In Super Bowl XLIV, New Orleans Saints head coach Sean Payton made the gutsiest of decisions, instructing his team to conduct a surprise onside kick coming out of halftime. The Saints recovered the kick and ended up riding that momentum to a Super Bowl victory.
Now, the call was really only "gutsy" in that it defied conventional football wisdom. The safe and traditional play in the NFL is to kick the ball deep, play great defense, and force a punt. But that doesn't make it the right one.
In statistical terms, surprise onside kicks aren't all that risky. Since 2000, teams that are leading games recover onside kicks around 55 percent of the time. With a lead, any onside kick is unexpected, and thus easily classified as a "surprise."
Now, the question is whether or not a 55 percent recovery rate is worth the risk of giving up field position. When a team fails to recover an onside kick, they're basically giving away about 30 yards of field position for free.
Well, it turns out that the value of possessing the football is far, far more valuable than 30 yards of field position (especially 30 yards on the opponent's end of the field). Teams that return normal deep kickoffs (and thus start around the 20-yard line) typically score around 0.8 points-per-drive. That number jumps to around 2.0 points-per-drive if they recover an onside kick at midfield (i.e. the kicking team fails to recover their surprise attempt). That's a pretty big jump.
However, don't forget that recovering a surprise onside kick attempt is basically "stealing" a possession. With a 1st and 10 at midfield, the kicking team that recovers has the same 2.0 point-expectation as their opponent. Since the expectations are the same and attempting an onside kick allows the kicking team to potentially forgo giving up possession, it can be extremely valuable. Actually, with the points each team can be expected to score/give up in each situation (a deep kick, an onside recovery, an onside failure), attempting an onside kick is statistically a smart move if you anticipate a 40 percent recovery rate.
Let me be the first to say that head coach Jason Garrett's decision to attempt an onside kick with 2:51 remaining in the first half was a brilliant move. I absolutely loved it. Garrett's decision was every bit as good as Sean Payton's onside kick attempt in the Super Bowl, and that the Cowboys didn't recover does nothing to change the fact that it was a smart call.
In the NFL, coaches and teams are often wrongly criticized based on whether their decisions work out. Many will point to the failure of a fourth down try or, in this case, a surprise onside kick attempt and use it as justification that a poor choice was made. Similarly, Payton was labeled as a genius for his decision in Super Bowl XLIV. You think that would have been the case had the Saints not recovered the kick?
As an NFL coach, Garrett's job isn't to have his team never fail, but rather to maximize their chances for success. If a particular offensive call doesn't work, that doesn't necessarily mean it was a poor one. Similarly, a positive play can result from a low-percentage start. Remember the2007 Shotgun snap that traveled over Tony Romo's head, bouncing 25 yards behind the line of scrimmage before Romo scooped it up and somehow eluded enough defenders to run for a first down? Anyone want to try one of those snaps again simply because it worked one time? Of course not. We want the Cowboys to consistently play in situations that have a high probability of working out in their favor.
Garrett's decision to call a surprise onside kick certainly maximized the Cowboys' chances of winning on Sunday, regardless of the outcome. The kickoff team undoubtedly had over a 40 percent chance of recovering the kick, making it the statistically correct call. Again – I can't emphasize this enough – the fact that it didn't work out this time doesn't make the decision a poor one.
On top of that, the timing was absolutely perfect. Looking back at surprise onside kick attempts since 2000, the ones that lead to the greatest winning percentage (by far) are those just before halftime and at the end of the game.
When the Cowboys tried their onside kick with 2:51 remaining before the break, the team was in a position in which they'd likely be facing two straight possessions from the Buccaneers. If the 'Boys kicked deep and Tampa Bay got at least one first down, chances are that Dallas wouldn't have gotten the ball back in the first half (or else without enough time to do significant damage). Meanwhile, the Cowboys were scheduled to kick off to the Bucs again to start the second half.
By attempting an onside kick with fewer than three minutes remaining in the half, Garrett was attempting to truly "steal" a possession. Had the Cowboys recovered, they would have had plenty of time to move down the field, but not so much that Tampa Bay would have also been able to mount a scoring drive before halftime.
You can be upset that the onside kick didn't work out, but try your best to realize that, despite the result, the call was a magnificent one. As crazy as it may seem, the odds were in the Cowboys favor in a big way on the kick, from the stats to the score to the time left on the clock. It didn't work out, but the style of coaching it exhibited, one of aggression and maximizing win probability, will lead to good things for the Cowboys down the road.