Sunday, October 15, 2017

NBA MVP Voting

We mentioned in class briefly that many sports polls are done by the Borda Count in which the voters rank teams or players and the teams/players get points based on how they are ranked. I decided to explore if there had been any controversies due to the use of this type of election system. As a basketball fan, I looked into NBA MVP voting over the last couple decades. For context, the NBA MVP voting system is a modified Borda count method in which voters rank their top five players and the points are distributed as 10 for 1st place, 7 for 2nd place, 5 for 3rd, 3 for 4th, and 1 for 5th.

There has only been one time since this system's inception that the player who garnered the most first-place votes did not win the MVP. In the 1990 race, Magic Johnson got 69.1% of the possible points, while Charles Barkley got 66.7%, and Michael Jordan got 62%. However, Magic Johnson got only 27 first-place votes as compared to Charles Barkley's 38 and Michael Jordan's 21. Despite Charles Barkley's significantly higher number of first-place votes, Magic Johnson received 38 second-place votes compared to Barkley's 15 and Jordan got 25. Barkley would have won a plurality system just based on voting for your top candidate, but his inability to successfully amass non-first place votes hurt him in this modified Borda count method. As the NBA didn't release full voting data until more recently, I couldn't determine the Condorcet winner in this example, but I did look into the full voting data for the last three races and all three winners using this system were also the Condorcet winner.

No comments: