Clipper name will finally be pertinent.
Deferred Rebounds as Proxy for Team Rebounding
As the season began, the densely packed NBA stats site revamped a few sections, including SportVU rebounds. One major addition was the deferred rebound chance, which is when a player is within 3.5 feet of a rebound that was picked up by a teammate. This was introduced as a way not to penalize players for missing rebound chances because the point is that the team gets the rebound, not the player. However, it may also be a proxy for team rebounders, i.e. players who add to the bottom-line without gaudy individual rebounding stats of their own. Grabbing the rebound isn’t as important as making sure your team secures the board.
To test the utility of this new stat, I’m using one of my favorite data sources, a plus-minus like model with rebounds instead of a team scoring margin, to tie a few key individual rebounding stats together. Essentially, the model measures how well a team rebounds with specific players on the court. For the individual stats, I experimented with a few you can access here as well as the standard defensive rebound percentage.
Based on a preliminary investigation, deferred rebound chances have a small but positive relationship with team rebound rates. Unfortunately, the adjective “small” has to be emphasized clearly — the r^2[2. The variation explained by the model.] increased by a mere 0.3% and the added variable did not statistically significantly improve the model. This was done with only two other variables: defensive rebound rate (proportion of available defensive boards grabbed) and contested rebound rate[1. This was just defensive rebound rate multiplied by the proportion of contested boards to total.]. For some useful information, contested defensive boards were roughly two times as valuable.
Referring to the graph below, there’s almost no pattern to deferred rebounds versus the model results, which include defensive rebound rate and contested boards. Having a positive differential on the y-axis means a player is having a greater effect on his team’s rebound rate than his stats indicate, and a flat trend all along the x-axis means the relationship is virtually nonexistent. This is true of many iterations that were also attempted.
The above results aren’t a complete condemnation of deferred rebound chances. They did positively correlate with team rates, and this was only using a limited one-year set of data. Integrated more intelligently with traditional stats, there’s a slim improvement. But it’s not a magic elixir to our stats. There’s a fundamental mechanical issue here: being near a teammate rebound doesn’t mean one would have grabbed that same rebound. A lot of perimeter players have high deferred chance rates, but they’re not as likely to grab those rebounds and other helpful activities, like boxing out, are not covered by the deferred chance. What’s surprising is that an indirect stat like this was given but not something more obvious like number of rebounds grabbed with a teammate but no defender nearby — that stat, a “stolen” rebound, might be even more useful.