I’ve been thinking about football stats now that the NFL and NCAA seasons are upon us, and I’ve been wondering (as usual) how we can better rate teams and make predictions. One big issue in football that makes prophetic prediction more difficult is injuries. There was a great article that came out in the New York Times about the costliness of certain injuries, and each position’s relative value to the team. Since injuries, the article argues, are almost impossible to foresee, our predictions are limited to the other aspects of a team that we can predict.
Statistics that are common for football teams include points scored per game and points allowed per game, and we also rate offenses and defenses on yards per game, in addition to points. Here’s my argument: some teams play faster-paced games and therefore get more possessions, while other teams like to grind it out on the ground, increasing time per possession and reducing possessions per game. I think it would be illuminating to look at points and yards per possession. With this data, we could look at how a certain team performs – in terms of scoring and yards per possession – in games where possessions are limited versus how that same team performs in games where possessions are numerous.
Understanding a team’s efficiency at various game paces could potentially help a lot in predicting the outcomes of games between vastly different-styled teams. While breaking down every one of 256 games is a daunting task for my Monday morning, I took a look at the 2008 season as a whole to see if anything looked remotely interesting.
New Orleans led everyone and his mother offensively in yards per game, touchdowns per game, and of course, points per game. The Saints scored 7 more offensive touchdowns than next-best San Diego, outscored San Diego by 1.5 points per game, and outgained the Chargers by more than 60 yards per game*. However, the Chargers utilized more of a balance between running and passing, thus eating up more clock and limiting their total possessions. In terms of points per possession, the Chargers actually came out on top. With fifteen more possessions on the season, enough to match the Saints, the Chargers actually would have scored 18 more total offensive points than New Orleans.
I think any stat per possession is going to be more telling than that same stat per game. Using more clock up, the Chargers limited their own scoring potential, but they simultaneously reduced the scoring potential of their opponents.
We can do the same thing with defensive efficiency, looking at how well a defense holds opponents to limited yards each set of downs. Looking at the college game, the Oregon Duck defense has given up an unimpressive 26.3 points per game thus far. But the quick three-and-out offensive strategy employed by Masoli and company – they had the ball for a grand total of 1:30 during the first quarter of Saturday’s game – puts the defense out on the field for more possessions per game. I would venture to guess, though the stats are not available to confirm, that Oregon’s defense has faced more possessions this season than most other teams, and that their points allowed per possession could be one of the best (lowest) in the country.
Once this NFL season gets rolling, I hope to be able to start breaking down each team’s games in terms of slow-paced, normal-paced, or fast-paced games, and then determine if certain teams tend to have advantages over other teams that play vastly different styles.
*The Chargers were second overall in touchdowns and offensive points, but 11th in total yards. I counted a team’s offensive possessions by adding up offensive touchdowns, field goals, turnovers, 4th down failures and punts, the five basic ways a possession can end.