Does it make sense to start overtime with a surprise onside kick?

The stadium scene.
Jan. 7 2011 11:10 AM

So Crazy, It Just Might Work

A pro football conundrum: Does it make sense to start overtime with a surprise onside kick?

Football on a tee. Click image to expand.

Overtime in the NFL used to be simple: The first team to score won the game. Starting with this year's postseason, it's gotten a bit more complicated. Last March, NFL owners voted to change the rules for playoff games exclusively, arguing that the team that won the overtime coin flip received an unfair advantage. In the past decade, as field goal kickers have become more accurate from long distances, the coin-flip winner has won 60 percent of overtime games. In only two-thirds of overtime games does each team even get an opportunity to possess the ball.

The owners made just one modification to the overtime rules: A team cannot win by scoring a field goal on the first possession. This simple change has a lot of strategic implications. If a team faces a long field goal opportunity on its first possession, should it try the kick or go for a first down? (Keep in mind that a field goal doesn't have the same value as it used to.)If it's fourth-and-long, should you punt from the opponents' 30 rather than try a long field goal? (The right answer depends on weather conditions and the quality of your defense, but an average team with an average punter should probably pin the opponent against its goal line.) Down by three points in overtime, should you match the field goal to prolong the game or try to end it with a touchdown? (Don't forget that a tying field goal puts you on defense in sudden death, which is a sizeable disadvantage.)

Another interesting strategic question has to do with onside kicks. As the site Pro Football Talk explained on Thursday, the overtime rules stipulate that "each team is only guaranteed the opportunity to possess the ball." On account of that, the onside kick offers a small loophole to the team that's forced to kick off at the start of the overtime period. If the kicking team successfully recovers an onside kick, the game reverts to the old sudden-death format—all the recovering team has to do to win is kick a field goal.

It seems unlikely that we'll see an overtime onside kick this weekend. As New York Jets assistant coach Mike Westhoff told the New York Times, "Boy, that takes some serious nerves to do that." The fact that an onside kick is so unlikely, however, means that it's likely to work. When it comes to onside kicks, the likelihood of success depends heavily on the element of surprise. Over the past 10 seasons, surprise onside kicks—defined as when the kicking team, based on win probability statistics, has a better than 20 percent chance of winning at the time of the kick—are recovered around 60 percent of the time. Expected onside kicks—those that come when a team obviously must resort to an onside kick and the receiving team can plan accordingly—succeed less than 20 percent of the time.

The principle of win probability allows us to calculate when an onside kick is worthwhile. Win probability is, very simply, an estimate of how often a team in a given game situation will win a pro football game. For example, under the conventional sudden-death rules, a team receiving the kickoff will win 60 percent of the time. A team with a first down at midfield will win 68 percent of the time. These estimates are based on winning percentages from similar situations in actual NFL games.

Because the new rules are, well, new, there are no previous examples to use as a baseline. But we can make reasonable estimates by looking at the typical touchdown and field goal rates from the various yard lines. We can also look at instances when a team is down by three points on its final drive of the game but is not too rushed by the clock—roughly analogous to the situation where a team is trying to respond to an overtime field goal.

So, does it make strategic sense to start overtime with an onside kick? Under the old sudden-death format, the break-even success rate for an onside kick is 30 percent. In other words, a team will increase its probability of winning the game by attempting an onside kick so long as its chance of recovering the kick is 30 percent or greater.

Under the new rules, the price of failure isn't nearly as high—even if you don't recover the kick you'll have a chance to match or beat an opponent's field goal. Plus, a kick recovery reverts the game back to the old rules, where just one field goal ends the game.

Even so, an opening onside kick is actually a worse percentage play under the new system—the break-even rate has increased to 40 percent. Although it's true that failing to recover isn't as costly as it was before, a conventional deep kickoff has become even more valuable. If the kicking team stops the receiving team deep in its own territory, after all, it will get the ball back in good field position with a great chance to end the game with a chip-shot field goal.

Though the new rules have made overtime onside kicks less advantageous, the numbers still suggest that an onside attempt would be a smart play. Never once has a team opened an NFL overtime game with an onside kick. With the element of surprise, the chance of recovery should be around 60 percent—well above the 40 percent break-even point to make the onside kick a sound decision.

Another intriguing possibility would be an onside kick following a first-possession field goal. I think this would be even more surprising than an onside attempt on the opening kickoff—everyone on the opposing sideline will be deep in thought about what strategy they should use down by three points in overtime, a situation no one has ever seen. In this case, a successful recovery would end the game immediately. And even if the receiving team recovers, the kicking team can still give up a field goal and get the ball back; at that point, the game reverts to old-fashioned sudden-death overtime, where you get the ball and have the advantage.

Before doing the analysis, I was hoping this would be a slam-dunk easy call. However, the break-even recovery rate works out to be just under 40 percent, about the same as for the start of the new overtime. (The reason it's not more advantageous is that a deep kickoff is also very valuable when up by three points.)

The key in any surprise onside kick is reading the receiving team's tendencies. If they've been dropping back to set up blocks or if they're standing too far back—as the Giants were against the Eagles a few weeks ago—then an onside kick is worth the gamble, so long as you trust the numbers. But would any coach be willing to risk his reputation on such an outlandish maneuver? Well, an onside kick helped win last year's Super Bowl. Maybe it's time for it to win a playoff overtime game.

Like Slate on  Facebook. Follow us on  Twitter.

Brian Burke is a former Navy fighter pilot and founder of Advanced NFL Stats. He regularly contributes to the New York TimesFifth Down and the Washington Post's Insider.