# Evaluating The Surprise Onside Kick

By Kurt Bullard

In the early stages of the intra-division battle between Chicago and Green Bay last weekend, Aaron Rodgers was having his way with the Bearsâ€™ defense. Â The former Cal Bear led two straight touchdown drives to start off the game and showed no signs of letting up. So, when the Bears scored with 3:50 left in the second quarter to take a 17-14 lead, Chicago decided to try to keep the ball out of Rodgersâ€™ hands for the rest of the half. Rather than kick it deep, Robbie Gould attempted a surprise onside kick. However, it was unsuccessful; the Packers were able to recover the ball at the Green Bayâ€™s 39. Rodgers would go down to score in less than three minutes, taking a 21-17 lead that Green Bay would not surrender.

While Chicago did not recover the kick, the question still remains: Is it ever worth it to go for a surprise onside kick? Anecdotally, surprise onside kicks have led to great outcomes â€“ the Saintsâ€™ bold onside kick to open up the second half of Super Bowl XLIV and the Eaglesâ€™ onside kick down 14 points which helped stage the Miracle at the New Meadowlands immediately come to mind. And, with the kickoff now at the 35, the penalty for a failed attempt is slightly less severe than it was in the past. But do these outcomes happen often enough to risk conceding great field position when no comeback is needed?

To answer this question, I looked at the expected points values of different outcomes off kickoffs. According to a previous post, the average starting field position off kickoffs that traveled at least to the opponentâ€™s 10-yard line is the 23.5-yard line since kickoffs were moved back up to the 35-yard line. The expected point value for a drive starting at this position is -0.51 points for the kicking team (According to Win Probability Calculator on Advanced Football Analytics). This is the baseline against which to measure the expected outcomes of onside kicks. There are two outcomes for surprise onside kicks â€“ recover the kick or concede possession. The expected value of the surprise onside kick is the probability of recovery times the expected value of possession plus the probability of a failed kick times the expected point value at that spot. Of course, the expected point value would be negative in each situation where the kicking team gives away position, as they are now defending.

Assuming that surprise onside kicks are as effective as they were between 2001 and 2010 as they are now, the stunt is successful 60% of the time. Weâ€™ll also assume that the average spot of recover is now the 50-yard line â€“ five yards behind the line of recovery. The expected point value from midfield is +2.04 points if the kicking team executes a successful kick.

Thus, the expected point value of a surprise onside kick would be:

E(Onside Kick) = P(Recovery) * E(Points|Recovery)+P(Failure) * E(Points|Failure)

E(Onside Kick) = 0.6*(+2.04) + 0.4*(-2.04) = +0.41 points

This expected point value is indeed much greater than kicking it away to the other team, which confirms my suspicion that surprise onside kicks have merit.

Obviously, teams should not try to execute â€śsurpriseâ€ť onside kicks every time they score. The element of surprise is not fixed â€“ it is dependent on the number of times teams try to catch their opponents off guard, as is evident by the success rate of regular onside kicks: 9%. Therefore, the next â€“ and key â€“ step in this analysis is to see how low the probability of executing a surprise onside kick would have to fall so that it broke even with kicking off to the other team. This is a relatively simple calculation â€“ just set the probability of success to x, and set the expected points value to -0.51 points. Solving this equation, we find that the probability of surprise onside kicks would need to drop to 37.5% to break even with the expected value of a normal kickoff.

Since the sample size for surprise onside kicks is so low, thereâ€™s no way to find out how frequent teams would need to try to surprise teams in order for the percentage to drop this low. But, until that saturation point is reached, one thing is clear: teams are missing out by not trying to throw in a little surprise into their kickoff plans.

View all posts