website statistics
After a crazy reversed fumble out of the end zone, taking a touchdown away from the Redskins at the end of the first half, the referees hit Washington with two unsportsmanlike penalties to be assessed on the opening kickoff of the second half. That means Josh Brown would kickoff from the Washington 35-yard line, down three, with a full half of football to go. Surprise onside kicks convert around 55%, although you have to assume Washington anticipated one - and they did, as they put their hands team on the field. Defended onside kicks convert only around 15%. So, what was the break-even point to kickoff onside? Seems like a no-brainer, right?

A successful onside kick would result in a 67.1% team-strength adjusted win probability for the Giants. A failure would drop that to 56.4%. A touchback would mean essentially the same thing, 56.7% win probability. But, in all likelihood, the Giants could do a high pooch kick to force a fair catch inside the 10. So, let's assume the Giants could pin Washington at their own 5-yard line if they kicked off normal. That's a 58.1% win probability for the Giants, which we will use for our expectation and break-even point calculation.

E[WPOnside-Kick] = PRecovery * 67.1% + (1 - PRecovery) * 56.4% > E[WPNon-Onside Kick] = 58.1%

That makes the break-even point PRecovery = 15.9%, which is right around the expected conversion rate. 

Remember, though, we did this analysis assuming the Giants would be able to pin the Redskins at the five 100% of the time. This is a huge assumption, and we do not have any data to build a distribution since a kickoff from the opposing 35 is so rare. In reality, we would assume there is a 100% chance they could force a touchback. If we use the touchback win probability value of 56.7%, the break-even point drops to 2.8% - significantly lower than the defended onside kick conversion rate. In truth, the true WPNon-Onside Kick is probably somewhere between 56.7% and 58.1%, resulting in a break-even point between 2.8% and 15.9%.
So, yes, the onside kick was the right call. Though, if the Giants believed they could pin Washington inside the five, it is really a coin flip.

Keith Goldner is the Chief Analyst at numberFire.com - The leading fantasy sports analytics platform - and creator of Drive-By Football.  Follow him on twitter @drivebyfootball or check out numberFire on Facebook.

Leave your comments

Post comment as a guest

0

Comments (2)

  • Guest - Tank 53

    Who cares!!!

  • Guest - biggerzhang

    The Cowboys would not face another third down on the drive, converting two more first downs before Romo connected with Terrance Williams on a quick slant for the opening score of the game. More Info: http://www.replicahandbagsshop.co.uk

    Before exiting the game for Brandon Weeden to get some garbage time reps, Tony Romo posted a +0.33 WPA, an 80.0% pass success rate and 90.0% completion percentage (18 completions on 20 attempts). Heading into the playoffs, the Cowboys offense is firing on all cylinders, Romo in particular. Over his past three games, the Cowboys' quarterback has averaged +0.29 WPA per game.