In looking at Offensive Footprints yesterday, I probably didn't make something entirely clear: the goal of the footprint is to serve as a supplement to performance data. So what I have in my head is something that serves as a "how they achieved the quality (or lack thereof) that you are seeing in other data sets." The best illustration of this is once again on the Pomeroy team profiles that I referenced yesterday. In the "Style Components" section, Pomeroy lists 3PA/FGA (the ratio of 3-point attempts to overall field goal attempts), A/FGM (Assists per Field Goal Made), and Defensive Fingerprint, which uses a team's statistical profile to guess how much a team likely plays man defense or zone. If opponents shoot a large number of 3-pointers on you (making them is irrelevant), if you don't foul a lot, if your opponents have a high A/FGM (meaning they had to pass to find open shots instead of creating on their own), and if you aren't great at defensive rebounding, signs point to you running a lot of zone.
One other point I should make: while charting is something we're trying to coordinate at Football Outsiders (it is incredibly difficult with 120 FBS teams and iffy TV availability), for the most part we are dealing simply with a line of play-by-play like this one.
1 and 10, FS47. Manuel, EJ pass complete to Reed, B. for 16 yards to the VT37, 1ST DOWN FS (Whitley, E).
With this line, we have no idea about either the length of the pass (Was it 16 yards in the air? Was it a screen that Reed took upfield?), the formation at hand, or where Reed was lined up to start the play. Our information here is limited ... but so is Pomeroy's. Over the course of 800(ish) offensive plays and 800(ish) defensive plays, you can still derive a lot about a team's general style and personality.
So. With this in mind, what information would one regard as useful in terms of a team's "defensive footprint"? Here are some options I have been tossing around:
- Opponent Run-Pass Splits. The same information shared the Offensive Footprint post yesterday. One of the best ways to look at a defense might be looking at how opposing offenses chose to attack it.
- Aggression Level. I have learned over time that you can derive a team's personality a bit from the ratio of their success rates to their Points Per Play (PPP). (Both of these measures are key aspects of my S&P+ rankings.) If a defense allows a large success rate (efficiency) but maintains a rather low PPP (explosiveness), then they almost certainly qualify as a "Bend, Don't Break" defense. On the other hand, you sometimes see defenses with low success rates and higher PPP averages, suggesting they take risks and shut down a lot of plays ... but when they suffer a breakdown, it's a large one. It wouldn't be hard to create an aggressiveness/passivity scale that could be rather telling.
- Biltzability. This would be quite a generalization, but when you look at Standard Downs and Passing Downs sack rates long enough, you start to conclude that teams with high Standard Downs sack rates are much more likely to be able to generate pressure without blitzing. This makes sense, of course -- you're infinitely more likely to blitz on passing downs than standard downs. If you ranked high on standard downs sack rates but were average or bad on passing downs, then you weren't a very good (or frequent) blitzing team. If you were the opposite -- poor standard downs sack rates and nice passing downs rates -- you were a team that potentially had iffy pass rush ability from your ends, but you blitzed both well and often when you got the opportunity.
- ?
That's really about all I have for the defense. I loved some of the (realistic) offensive suggestions, and I am downright counting on some suggestions for the other side of the ball. Thoughts?
Loading comments...