O-Score: Measuring Offensive Efficiency (A Preface)
How do we determine how efficient a team’s offense is? It’s a tricky question to answer and perhaps more complex that it seems at first. This is the first part of what will be a week-long look at offensive efficiency. Lets start off with what appears to be an obvious statement:
Football games are won by scoring more points than your opponent. The vast majority of points are scored by a team’s offense (93.5% since 1970). Therefore, a good offense is one which scores the most points.
Right? Well, not really. Imagine this scenario:
Team A has 12 offensive drives in a game. They score touchdowns on 4 of them (and go without points on the other 6). At the end of the game, Team A has scored 28 points.
Team B only gets the ball 8 times and scores touchdowns on 3 of them. They end up with 21 points.
Which is the better offense? Team B scored more frequently than Team A but their overall points-per-game output is hampered by their lack of offensive drives. Further complicating this evaluation is that the reasons for Team B having less drives could be related to any number of factors – some of which are offense driven and some of which are out of the offense’s control. I’d probably prefer to have Team B’s offense which is most likely (but not necessarily) more efficient than Team A’s. It presents us with an interesting challenge in determining offensive efficiency.
In my mind there is a (perhaps subtle) difference between “value” and “efficiency”. In terms of the way it is commonly evaluated: value is situation and play dependent, whereas efficiency is best defined in a broader view. Therefore, efficient offenses are almost always valuable but valuable offenses are not necessarily the most efficient. This is especially true once we get beyond the extreme ends of the statistical distribution. To illustrate this here is another hypothetical situation:
Team A has 9 offensive drives in a game and scores a TD on all 9. They win the game 63-0
Team B has 9 offensive drives in a game and scores a TD on all 9. They win the game 63-62
Both teams have equally efficient offenses (in an overly simplified touchdowns-per-drive way) but team B’s offense would have more value. In Team B’s case, their offense is able to keep pace with their opponent and the most common “value” metrics would give them a bonus for consistently getting go-ahead scores after their opponent scores. In other words, their 9th touchdown would have much greater value than Team A’s 9th TD.
For the purposes of this discussion, I’m going to stick to offensive efficiency. If you’re interested in a value-driven discussion, the folks over at Football Outsiders do an excellent job with their DVOA metric.
So, how do we measure offensive efficiency? I haven’t found a specific stat which answers that question to my liking, so I decided to create my own which I call simply “O-Score” (or Offensive efficiency Score). It’s essentially a team’s performance relative to league average in a few different efficiency stats with a strength of schedule adjustment.
Here are the statistical categories I’ve chosen to include:
Points Per Drive
Yards Per Point
Yards Per Drive
Over the course of this week, I’ll be discussing each one of these variables separately and then unveiling the final scores over the weekend. Until then, check out the best and worst point-per-game offenses since the AFL/NFL merger:
Where will these offenses rank in terms of efficiency? Check back later this week to find out.