Explaining Team Efficiency Ratings

If there is just one thing you learn in sports today, it should be this: Ranking football offenses by how many yards they gain is wrong.

For as long as most fans can remember, NCAA team leaders in "Total Offense" and "Total Defense" have been ranked in terms of how many yards they gained and gave up, not turnovers, not third down conversions, not points, and not first downs. This is both a bad concept and awful labeling. New Coke was a bad concept that didn't deserve the label of "Coke," but that was 30 years ago; the company destroyed both the product and the label within 10 years. Somehow this "Total Offense" thing has lingered for decades.

There is a better way and we will be using that for college football beginning this year. We are calling the rankings Team Offensive Efficiency, Team Defensive Efficiency, and Special Teams Efficiency. The numbers are in terms of scoring margin -- actual points on the scoreboard -- and they are a lot more "total" than yards.

Before explaining fully why yards are not as useful, let's clarify what these ratings are. In 2012, Oregon had a 12-1 record and an average scoring margin of about 28 points per game. These team efficiency rankings, when not adjusted for opponents, essentially answer the question, "How much of that 28-point margin was associated with their offense, defense, and special teams?" Because college football schedule strength varies so much, the typical efficiency values that we will use include an opponent adjustment. So Oregon's opponent-adjusted point margin is actually about 32 points per game and the question is which unit is responsible for those points.

The answer to this question, which gives the Oregon team efficiency ratings, is:

• +21 for their offense,
• +11 for their defense, and
• +0.3 from their special teams.

In other words, against an average defense, Oregon's 2012 offense would add 21 points towards the final scoring margin of the game. Against an average offense, their defense would add 11 points by forcing turnovers and improving field position for their offense. Against average special teams, the Ducks' special teams would add essentially nothing, about 0.3 points, though punt returns and field goals can obviously add some. Against a uniformly average team on a neutral field, Oregon's 2012 team would win by 32. Maybe the final score would be 49-17 or 42-10 -- these ratings don't say that, but they do capture how much the defense shortens the field for the offense, defensive touchdowns, punt and kickoff returns putting the offense in good position, field goal kickers who make long field goals, blocked punts, turnovers, and, of course, yards gained.

Among all the metrics ESPN is putting out, these are the most inarguable ones. First of all, the method is not new -- the expected points added concept underlies this and it has been implemented in football for years by other people (such as shown here and here). Second, they are based completely on historical averages for how many points a team has scored in a given situation. This nearly completely untangles the interaction between offense, defense, and special teams. Defense and special teams set up an offense with a shorter field and good offense allows the defense to force opposing teams to go farther.

Because they are so useful, team efficiencies form the basis of QBR, of the Football Power Index, and of many discussions about teams and players. Here are a few examples:

• Alabama was National Champion in 2009, 2011, and 2012. In those years, they also had the NCAA's best defense by defensive efficiency. Somewhat surprisingly, the 2009 defense was actually the best of these three, even though the 2011 defense was setting records for fewest points and yards. That was because the 2011 defense faced much worse offenses than the others. Though we prefer to account for opponent strength with defensive efficiency, if we don't, that 2011 defense was the best.

• The LSU special teams have ranked in the top 10 every year from 2008 to 2012, reflecting Les Miles' emphasis on using athletes on his special teams units.

• The effect of Jadeveon Clowney on South Carolina's defensive efficiency is not as strong as one might think. In the two years before his arrival, the defense ranked 17th and 11th, adding an average of 10 points per game to their scoring margin. In the two years after his arrival, the defense ranked 12th and 8th, adding an identical 10 points per game to the scoring margin. He may be good, but whatever South Carolina was doing before he arrived -- with its system and with good players -- was already pretty good.

• Oregon was 44th in the so-called "Total Defense" in 2012, but that ignores the fact they forced 40 turnovers -- most in FBS. That played a big role in the Ducks being ranked 6th in defensive efficiency.

• Oklahoma State in 2012 faced 7 of the top 14 offenses by offensive efficiency. Partly as a result of that and partly because they gave up a lot of yards (a stat that lies, having had them 82nd), the Cowboys' defense was not viewed to be very good. But adjusting for their competition -- and incorporating everything else we've mentioned -- the Cowboys rose to 17th in defensive efficiency.

So what's wrong with yards?

• A team that gains 80 yards to go from their 10-yard line to the opponents 10-yard line before having to kick a field goal -- that is a bit less than a team that gains 80 yards from their own 20 to score a touchdown.

• A team that gains 80 yards to go from their 10-yard line to fumble at the opponents' 10 -- that is A LOT worse than a team that gains 80 yards from their own 20 to score a touchdown.

• A team that gains 80 yards, but has 40 yards worth of penalties before having to punt, only goes from their 10 to midfield.

All of these situations are captured with team efficiencies. And that's what you should learn if you learn two things in sports today.