NBA Week in Review: Back to the NBA

The NBA season officially began mere days ago, so it’s time to wildly extrapolate a team’s future based on three games and anoint new stars based on one or two good games. But seriously, this is our first look at an impressive rookie class and a number of new team setups. LaMarcus Aldridge is in San Antonio. The Chicago Bulls have a new starting lineup and coach. With the new season comes my first of week in review, covering a random assortment of topics from the general to the hopelessly specific and from basic basketball philosophy to statistical experiments.

The Latvian Giant in New York

There are many things intriguing about Kristaps Porzingis, but there’s something historically unique about him already: he’s maybe the tallest non-center in NBA history. Listed at 7’ 3”, he’s purely one of the tallest players in NBA history too, as few players his size or taller have tallied considerable minutes. The only contender is Ralph Sampson, who was part of the famous twin towers alignment with Hakeem Olajuwon[1. Most sources call Sampson the power forward, and some anecdotal evidence shows that Sampson usually guarded opposing power forwards.]. But both guys were known as centers even when they played together, hence twin towers. Porzingis has been labeled a PF, and so far that’s been accurate because all of his minutes have come with Robin Lopez at center.

Why does this matter? Being super-tall for your position while still being able to credibly defend and offer enough skill and floor-spacing is highly beneficial. It means Porzingis will be able to contest shots from further back and he’ll be able to shoot over his opponents on offense. Of course, he’s also young and position drift is a real thing, as Kevin Garnett used to play at small forward often and Kevin Durant was a 6′ 9″ shooting guard as a rookie. Once the tall Latvian adds more weight, he could cover more of the center position, but New York has already invested in Robin Lopez, who can definitely only play center. It’ll be interesting to see how the team responds to this and how quickly they move Porzingis to center full-time, if that happens.

As a side note, does anyone else find it odd that a good portion of the super-tall players (7’3” or taller) have soft skills, like outside shooting and passing (Arvydas Sabonis, Rik Smits, Ralph Sampson, Zydrunas Ilgauskas, Yao Ming, et al)?

The Mamba in the Room: Kobe Bryant and Adjustment

When a star scorer ages, there’s a bifurcation: either they adjust well and take on a smaller role, from Paul Pierce’s versatile use on playoff teams to Michael Finley as a spot-up shooter, or they refuse to take on a smaller role like Allen Iverson or Michael Jordan in Washington. Kobe Bryant hasn’t managed to stay healthy in a long time, but when he’s been on the court he’s been taking shots like he’s still 28 years-old and that hasn’t changed this season. The problem is that he can no longer maintain adequate efficiency for the shots he’s using. His free throw rate has dwindled, and his proportion of shots at the rim is abnormally low for a big scorer and it’s been declining for a while. Since he’s been chucking three-pointers and missing them, that leaves his midrange shooting.

As seen in these videos, his top-notch footwork and balance leave the crowd wowed, but you don’t receive extra points for style and one cannot subsist solely on a diet of those shots for a healthy offense. This also matters because the Lakers need to be developing their talent for the future, not existing as a vessel for a gunner adding to his gaudy totals. D’Angelo Russell is averaging a paltry nine points per game, for instance, and Jordan Clarkson has taken a backseat as well[2. I’ve always wanted to see Roy Hibbert on a good offense because I think he was overused as an option in Indiana with a usage rate near 20, explaining some of his shooting problems, but this is close enough. His usage rate has fallen to 14% due to the exploits of his teammates, and his efficiency has skyrocketed, as he has a 65 TS% now.]. Watching old man Kobe is perplexing, and there’s some mortal pain inside that pings with every game he challenges Father Time and fails.

Sanity Check: Statistical Leaders in Week 1 Last Season

After one week, the statistical leaderboards have some surprising names, but some of those players will drop out. For example, last year Joe Johnson was the third leading scorer in the league with Rudy Gay not far behind on an efficient 65 TS% and Westbrook was down the list with a paltry 20 points per game. Westbrook ended up leading the league in points per game, and Joe Johnson tumbled down the rankings – the world ended up making sense. But for teams it can be even wilder. Miami started the season 3-0 and ended up missing the playoffs, and a large number of eventual playoff teams had 0.500 records like the Spurs and Hawks or even losing records like the Bucks and Blazers. There are other weird occurrences like Houston, who should have been one of the best teams in the league, losing every game by (exactly!) 20 points and the Lakers leading the league in three-point rate. Obviously, there’s only a limited amount of information in two to three games, and large random events, like injuries, can change the landscape. So let’s not write anyone off yet.

Karl-Anthony Timberwolf

As the number one overall draft pick, watching Karl-Anthony Towns play during opening week is like unwrapping the biggest Christmas present. With two games, we shouldn’t focus too much on his stats[3. Towns is averaging 21/13/2 right now. I very much doubt a rookie right one year of his college under his belt can keep up those averages.], although they’ve been great for someone his age, but he does look like the multi-skilled center as he was advertised. In his second game, he scored 28 points against Denver, and it was the breadth of his scoring that was impressive. You can watch every field goal made in this playlist. In the second video, you can see his ability to catch-and-shoot from 20 feet with a defender nearby; he already looks like a capable outside shooter, which is quite valuable for a center. In the third video, he gets a pass in the midrange area and then drives to the basket, showing off plus ballhandling skills for a big man and finishing in traffic. In the last video, he’s used as a roll man well in a pick-and-roll with Rubio. He’s a multi-threat scoring center, which is rare, because he can catch the ball and shoot, drive, pass, or post-up.

Towns, however, was taken number one overall over guys like Jahlil Okafor and D’Angelo Russell because of his defense. He looks like a good rim protector already. He’s also quick enough to cover guards and there’s this wonderful possession where Joffrey Lauvergne tries to drive on Towns and gets stuffed, which leads to him being switched on Mudiay, who’s forced into a tough shot and a shot clock violation. He’s a really good rebounder too, and his skillset has an added value where he can mesh well with virtually any other frontcourt partner or lineup. As a number one pick, we expected him to be good, but he’s looked great awfully early.

Ricky Rubio and Shooting

Tracking shooting percentages through a paltry three games is next to meaningless, but in Rubio’s case there is something quite encouraging: he’s been efficient so far and he’s been playing like an improved scorer. For most wayward bricklayers, the three-point line is an obstacle – and for some the free throw line is. Yet for Rubio it’s been everything inside 16 feet. His free-throw shooting is good and his 3PT shooting is poor but not his worst feature. His shooting percentages from 3 to 10 feet, per basketball-reference, are usually below 30%, and he’s one of the worst finishers at the rim in NBA history. With a minimum of 100 attempts, in 2012 Rubio was the fifth worst finisher, but he was the worst in both 2013 and 2014. He also had an inside track on 2015 as well since he shot an amazingly low 31.7%, but it was in only 41 attempts.

So far, in a gargantuan sample size of two games, Rubio has been shooting well inside the arc. But he’s been shooting the ball in atypical Rubio fashion, as seen in this video when he hits a running three-pointer or when he hits a layup after getting bumped. He’s still a bright young point guard with a lot of skill, but his development curve has been flat since entering the league. If he makes substantial progress on his scoring ability, it’ll be hard for even the mainstream media to keep him out of all-star discussions.

Nikola Jokic’s Luis Scola Impression

NBA drafts are tantalizing but there’s a big delay between scouting a prospect and seeing him in action. Jokic had fantastic translated stats yet he was not seen as a lottery talent, offering some intrigue both as a player and as a case study. But now we finally get to see him against NBA talent. In Denver’s game against Oklahoma City, he scored over Ibaka in the post with some nice footwork and without every really leaving the ground. Then later he recovers a board and drives down the baseline for a layup against Steven Adams. If you want the vaunted big man you can throw the ball down to for a quick bucket, he can sling his hook shot against anyone. His defense could use some work, but there are some encouraging signs. He has outstanding footwork, and he hasn’t yet fully demonstrated his excellent passing and outside shooting.

Eerie Number Bias

There’s a strange number bias with NBA stats, and it’s only natural this is discussed right after Halloween. One may think that the leading digit for a number of stats should be some random number, 1 through 9, but it’s not true. The leading digit is much more likely a one than a nine or even a two, and this is consistent with a variety of measures even with modest sample sizes like a single season.

For example, looking at season point totals from 2015, the first digit is a one 125 times while it’s never greater than 70 with any other digit. There were only 29 eight’s. The pattern holds with rebounds too. There were 140 one’s for a leading digit, and that declined to 109 for two’s, falling down to 78 and then declining all the way to 19 with the nine’s. And of course this is true with assist totals. Putting the results together in the table below, the percentages follow a natural declining slope with a drop every step of the way.

Table: leading digits for season totals in pts, asts, and rebs, 2015

This isn’t a spooky coincidence and it follows a well-documented phenomenon: Benford’s law, which states that in naturally occurring datasets the leading digits are disproportionately smaller. Specifically, the percentage the first digit is a one is 30.1, then 17.6, 12.5, and then declining all the way to 4.6 — that matches the above table well. It was noticed by a man named Simon Newcomb in the 1800’s when he discovered that certain pages from books of logarithm tables[4. Before calculators, people needed to use tables of values for more complex computations.] were more worn than others, and found this pattern. It was later confirmed by a physicist named Benford.

There is actually no accepted full explanation of the law and why it works. The full range of this phenomenon is amazing. You can see it, for examples, with the listed height of skyscrapers … in different scales. In the NBA, I think this law makes sense: there are more possible paths to a low total than a high one because it’s simpler. For higher totals, you need both the minutes and the ability to collect the stats, and this pattern holds across a few magnitudes. Totals spread out as the numbers increase, and thus, you see more numbers starting with one’s than two’s and so on down the line.

There are actually some pertinent applications: fraud tracking. When people create numbers, they do so more randomly than what happens in reality, and the results do not adhere to Benford’s law. Lucky, the NBA follows this natural phenomenon, disproving the conjecture that the NBA universe is a simulation created by Gregg Popovich.