Nylon Calculus Week 6 in Review: Our hallowed Hall of Fame

We’re about a quarter of the way through the season already, and except for fans of a few disappointing teams I would think there are few complaints. It’s been an entertaining season so far, and there’s still so much potential brewing out there on the horizon. The Golden State Warriors are obliterating the league, and we’re primed for one of the better MVP races in recent history.

But let’s take a step back first: what happened last week?

Westbrook’s path of destruction

Russell Westbrook has had six triple doubles in a row, and as of this week he’s averaging a triple double for the season. This is unprecedented for the modern NBA era, and it’s akin to, say, averaging 0.400 in baseball. The season-long triple-double is associated with the legendary Oscar Robertson, and it’s one of the most impressive and enduring per game stats in NBA lore. Yet it’s one of the most overrated too, unevenly dominating the conversation about Oscar himself. He was more than just a stat-line. It’s about arbitrary milestones that ultimately are set because of our number system — there’s nothing better about 10-10-10 than 10-12-8.

Read More: Best dunk, assist, block, and buzzer beater from the month of November

The obsession about these numbers can override real basketball discussion, but at the very least we can steer this obsession into useful waters. Westbrook doesn’t have the same rebounding might as big men; he’s been stealing the easiest boards, the uncontested ones, and this sort of stat-padding has been a common feature of his career. But it’s also led to more transition opportunities for the team as a whole, as their offense is healthier when he grabs the board. That’s actually the mechanism behind a theory I had on the offensive prowess of Oscar Robertson’s teams — it’s a huge advantage to start your offense immediately with the ball in your star guard’s hands, especially when half-court offenses are limited. But that’s something we can appreciate without Westbrook hitting 10 rebounds a game.

The Clippers’ strange week

The Clippers have fallen back to Earth over the past week or so, which was punctuated by a wild double-overtime loss to the Nets. It was a weird game where Sean Kilpatrick was a star, Doc Rivers went berserk, and the Nets let a four-point lead with four incoming free throws go to waste in the waning seconds thanks to three missed foul shots and a Jamal Crawford 3-pointer. The Clippers need every win they can get too with how competitive the top of the conference is. If you slip to the four-seed, then you have to face the Warriors in the second round. They’re still first in defensive rating though, and I doubt that’ll hold. Since November 11, they’ve actually had a defensive rating of 109, which is below average. I don’t think that’s their true level, but first in the league isn’t either — and it’ll cost them some good seeding.

Mike Conley and the tribulations of the Memphis Grizzlies

The Grizzlies once again are marred by injuries, as Mike Conley suffered a transverse process fracture to one of the vertebrae in his lumbar spine — in other words, a broken back. He could be out for six to eight weeks, which is an outright shame for a guy who’s having a career season after signing the most lucrative contract in NBA history. The Grizzlies have been cursed with injuries recently, from the deluge of injuries last season to the problems with Chandler Parsons’ health this year. The team is in such disarray that they’re leaning on heavy minutes from Half-Man, Half-Senior Vince Carter. It’s disappointing because we didn’t get to see much of the team intact last season and this season we get a healthy, 3-point-shooting Marc Gasol.

What’s most distressing for the team is their lack of a backup or replacement options for Mike Conley. There’s the rookie Wade Baldwin, and then the closest option behind him is combo guard Troy Daniels. They’ve actually gone several games without any natural point guard starting; instead they’ve had backcourts with Tony Allen and Andrew Harrison or Troy Daniels. The team has obviously expressed interest in available point guards, but it’s highly unlikely they’ll unearth someone who will help them win games — those guys are truly, and by definition, replacement level. Memphis will win a few more than expected thanks to their close game voodoo, but they need Mike Conley back.

But hey, at least they still have Marc Gasol.

https://twitter.com/SBNationNBA/status/804508025495162881

Barkley and Amazonian Warriors

One of the most enduring subplots to the NBA-media world has been Charles Barkley’s abject hatred of the Golden State Warriors. Last week he went over the line in his criticism, calling their style “girly.”

Barkley’s tirades are getting tired, and his schtick of talking bluntly and often ignorantly has lost of a lot of its appeal. I also don’t know why he should be the final arbiter of title-seeking teams either, since he infamously never won a championship and his team-hopping ways were the most egregious, forcing his way onto a successful Kevin Johnson-led Phoenix Suns team and then later one with Hakeem Olajuwon and Clyde Drexler. People still love him for his bluntness, but his unfiltered mind has few insights worth listening to.

We can do better than this.

Trick shot magic

There are no extra points given for degree of difficulty, despite the protests of Kobe Bryant fanatics. In fact, this is why players like Kobe or DeMar DeRozan can be overrated not just by the media but by their own peers. Professional players know as well as anyone how difficult some of their moves and shots are, so they, consciously or not, rate those players higher. But a point is a point is a point. It’s too bad, because occasionally there are wild field goals that deserve a few extra points on their own, like the two in the below clips. The Steph Curry shot counted, but the Isaiah Thomas one did not — and both were so difficult you wouldn’t even want to seriously try them in a game of HORSE.

Minnesota’s stumble

The Minnesota Timberwolves are sloshing around in the bottom of the Western Conference, which is derailing preseason prognostications about the young team’s ascent. In retrospect, a team that had just won only 29 games with a -3.5 point differential and no major personnel changes should not be expected to make a major leap forward. It’s why stat-based projections were generally so pessimistic. Tom Thibodeau’s defensive magic hasn’t permeated; they’ve one of the worst teams in the league at that end of the court.

Read More: Timberwolves are struggling to execute Thibodeau’s signature defense

However, the team actually has improved, and most of the improvement has been on offense. They’ve won only six games, but they have the point differential of a team with nine wins — extrapolated over the full course of a season and that resembles a team with roughly 36 or 37 wins. They’re 1-5 in games decided by five or fewer points; that’s not a trend that typically holds.

Minnesota still has its issues though. Andrew Wiggins has to evolve beyond the one-dimensional scorer mold, and he’s gotta develop some legitimate defensive power. I can understand why people praise a guy for being such a high-volume scorer at a young age, which is indeed rare, but it’s not as valuable as a multi-tool star. Reinforcing the belief that just scoring a lot of points is fine on its own won’t sway his, or any future player’s, opinion on developing further.

Finally, I’ve seen some blamed targeted at Ricky Rubio, but he’s been one of the few bright spots they’ve had over the past few years. And the idea of replacing Rubio’s shooting blight at point guard with a college senior noted for a shaky outside shot and a career FT% under 70% has not looked too bright lately. Dunn has awful shooting percentages, and Rubio’s terrible raw +/- stats this season are an aberration, not a final verdict. Raw +/- is poor statistic, which is why +/- needs advanced math and multiple seasons of data to form lucid conclusions. The Timberwolves are a mishmash of disappointment and hope, of spurts and fits, like any other young team. They won’t be great yet, but the future is still there.

Hall of Fame odds

Hall of Fame discussions have been trickling through the league at certain points, as we appreciate the 2016 class and talk about the chances of current players like Vince Carter. I’ve approached this topic before with a discussion on Allen Iverson; it’s always something that eats away at the back of my mind as people discuss the great players. Unlike the MLB, there are few crucial milestones we can use to compare and compile players historically. One that’s frequently used, unfortunately, is the number of All-Star games, which creates a host of issues. You can see this problem clearly with a handful of cases, but the real problem is the number of cases that were unconsciously subverted by an unfair adherence to All-Star games or other dubious milestones.

All-Star games are a terrible measure for player greatness for a few straight-forward reasons. Firstly, it’s an award based on roughly just half a season, and many votes are cast early in the season. Secondly, fan voting has obviously led to many poor decisions, like when guys are chosen who haven’t even played ten games and can’t even suit up for the game. Then there’s the issue of the conference disparity, where making the team as a forward in the early 2000’s in the West, for example, was much more difficult than making it as a guard in the East in the same period. Finally, selections are largely made by position, and combined with the conference problem you can get some weird inclusions, especially when center was its own position. Both Kevin Duckworth and Terry Porter made two All-Star teams, but they were not equal to Portland’s success in the early 90’s.

The shining example here is Joe Johnson, a steady player with seven All-Star Game selections, despite the lack of any magnificent stretch of his career where he could be considered one of the best players in the league. He was never top ten in MVP voting, and he was all-NBA only once and that was the 3rd team. His career stat totals are notable, but that’s through sheer volume and he only once averaged more than 22 points per game. Known for primarily his scoring, that’s damning, and the fact that he’s never won a title should close his Hall of Fame aspirations for good. But why are his odds 51 percent, according to Basketball-Reference?

Looking at the formula, the culprit is the All-Star Game variable. The other factors — height, titles, top-ten leaderboard appearances, and peak Win Shares — are negligible to neutral for Joe Johnson, but his All-Star appearances alone are enough to push him into likely Hall of Fame territory. Other players with seven selections include Scottie Pippen, Alonzo Mourning, and Willis Reed — there are some heavy hitters in that group. The only other players in that group who aren’t in the Hall of Fame don’t qualify yet except for Jack Sikma, one of the better borderline cases out there.

For a clearer illustration of how Basketball-Reference’s odds vary with all-star game selections I prepared the graph below. Each line uses the career stats of the specified player where All-Star games are varied along the x-axis (the triangle is based on the player’s true number of games.) You can see how Chris Webber’s career is actually comparable to Johnson’s, except for the number of All-Star appearances. You can see how Yao Ming’s case is helped appreciably; he was selected every year he was in the league too, even as a rookie and when he played a mere 5 games. The model loved that fact, and ignored his special circumstances. A player who accrues even two or three undeserved selections can dramatically change his odds.

If you saw that height was a factor in the odds, you may wonder why and how it affects the results. Yao Ming actually was penalized heavily for his gargantuan height, and while he had a few unfair all-star selections to make up more than enough ground it does hurt the candidacy of other big men. Fun fact: Yao Ming had 60.8 regular season games per All-Star Game. The All-Star Games themselves are supposed to be a proxy of star-level dominance, but I’ve always thought another simple stat was better: MVP shares, which show the proportion of the total possible vote a player receives.

Building a similar model to the one found on Basketball-Eeference, I settled on four variables: MVP shares, number of 1st-team all-NBA pre-1981, number of 2nd-team all-NBA pre-1981, and total Win Shares. This was done using every player with at least 400 games from 1956, which was when the MVP award was introduced, to 2011. MVP shares show dominance and whether or not a player was considered one of the very in the league, which is a common description of what a Hall of Fame player is. MVP shares only use first-place votes before 1981, so I supplemented the model with all-NBA teams; they’re a better proxy for peak-level dominance. The 3rd-team all-NBA has only existed since 1988. Win Shares were used to boost guys who were very good for a long period of time — these are the longevity cases, like Paul Pierce or Robert Parish. I’d rather use a different metric, but for now this is pretty much the only all-in-one metric available going back to the 50’s. You can play with the formula below.

100/(1 +exp[-(-6.06 +4.81*MVP Shares +1.34*1stTeam(pre-81) +1.40*2ndTeam(pre-81) +0.0494*WS)])

The players with the highest odds are (in no order): Kareem Abdul-Jabbar, Michael Jordan, Wilt Chamberlain, and Bill Russell. Remember that this doesn’t count recent superstars, like LeBron James; but that’s a decent ordering for the best players ever. The model’s greatest misses at one end of the spectrum — this is for players with the best cases statistically who didn’t make the Hall of Fame — were Paul Westphal, Sidney Moncrief, Marques Johnson, and Chris Webber, respectively. Those guys are known for being Hall snubs, especially Moncrief. At the other end of the spectrum were Ralph Sampson, Guy Rodgers, Wayne Embry, and K.C. Jones. Some of those guys, in particular Sampson, had wondrous collegiate careers too, which bolstered their cases, but there’s another factor I failed to include and it’s apparent with K.C. Jones, as well as a few other players with high errors like Frank Ramsay and Tom Sanders: rings were not part of the model. You can see it too with the other side of that spectrum, as Westphal, Moncrief, Johnson, and Webber all failed to win titles.

Listen: Meet the crazy outlier NBA players who inspired rule changes

I could haphazardly throw in total rings for every player, but I think this deserves a more careful treatment — and thus this little study will just be part one and I’ll finish it another week. I don’t want to equate Bill Walton’s magical title run with, say, an eighth man picking up a championship thanks to his teammates. But in fixing the veracity of the model, in a way I’m moving further away from the truth in how people pick Hall of Famers, because the flawed system of using stats like All-Star selections or ring counting can influence opinions and distort memories. I can explain why Joe Johnson’s odds are inaccurate, but then I unravel a tricky thread and have to deal with the consequences of retroactively grading previous selections. I don’t know if I can in good conscience build a stat that aims for pure accuracy if it perpetuates historic issues. Sometimes the best model is the least accurate one.