As the regular season nears its conclusion and with the play-in tournament just days away, it's time for my year-end edition of the NBA mailbag.
Throughout the NBA season, I answer your questions about the latest, most interesting topics in basketball. You can tweet me directly at @kpelton, tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.
This week's edition of the NBA mailbag focuses on end-of-season awards:
Is there a formula that predicts MVP? What about other awards?
Which past award result would I change?
Is the Coach of the Year award a trap?
"What are the criteria for winning MVP?"
-- Kevin D., Brooklyn
This question is inspired by Brooklyn Nets star (and 2013-14 MVP) Kevin Durant telling ESPN's Rachel Nichols last month that he doesn't care about who wins MVP because, "I feel as though the criteria changes every day." I get why Durant feels that way, as I do think the explanations for why reporters vote the way they do for MVP tend to change from year to year and even from day to day as the debate plays out over the course of the season.
However, I think when we look at how voters vote rather than what they say, the way MVPs are chosen is remarkably consistent. In fact, about three-quarters of the time I think you can predict the outcome based on two factors: team record and the player's rank in an all-in-one metric. (I'll use my wins above replacement player, aka WARP, for this task.)
There have been 40 MVPs since voting for the award went to the media in 1980-81. Of those, 29 have gone to the player with the lowest combined ranking in terms of individual WARP and team winning percentage. Another three (Dirk Nowitzki in 2006-07, Kobe Bryant in 2007-08 and Derrick Rose in 2010-11) went to the player with the second-best MVP score. That leaves just eight MVPs in the past four decades as outliers by this formula:
A more advanced formula that accounted for the extent to which players dominated one category rather than simply going by their ordinal rank would probably show some of these MVPs to be predictable too. Of the past 40 MVPs, just three players -- Bryant, Iverson and Nash's second win -- did not rank first in either WARP, team record or the combination of the two.
Here's what the formula says about this year's MVP race:
I doubt that will predict the actual voting, as Joel Embiid (who lags in WARP because of the time he has missed) and Stephen Curry (whose Warriors have the league's 14th-best record) will likely jump many if not all of the players behind Jokic. Still, the odds are good the formula will again pick the MVP winner.
How about other awards? I wrote back in 2014 that Rookie of the Year could almost always be predicted by taking the rookie with the highest combination of points, rebounds and assists per game. Since then, that formula has "failed" twice: Joel Embiid in 2016-17 and Zion Williamson last season, when they played a limited number of games. So we'll update that to a minimum of 40 games. That does leave us with 2016-17 as our only failure since Amar'e Stoudemire over Yao Ming in 2002-03: Malcolm Brogdon (17.3 combined PPG, RPG and APG) beat out Dario Saric (21.4) for the award.
This year, that formula gives LaMelo Ball the edge over hard-charging Anthony Edwards:
Looking at other individual awards, I was surprised to find out that the Sixth Man of the Year Award isn't automatically the highest-scoring eligible reserve. PPG has predicted just 20 of the 38 winners. Adding rebounds and assists or team success didn't help me do any better, suggesting that voting for this award is a little less predictable. Unsurprisingly, that makes Jordan Clarkson (18.2 ppg) this year's favorite.
For Defensive Player of the Year, the formula that worked best averaged the player's rank in combined blocks and steals per game (aka "stocks") and his team's rank in defensive rating. This still identified just 16 of the 38 winners with another four finishing second. We'll likely add another hit as this year's front-runner is tied for first with a more unlikely candidate.
As for Most Improved Player, my preliminary research didn't turn up anything that identified more than six winners as the most likely contender that year. I'm chalking that up as more unpredictable than any of the other player awards.
Coach of the Year appears to be best predicted by a combination of team record and the extent to which a team outperforms its preseason line over the 15 seasons for which I have those over/under totals, although that combination still predicts just seven of 15 winners in that span. Three other Coaches of the Year finished second in that rating.
This year, that history would favor Monty Williams. I would expect the top three here to finish atop Coach of the Year voting, although probably not in this order:
MORE: Boom or bust -- How each play-in team can get to the NBA playoffs
Jordan Clarkson is considered >95% to win by the betting markets. If the season were to end today and he did win, would you consider him winning 6MOY to be the worst awards decision of recent memory? #peltonmailbag
— Sam Garriock (@Sam_Garriock) April 1, 2021
No, I don't think it comes particularly close. As noted, Clarkson winning is generally in keeping with the spirit of the Sixth Man Award. He wouldn't be my pick, as I think his Utah Jazz teammate Joe Ingles has done more to contribute to winning this season -- Ingles leads eligible candidates with 6.5 WARP to 4.5 for Clarkson -- but that difference isn't enormous. Three of the past four winners have had lower WARP than Clarkson, although Montrezl Harrell's total is affected by the higher replacement level I utilize for centers.
I suppose my answer to the question depends on how we're defining "recent memory." One decision that was striking in my research was Jermaine O'Neal winning Most Improved Player in 2001-02 with 2.4 fewer WARP than he'd posted the previous season. (Other advanced stats suggest O'Neal was about as valuable both years.) O'Neal played five more minutes per game and played a larger role in the offense, boosting his scoring average by 6.1 ppg to win the award.
If we're limiting it to the past decade, I'd be inclined to pick Jamal Crawford's third and final Sixth Man Award in 2015-16. Crawford had a much stronger case the first two times he won, but by 2015-16, he wasn't even the leading scorer among eligible reserves, ranking fourth behind Ryan Anderson, Jrue Holiday and Will Barton.
In hindsight, I'm baffled Holiday didn't get more consideration after averaging 16.8 ppg and 6.0 apg while starting 23 of his 65 games as a precaution because of stress issues in his legs. Holiday got just one third-place vote. He wasn't in my top three; I chose Andre Iguodala from the 73-9 Golden State Warriors, who finished second to Crawford in the voting.
These are some good takes! #peltonmailbag does COY predict being fired a few years later?
— Andrew Schein 🐢🍦 (@a_r_schein) April 25, 2021
This question came up in the context of my point that Coach of the Year has become primarily about teams exceeding expectations, which makes it harder for coaches who are strong year in and year out to win the award. Nonetheless, there's not much recent track record of the Coach of the Year being a poisoned chalice for winners by creating unrealistic expectations they can't meet.
This might have been the case before the past decade. From Mike Dunleavy in 1998-99 through Mike Brown in 2008-09, nine out of 11 winners lasted no more than two full seasons after the award in the same job. That was clearly voluntary in the case of Hubie Brown, who retired during the season after winning in 2003-04, and some manner of mutual in the cases of Larry Brown in Philadelphia and Mike D'Antoni in Phoenix. But six of those 11 coaches were fired not long after winning Coach of the Year.
Since Mike Brown's win in 2008-09, just two Coaches of the Year have failed to last at least two seasons in their job after winning: George Karl in 2012-13 and Dwane Casey in 2017-18, both of whom were fired that offseason. (In Casey's situation, before he even was named the winner at that year's NBA awards.) Aside from that, the earliest departures were by D'Antoni from Houston and Mike Budenholzer from Atlanta three years later, when D'Antoni elected to leave and Budenholzer and the Hawks mutually agreed to part ways so he could seek another head-coaching job.