In the previous two iterations of this series, we took a look at PER and WS/48 in an effort to determine what constitutes a “league average” NBA player—that is, a guy who is basically a 41-41 team over 82 games (man, remember 82-game seasons? Such nostalgia) distilled into one player across one season.
Today, we take a few steps back and consider not just the VORP per 82 games of the guys we’ve considered so far, a cross-section of players from the 2019 and 2020 seasons who had league-average PER (between 14.8 and 15.2) or WS/48 (between .095 and .105) while playing at least 1000 minutes, but the overall VORP formula, what we’ve learned about its correlation with team wins, and what that means to a 41-win team trying to distribute the minutes across the 19,680 minutes available (not counting overtime) to players at five positions for 48 minutes over 82 games.
For the team analysis, we’ll be sticking to 2019 data for this because COVID made everything weird last year in terms of comparability and completely broke efforts to compare teams around the league.
On average, since the formula for win expectation is 20.34+2.07x, where X is total team VORP across the entire roster, that makes expected team VORP just about exactly 10. It’s actually 9.98, but 10 is a round number that divides easily by 5 (two VORP per position per 3,936 minutes of an 82-game season.)
So let’s take all the guys from our previous research and see if their VORP lines up with their expectations based on their PER and WS/48.
We’ll start with Jayson Tatum‘s 2019 season, since he appears on both lists.
Tatum had 1.1 VORP in 2,455 minutes played in 2019, his age-20 sophomore season in Boston. That gives him 88 percent of the production (1.76 VORP per 3936 minutes of a full season playing 48 minutes a game) of our hypothesized completely league-average player.
Given that we allowed a 2.6 percent swing in PER and a 10 percent swing in WS/48, a 12 percent swing below normal is a bit high, but VORP is also rounded to just one decimal place (and I’m not going to calculate it myself on a deadline, just work with me here) so that is absolutely close enough.
But let’s back-burner “Sophomore Jayson Tatum is our hypothetical perfect league-average player” for a minute and look at the rest of the data.
There’s Reggie Jackson‘s 2019 season to consider as well. At age 28 in Detroit, Jackson posted a 15.2 PER and .105 WS/48, slightly on the high side of the expectation for league-average.
His VORP? 1.4 in 2289 minutes, or fully 2.41 VORP per 3936 minutes. That’s just over 20 percent above expectations for the average player on a truly mediocre team, while playing for the very definition of a “truly mediocre” team, a Pistons squad that went 41-41.
How about Jalen Brunson? His 2020 season featured 14.9 PER and .101 WS/48, almost bang-on even with the hypothesized league average.
His VORP? 0.3 in 1022 minutes, or 1.16 VORP per 3936 minutes. That’s miles below what you’d expect from the average player on a 41-win team, and Brunson was on a Mavs team that went 43-32 in a shortened 2020 campaign.
But here’s the thing. Brunson was a bench player. Keep this in mind as we move forward.
At the last, we have Derrick White. He lives in that no-man’s-land between starter and bench player. In 2019, he started 55 of the 67 games he played, but he only played 25.8 minutes per game, a role he has largely continued in San Antonio in the 2020 and 2021 seasons, starting less often last year but still ultimately getting his 25 minutes or so in.
He posted a 14.8 PER and .101 WS/48 in 2019. But his VORP/3936MP was off the charts at 2.96, a whopping 48 percent above “league average”!
So there doesn’t seem to be a lot of correlation between player role and prorated VORP. Part of that is rounding (where between 0.25 and 0.34 was Brunson’s VORP last year?), but part of that is the simple fact that three advanced stats are not necessarily all measuring the same things the same way—we haven’t even brought in stats like the various iterations of pace-adjusted, minute-weighted plus-minus (BPM, RPM, PIPM, and the extra-crispy versions that teams keep to themselves.)
But part of it is that it’s fiendishly difficult to find in any given sample a player who’s perfectly mediocre in every way you can possibly measure mediocrity, that nebulous sense of “the average player on an 8 seed” that is, by definition, the exact median of NBA production.
But in 2019 we came real darn close; Jayson Tatum at age 20 was just slightly below average as the league figured him out, but he made a quantum leap forward in 2020, making his first All-Star team, and he seems an absolute lock to be an All-Star again in 2021.
And Reggie Jackson, the ultimate journeyman’s journeyman, put up a slightly above average (rounding and context!) season by those same metrics.
I don’t think it’s fair to Tatum to call an age-20 season, an age when a lot of players are still in college who then go on to have decent NBA careers as, well, league-average players, “average” by any stretch. He was on the cusp of greatness; he just needed to learn how to handle his business on a good team.
But Jackson? Yeah. I think if you told me the average NBA player is a veteran late first rounder who packs a decent per-36-minute scoring punch, is an OK shooter for his position, and dishes a few assists as a guard or grabs a few boards as a big man (not for nothing did Myles Turner show up on the fringes of this study, for example), is nobody’s idea of an All-Star, but gives you all this on a relatively team-friendly contract under the modern cap (Jackson made $17 million in 2019), I’d be willing to believe that.
So congratulations, Reggie Jackson. You are, and I mean this in the nicest way possible, average.