The season is still early, but we've already had several players throw their hats into the MVP ring, especially two Splash Brothers in California with separate 50-point games. Young players such as Giannis Antetokounmpo and Anthony Davis have led their teams to excellent starts, and we've even seen resurgences from veterans such as Kyle Lowry and Blake Griffin.
So based on the first few weeks of action, who is in the driver's seat for the prestigious award?
Which Metrics Best Predict the MVP?
To answer that question now and throughout the season, I wanted to build a MVP model of sorts. I was curious what metrics were the most predictive for MVP voting, so I looked at all the races going back to the 2000-01 season.
To put things simply, several of the publicly available all-in-one metrics did a pretty good job of predicting the award. Based on correlations, total Win Shares for a season explained about 60% of the variance of MVP voting, and even something like PER, which is maligned at this point by stats gurus, has been predictive, explaining about 59% of the variance of MVP voting among likely candidates.
Of the players to produce the 14 best Win Share seasons since 2000, nine of them won the MVP award that season. And the players who didn't win were bested in win shares by a guy that season — think LeBron winning over KD in 2012-13, for example.
The season leader in Win Shares has won 11 of the past 18 awards, which is a good start.
Interestingly, on/off or plus/minus metrics didn't backtest very well. Simple on/off explained about only 25% of the variance, and more advanced plus/minus numbers — such as ESPN's Real Plus-Minus, which takes into account context and teammates — just aren't publicly available going back 20 years.
So I decided to keep it simple and use the more predictive all-in-one metrics as the heavier weights.
Anyway, putting all that into a simple model, I am rating each player on a 0 to 10 scale of MVP candidacy. I'll update this throughout the year, as it will certainly shift as we get more data. But even now, I think we can get a nice peek into who the frontrunners are currently and whether there's any betting value to be had.
One quick note, though: A good chunk of the players in the NBA will shift way down toward zero as the season progresses, but they're higher now because of the smaller sample.
Stephen Curry, who leads the NBA in scoring with 260 points, has only 109 more points than Buddy Hield. As the gap widens, the majority of the field will drop down and only the major candidates should emerge.
Notes
Stephen Curry is far and away the leader right now, given his individual output — he leads the NBA in scoring — plus excellent marks in advanced metrics and team win rate.
Not all sportsbooks have MVP odds up every day, but bet365 does, and currently Curry is just the fifth-largest favorite there at +650. Considering Curry's hot early start and how dominant the Warriors look, I think he's the best value on the board right now and should be the frontrunner.
Of course, the biggest argument against him is that his teammates, specifically Kevin Durant, will cannibalize votes.
However, I think that's probably a bit overstated at this point: James Harden won last season with Chris Paul in town, and LeBron James won multiple MVPs in Miami after getting publicly ridiculed for forming a "super team" with Dwyane Wade and Chris Bosh.
If Steph continues his current pace, he should win — regardless of how good his teammates are.
As I mentioned above, players will move up and down as we get more data. The best example of that recently is another Warrior, Klay Thompson. Before his 52-point explosion, he had an MVP Rating of 3.56. He now sits at 4.50. A single game won't mean nearly as much in February as it does now.
One last note: Several of the favorites in the betting markets — Giannis Antetokounmpo (+375), Anthony Davis (+450), LeBron James (+450) and Kawhi Leonard (+650) — are still very much in the running.
However, guys such as James Harden (+850) and Russell Westbrook (+1700) have incredibly low MVP Ratings after the first couple weeks of the season.
They can put up numbers in a hurry, of course, but their slow starts could prove costly in such a loaded field.
Stay tuned for weekly updates this season!
