Simplified Analytics: Player Efficiency Rating and the Milwaukee Bucks

In the past decade, the basketball community has seen a shift in the way we look at player performance. Those like Memphis Grizzlies VP of Basketball Operations John Hollinger, Houston Rockets General Manager Daryl Morey and Milwaukee’s own Director of Basketball Research Seth Partnow are using more and more advanced data analytics to look at player impact. This change has brought a shift in both the style of play and in how we evaluate player production during roster assembly.

This series will break down these advanced, complicated statistics and discuss their relation to the Bucks’ roster. We’ll start with one of the most well-known stats, player efficiency rating (PER).

What is PER?

Player efficiency rating was created by the aforementioned John Hollinger. It gives players an all-inclusive rating for their per-minute production over the course of a season. By incorporating both offensive and defensive stats, PER provides an easy overview of player performance.

Calculating PER

In order to provide a single comprehensive number, the PER formula is pretty complex. For the full formula, check out basketball-reference or Hollinger’s article on ESPN Insider (here’s a brief primer from the stat god himself). Below are the included factors, weighted based on importance:

-Field Goals, Three Point Field Goals, Assists, Blocks, Rebounds, Steals, Minutes Played, & Pace

Interpreting PER

The PER for an average role player is 15.0, higher being better and vice versa. Players scoring in the 20’s generally contend for All-NBA or All-Star spots, while single-digit scorers tend to look for new teams.

Criticisms of PER

One number that tells you everything you need to know? Yes, it is too good to be true. The most common criticism with PER is that it heavily favors offensive players by only including two defensive stats (steals and blocks). This helps big men look gaudy by comparison to their wing counterparts who may not tally rebounds or sport a stronger shooting percentage by virtue of their perimeter oriented game. PER also fails to address the difficulty of the opponent. It’s easy to put up big numbers against a bench, but that shouldn’t automatically earn you a higher PER. A final issue is that PER doesn’t factor in everything. Some players provide huge value on the court without necessarily putting up the high stats PER is looking for. Where would the Bucks be without Hype Man Jason Terry? Unfortunately, “hype” isn’t measured in PER.

The Bucks and PER

Unsurprisingly, Giannis led the 2016-2017 team with a PER of 26.1, considered a quite strong season and ranked 10th overall in the entire league. Shown below is how the rest of the team stacked up.

Henson gets a bump in PER due to his ability to block shots but I wouldn’t say he had a better season than players like Snell, Brogdon, or even Terry. I thought Monroe would have a higher PER this season compared to last but that wasn’t the case (21.9 in the 15-16 season vs 21.2 in 16-17). This could be due to fewer rebounds and points scored despite the increase of steals. Overall, the PER rankings of the Bucks really don’t correspond to the fan rankings of their performances, apart perhaps from Giannis, Vaughn, and Toupane (remember him?).

Snell seems like the guy who gets the most short shrift here, and some of that is certainly due to PER’s inability to value wing players who don’t dominate the ball. Klay Thompson, for example, is ranked 74th in PER at 17.53, just two slots behind the bowling ball of mediocre that is Marreese Speights. So, take Snell’s ranking with a grain of flawed stat salt.

Overall PER is an easy stat for to provide an overview of player performance, but it’s far from a perfect rating. It isn’t the most accurate or simple formula, but it has laid the foundation for the catch-all performance-based advanced stats used in the past decade.