2016 NBA MVP in Data

Last season, I reported on two metrics that seemed to be decent predictors for NBA's Most Valuable Player award: effective field goal percentage (eFG%) and usage percentage (USG%). Here, I revisit these metrics and try out a new metric that uses these two metrics in a simple way.

Voting share for all MVP candidates

For those unfamiliar with how the NBA MVP is selected, reporters and broadcasters from across the nation rank five players of their choosing from 1st (most valuable) to 5th. 1st place gets 10 points, 2nd gets 7 points, 3rd gets 5, 4th gets 3, and 5th gets 1 point. The most amount of points you can receive is 1310 points -- that would be receiving every single first place vote from all voters. Stephen Curry was the first player in history to achieve the unanimous MVP, earning all 1310 votes.

The score you receive can be easily converted into a percentage when divided by 1310 points, known as a voting share percentage (VS%). Since Curry had 1310/1310, this simplifies to 1. Kawhi Leonard earned 634 points (0.484 VS%).

Effective field goal percentage

Effective field goal percentage (eFG%) is an advanced shooting statistic that takes into account the value of a 3-point field goal. In a previous post, I showed the MVP as being someone who has a relatively high eFG%. Out of the MVP vote-getters, Curry had the greatest eFG% (0.63). In comparison, Curry's eFG% from last year (when he also won MVP) was 0.59.

Usage Percentage

Usage percentage is a metric that includes field goal attempts, free throw attempts, minutes played, and turnover. When computed, this advanced statistic tries to determine how many scoring opportunities a player generated in comparison to the total team opportunities, with respect to the amount of turnovers a player incurred. Both Curry and James Harden were used to a nearly identical percentage (0.326 and 0.325, respectively). Last year, Curry had one of the lowest USG% for an MVP in the last decade (0.28). This year, Curry's USG% was similar to the last four unique MVPs (Kobe Bryant, Derrick Rose, Kevin Durant, and LeBron James).

My simple MVP rating system

Since I think both USG% and eFG% are good predictors, I averaged the two statistics together to create my own MVP percentage (MVP%). With this rating system, my MVP order would have been 1st-Curry, 2nd-Durant, 3rd-James, 4th-Harden, and 5th-Leonard. This ranking represents 4 of the top 5 MVP candidates this year, with Curry and James being in the same position they were voted into. Kawhi sees a drop from 2nd to 5th, Durant from 5th to 2nd, and Harden goes from 9th place to 4th place. The biggest loss here is Russell Westbrook dropping out of the top 5.

A debate over "most valuable" and "best offensive" player

LeBron James made some interesting comments about what makes a player "valuable". He acknowledged statistics, and that a player with the best statistics should absolutely be in the running (he used the word deserving) of the MVP award. However, he also proposed what a player's value is to a team -- which he did not go into detail over, but it was apparent that he was implying stats don't mean everything. How true is this?

On one side of the debate, the MVP voters have some agreement with James's statement. Durant and Westbrook are teammates, but they both appear on the top 10 most valuable players -- same with Draymond Green and Curry. Logic would state there could only be one "most valuable" player from each team, but what if certain players contribute in ways not seen on the stats sheet? Green has been touted as one of these types of players, noted as "the heartbeat" of the Warriors. Statistically, he had the lowest USG% out of all MVP candidates, indicating he was not typically the team's primary point scorer (that would be Curry). But his value is not necessarily points, but everything else. On the Warriors, he was 1st in total rebounds, first in assists, second in steals, and second in blocks. He also led the team in fouls and was 4 TOs shy of leading the team in turnovers (Curry led the Warriors with 262 TOs).

But value, from a baseline standpoint, is directly related to wins. And you win by making points. Curry dominated in eFG% for all guards. He touched the ball most out of all guards. He led the league in points per game, free throw percentage, steals, and total field goals attempted. He scored 1 less point than James Harden, ranking him 2nd in the league for points total. However, Harden played a league-leading 3125 minutes this season while Curry generated all his points within 2700 minutes. Scoring a lot of points in less amount of time. That is value. Even Harden has immense value, contributing to the majority of Houston's offense.

I believe the MVP is fine as it is. There is a slow change to represent how important guards are to the current style of NBA play. In the last 20 years, only 6 MVPs were guards, with two of them being Steph Curry these last two years. Media votes are good indicators of the spirit of the game, and the current spirit is Steph Curry. Brute physicality is changing for precision shooting, and Steph Curry embodies this spirit. And until the rules change to value something other than the power of the 3 point shot, I believe the MVPs of the era will be best represented by USG% and eFG%.

Leave a Reply