Like any election, the public is fiercely divided on the issues. Virginia Tech's Erick Green's candidacy hinges on whether voters believe a team's record should factor into Player of the Year. Duke's Mason Plumlee's prospects depend on whether people believe POY is a full-season award or just based on the conference season.
Campaigning is in high gear as the regular season winds down. One-page hand-outs touting a school's candidate are in media rooms across the conference, right next to the game notes, and coaches are stumping on behalf of their guy.
After a loss to Duke, Virginia Tech's James Johnson pointed out that Green's eye-popping scoring totals have been done against double and triple teams, and that Green usually defends the other team's best player.
As the race moves into the final days, media members are going Nate Silver, crunching numbers to find a way to determine who truly is the best in the ACC. ESPN's Len Elmore assigned weights to various statistics to come up with an Elmore Index to rate the players, an undertaking mocked by ACCSports.com, who created their own index declaring UNC's P.J. Hairston POY.
Rather than subjectively determining how much a rebound is worth relative to a point, it's perhaps best to look at what's most important in basketball—outscoring the other team.
Plus/minus is one statistic that can do that. It tracks how the score changes when a player is on the floor. The problem is, by itself, plus/minus depends heavily on how good a player's team is. No one can deny that Green is a great player, but the Hokies are outscored by 2.6 points a game while he's on the floor. That's lower than every player on the Duke roster.
So instead of merely looking at plus/minus, let's look at how a player's team does with him on the floor, versus how they do when he's NOT out there.
For instance, the Hokies were outscored by 78 points in Green's 1095 minutes on the court this season. But in the 125 minutes that Green was out of the game, they were outscored by 39.
In other words, without Green, the Hokies were outscored by .31 points per minute. With him, that fell to .07: a .24 points per minute improvement.
Here's how that ranks against the other candidates:
And what about Duke? Based on the full-season numbers, they have a definite POY candidate, but it's probably not who you think.
Both Curry and Plumlee have negative impacts, meaning that the team actually does better, per minute, when they're not on the court. Curry was hurt by the fact that the team outscored Delaware by 38 in a game he missed. Take that out, and the team did about as well with him as without him.
Duke's candidates actually did better in conference play. Here are the numbers for the contenders in ACC games only:
Miami actually did better with Larkin off the court in ACC games, while Harris and Howell established themselves as front-runners. Here are Duke's best candidates:
People wondering why Thornton and Hairston are among Coach K's favorites will notice that they scored highly in this measure. It's also interesting that in ACC games, Duke was outscored by opponents when Seth Curry was on the bench.
But for consistency all year long, the numbers don't lie. It's clear that there's really only one choice for ACC Player of the Year: Ryan Kelly.