Curry the greatest MVP candidate ever? NBA's first unanimous MVP?

ESPNAPI_IMG_NO_ALTEXT_Value

The 2015-16 NBA season marks the 60th anniversary of the first time the league handed out its MVP trophy. In those six decades, no player has won the award unanimously.

There have been near misses. Shaquille O'Neal in 1999-00 and LeBron James in 2012-13 secured 120 of the possible 121 votes, coming agonizingly close to achieving unanimity.

Is this the year? Will Stephen Curry make a unique kind of history by sweeping the voting? 

And is Curry the best MVP candidate of all time? We can take apart these questions and answer them one by one.


What makes an MVP

Predicting MVP voting with reasonable accuracy is relatively easy, looking at three simple factors:

  • The gap between the best player and the next-best player (as measured by my wins above replacement player metric, WARP).
  • Points per game.
  • Wins by a player's team.
  • This is what my model shows, based on the NBA from 1980-81 to now, the era in which the media has voted on the award, replacing players as voters.

    Accounting for these three factors correctly picks 23 of the 35 MVPs since 1980-81, and projects the player who finished second or third in the balloting 11 of the remaining 12 times.

    (The one exception was 2004-05, when Amar'e Stoudemire -- who actually finished tied for ninth -- was projected over teammate Steve Nash, who won the award.)

    Of the three factors that tend to consistently influence MVP voting, team success has been most important. All other things equal, each extra win has been worth 1.8 percent of possible first-place votes on average.

    So the difference between playing on a 60-win team and a 50-win team is an 18 percent swing in first-place votes, typically enough to make the difference between winning MVP and finishing runner-up.

    By contrast, individual performance usually isn't all that important. Last season, Curry finished 0.6 WARP ahead of MVP runner-up James Harden. The history of the voting suggests that difference was worth only about 1.4 percent of first-place votes to Curry.

    The Warriors' 11-game advantage over Harden's Rockets in the 2014-15 standings, by contrast, was worth about 20 percent of first-place votes on average.


    Curry's current MVP case

    Barring an injury -- knock on wood after he briefly had to leave Saturday's win over the Oklahoma City Thunder when Russell Westbrook landed on his ankle -- Curry long ago locked up MVP honors. The only question now is just how many votes he might get.

    Take a look at the three factors we just identified. Curry leads the league in WARP (17.0), though Westbrook (15.4) has managed to stay close behind. Projecting out through the remainder of the schedule, that should translate into a 2.8-WARP advantage for Curry, which would rank him 10th among MVPs since the media began voting.

    Ordinarily, I'd point out the possibility of Curry regressing from his historic offensive pace, but if anything, he seems to be getting stronger as the season goes on. In terms of John Hollinger's game score metric, Curry's six games since the All-Star break -- all on the road, and in nine days with a pair of back-to-backs -- have been his second-best stretch, trailing only the first six games of the season. Curry's average game score for the season has risen from 23.8 (as of Jan. 8) to 25.3 now.

    Curry's recent hot streak has lifted his scoring average to 31.0 points per game, which would trail only Kevin Durant's 2013-14 MVP campaign (32.0) in the past nine seasons.

    And then there's Golden State's pursuit of the best record in NBA history, which became more realistic after Curry's deep 3-pointer won Saturday's game in overtime.

    The victory lifted the Warriors to 52-5, and even before the game,  ESPN's Basketball Power Index projected Golden State to finish with an average of 72 wins, which would match the 1995-96 Chicago Bulls for the most ever.


    Best MVP candidate ever

    Add it all up and there's a strong case that Curry is the best MVP candidate ever, at least in the era in which media has voted for the award.

    (For whatever reason, players tended to be much more unpredictable and diverse in their voting. The highest percentage of first-place votes achieved before media began voting was 133 out of 153, 86.9 percent, by Kareem Abdul-Jabbar in 1970-71.)

    My model tends to produce conservative estimates of the percentage of first-place votes the best candidates should receive, given that there is generally strong competition for the award. Regardless, the model described at the top of this article shows that Curry has the strongest case of anyone since 1980-81 by a fairly wide margin, looking at the model's projected shares of first-place votes.

    Of the eight highest first-place percentages as voted by media, six appear on this list -- all but Kevin Garnett in 2003-04 (projected to get 51.3 percent of first-place votes) and Durant in 2013-14 (projected 50.3 percent), both of whom played for teams that won fewer than 60 games.

    Based on the three key factors, Curry has a better chance at a unanimous vote than the two players who came closest, and they came very, very close.

    But relative to Curry and these Warriors, O'Neal and James played on slightly weaker teams, winning 67 and 66 games respectively in their MVP seasons. While O'Neal had a much larger advantage in WARP (5.9 more than the next-best player), Curry has the highest scoring average of the three.

    So despite the incredible MVP seasons posted by Larry Bird, Michael Jordan, Shaq and LeBron (as listed in the table above), it does appear that Curry will be the best MVP candidate ever by the standards the voters have set.

    It would take only one rogue voter to prevent Curry from taking the singular honor of first ever, so this prediction comes with a measure of uncertainty. But given everything he and the Warriors have done this season, and are likely to do, Curry will be the first unanimous MVP if history is our guide and no voters are determined to prevent it.