As you all know, the Texans' defense has reached into the crapulencesphere again. Assuming it ever left last year. Teams are targeting Kareem Jackson early and often, and he and the rest of the kiddie cornerbacks are playing off the line so often that the pass rush isn't getting much time to get pressure. Bernard Pollard has justified those of us who didn't want to give him an eleven year extension, and the wave of pain from DeMeco Ryans injury hasn't even been truly felt yet as the Texans avoided their base 4-3 against the Colts most of Monday night.
So in my booze-induced hangover, I reached out to Brian Burke of Advanced NFL Stats with some questions about onside kicks. He's already authored two rather conclusive posts about the subject that can be found here and here. My questions were simple, but the answers to them were anything but.
1) How high could a team stretch it's percentage of onside kicks without opponents starting to game plan for it?
2) For a team with a defense as consistently bad as Houston's has been, is it worth it to chase onside kicks even if their opponents did start to game plan?
3) Given the answers to A and B, how often would it make mathematical sense to attempt an onside kick for a team with a porous defense?
Brian graciously responded to my questions, although of course there's not much conclusive to say as neither of us really have the type of conclusive game theory experience you'd need to fully answer the questions. Here are his answers:
1. There's no way to tell except by experimentation. A coach would have to test the waters and see how often he can get away with it before the cat is out of the bag. Once %onside becomes part of the standard team scouting report, it would be very difficult to do it more than once every few games.
2. Probably not. Two points. First, the worse your defense the MORE you want to gamble to keep the ball. Consider this thought exercise: If you had the galaxy's worst defense that gave up a TD on every drive, it wouldn't matter where the opponent gets the ball, your 40 or their 30. They'd score either way. So the worse your defense, the less important field position becomes and the more important possession becomes.
Second, when an onside kick is not a surprise, the recovery rate drops well below the break-even rate where it makes sense.
3. Just a hunch, but once every few games. It should be an option for the kicking team, based on a read of the receiving team's front line. Watch how they do the first kickoff, then see how they line up on the next one to make a decision. The kicker should have the option to boot it deep if he doesn't see a good read.
So at the very least, with this defense that the Texans have constructed, Kubiak and company should probably be attempting an onside kick every few games. Asides from perhaps the first game you brought it out, it looks like you'd be limited to one attempt a game. I don't think you could attempt one every game without coaches starting to game plan for it, but perhaps every three or four games would be the point where you could attempt it as much as possible without a team gameplanning against it.
Sadly, this still leaves the Texans defense on the field too much for my liking, but as a Texans fan I am used to the harshness of reality.