<
>

Reintroducing ESPN's team efficiency ratings

Yards per game doesn't provide a true measure of Alabama's offensive efficiency. Pouya Dianat for ESPN

With college football only one day away, we at the ESPN Stats & Information group want to reintroduce some of our metrics, starting with our revamped Team efficiency ratings (for offense, defense and special teams):

Definition: Efficiency measures a unit’s per-play contribution to a team’s scoring margin, adjusted for the opposing unit faced, expressed on a 0-to-100 scale.

-Efficiency is calculated on a per-play basis to better control for the varying pace at which teams play.

-The strength of the opposing units faced are taken into account in the efficiency ratings.

-The calculations reduce the importance of blowout situations.

-Efficiency is expressed on a 0-to-100 scale (50 is average).

-The ratings account for turnovers, red zone efficiency and the ability to convert yards into points, among other factors.

-Efficiency ratings can be thought of as “QBR” for offense, defense and special teams.

Why is this metric needed?

Traditional measures of a unit’s success (such as yards) are often misleading. Consider the example below:

A team gains 80 yards and…

-Scores a touchdown

-Goes from its 10-yard line to the opponent’s 10-yard line before having to kick a field goal

-Goes from its 10-yard line before fumbling at the opponent’s 10-yard line

-Incurs 40 yards of penalties on the drive, going from its 10-yard line to midfield before having to punt

In each case above, a team is credited with 80 yards of total offense, but the outcomes are drastically different. The main goal in football is to score points, and at times total offense (or even points per game) is not the truest measure of a team’s offensive scoring efficiency.

Cases in point: Alabama and Oregon

Alabama has ranked outside the top 15 in yards per game every year since the start of the 2009 season, but the Crimson Tide are one of the most efficient offenses (top 10 in efficiency each year) during that time. Total yards does not account for Alabama’s ability to convert yards to points (second in scoring rate), avoid turnovers (first in turnovers), control field position (fifth in field-position margin) and move the ball (third in yards per play) during that time.

Similarly, Oregon’s defense has not ranked higher than 34th in yards per game allowed in a season since the start of the 2010 season. But that doesn’t account for the fact that Oregon’s defense plays more snaps than many other teams’ defense (often in blowout situations), faces some of the toughest offenses in the nation in the Pac-12 and has forced the most turnovers during that time. Oregon has ranked in the top 30 of the efficiency ratings in each of those five seasons, including in the top 11 in 2010, 2011 and 2012.

ESPN’s team efficiency ratings account for all of these factors through the way it evaluates the success of every play.

How does efficiency work?

Efficiency is built off the concept of expected points added (EPA), a popular concept in football analytics. EPA accounts for context by measuring the impact of every play on a team’s potential points. It factors in situational aspects such as field position, down and distance to measure the increase or decrease in potential scoring with each play result (hence, expected points added).

For example, gaining 5 yards on third-and-4 is generally a good play in terms of EPA, and gaining 5 yards on third-and-6 generally delivers negative EPA.

For more on why EPA is the backbone of most of our metrics, you can read this article by Alok Pattani, associate director of Stats & Info’s Analytics Team.

For those who remember the previous efficiency ratings, the new system is built off the same framework. By accounting for the success of every play, team efficiency ratings better adjust for pace and are a truer measure of a unit’s effectiveness.

Related articles

Want to learn more about ESPN’s college football analytical metrics?

-A primer on the Football Power Index (FPI)

-Evaluating teams’ résumés