Independent events

Let's discuss briefly the Jets' end-game strategy during this past Sunday's NFL contest in New Jersey, won by the Browns.

Many people are criticizing Jet coach Eric Mangini's decisions to kick field goals and attempt on-side kicks instead of trying to convert 4th-and-long situations late in the game. Personally, I think Mangini's decisions were OK, except that out-of-bounds deep kick at the end was kind of weird. Getting that field goal with two minutes left to pull within a touchdown was not a bad strategic move, nor was kicking prior to that instead of trying to convert a 4th-and-10 with Kellen Clemens at QB. You could argue that playing for the touchdown would have been the right move, and I think you also would have been right. To me, both were reasonable strategies - I probably would have done it like Mangini but would be OK with a coach going the other route.

I'm not here to discuss that, though. No, I'm here to set the record straight on onside kick probabilities and how Mangini's choices should be evaluated. A number of people at my Browns club during the game and many of the people criticizing Mangini since the game have talked up "the odds" of recovering two onside kickoffs in one game or, as one blogger put it, having lightning strike twice. This is poor analysis.

The chances of a successful on-sider are roughly 1 in 4 in the NFL. So, starting from scratch, the odds of getting two in a row are about 1 in 16 (6.3%). That is not good. Considering the Jets actually tried three kicks, their chances of getting two were closer to 11%. Still not especially good, but in this discussion, not especially relevant either.

You see, those criticizing Mangini for trying a second kick, both at the time he elected to do so and after the game, failed to realize that the first kick had already succeeded and had no bearing on future on-side recovery probability. People kept saying, "what are the odds of getting two onside kicks in a row?" and I kept saying, "It doesn't matter - they've already gotten the first one so now all that matters is the odds of getting one again, still 1 in 4." Once lightning has struck, the odds of it striking twice are the same as they originally were for it hitting once.

The point is that the second (and third) onside kicks were independent of the successful result of the first. Likewise, a failed first try wouldn't have meant they were "due" or had any enhanced probability of getting a second attempt, nor should it have deterred them if a second one seemed like the right strategic move. The whole notion of athletes being "due" is entirely incorrect - the past does not strongly affect the present. Fans and columnists may criticize Mangini for trying a second short kick if they wish, but they cannot frame it in the context of trying to grab two on-siders, since the first had already been executed successfully.