In the age of advanced analytics, it makes no sense why onside kicks are seldom used

The onside kick. It's one of the most polarizing and exciting plays in all of sports. If done correctly, it can turn the momentum of the game on its heels, as demonstrated by the Saints in Super Bowl XLIV. But on the opposite dime, it can destroy a teams' chances by giving the opposition prime field position.

According to my count, 66 onside kicks were attempted last season in the NFL. Only nine of those panned out. That's about a 14% success rate. Given that sad statistic and the tendency of the onside kick to backfire, it's no wonder why we seldom see the play called upon, and only used as a last resort in the dying moments of a game.

By nature, coaches, especially at the highest echelons of their sport, tend to be creatures of habit who trust their gut first. If your judgement is based off of past success, it makes no sense to risk an onside kick unless absolutely necessary. After all, if you watched all of your teams' games last season, chances are that you never saw a successful onside kick. So why risk it?

But times are a-changing in the NFL, and with touchbacks now being moved up to the 25-yard line and the ball being kicked from the 35, the onside kick makes more sense now than ever.

The odds that an onside kick will work out for the kicking team are about 1:4, as the success rate is 26 percent. But that's a little misleading, because the other team is expecting it. When you reclassify the data and identify scenarios when the opposition wasn't prepared for the trick play, the success rate jumps to a much more appealing 60 percent.

As Brian Burke of Advanced Football Analytics explains, the probability of an onside kick succeeding is directly tied to the win probability at that point in a game. A large majority of onside kicks are attempted when a team's win probability is below 15 percent, but that also happens to be when the opposition is expecting it, as the success rate is about 20 percent when win probability is that low. But if you up the win probability a few notches to 40 percent, the success rate jumps to almost 70 percent.

(Advanced Football Analytics)

It's these stats that have Kurt Bullard of the Harvard Sports Analysis Collective arguing for coaches to pull out the surprise onside kick more often. With the conventional kick-off now much less appealing thanks to the new rules, Bullard thinks that the onside kick looks better than ever from a statistical standpoint.

"I found that teams should attempt them if they believe they have at least a 37.5% chance of recovering it, meaning that teams should be more aggressive—the reward is greater than the risk if the recovery rate is 60%."

While we rarely see onside kicks, let alone surprise ones in the NFL, one high school coach is a big believer in the onside kick, and it's brought his team great success. Kevin Kelley of Pulaski Academy in Little Rock, Arkansas, has become well-known for relying on statistics and forgoing punts. His squads have only punted four times in the past three years, and also opt to take onside kicks after almost every touchdown. The strange strategy has brought the small school five state championships in football since Kelley took over in 2003.

But if the past is an indicator of anything, it's that change takes a long time in football. While other sports have evolved to integrate metrics and models into personnel decisions and strategy, football still seems fixated on intangibles and attributes. Even with the rule changes that are favorable to onside kicks, I still think that the play will be taboo to coaches.