How Do You Fake Onside Kick?

When a team is losing by 9-to-11 points late in the 4Q, why is the trailing team's strategy to drive into FG position, kick it, and then try an onside kick?

• Every NFL announcer says this. It seems to be a mantra they use for meditation at broadcaster school. My point is: If you are close enough to kick a FG, aren't you already in a better position to score a TD than you would be with an onside kick? Score the TD first, then try for the onside kick and FG seems a better choice. An onside kick when expected has less than a 20% chance of success (but that stat is irrelevant to my question.) The kicking team kicks from its own 35. The kick must travel 10 yards before the kicking team can touch it, so it will be recovered at no less than its own 45. Let's be generous and say they get it at the 50. So now the team that kicked the FG must go 50 yards for a TD. When they kicked the FG they were probably at least 15 yards closer to the goal line. Is this a logical strategy? Is this strategy a disturbance in the force caused by Vegas? Enquiring minds want to know.

• Answer:

  Good question, and one that as far as I can tell hasn't been looked at critically by my favorite analysis sites such as http://advancednflstats.com .  In fact, this happened just this season with the Patriots down 10 to the 49ers at the 2 minute warning.  The Pats drove to the SF23 in 11 plays and 78 seconds, leaving 38 seconds.  They kicked a 41-yard FG and did not recover the onside kick, to lose the game.  If "coaching genius Belichick" does it, it must be right, right?  Hmmmmm. First of all, we're probably arguing about fractions of win percentages - the Win Probability is less than 1% either way.  This graph of Win Probabilities for the NE-SF game I mentioned shows that: http://live.advancednflstats.com/weekly.php?gameid2=55726&week=15.  That was also true for the NE-MIA game in week 13:  http://live.advancednflstats.com/weekly.php?gameid2=55686&week=13  So in a sense, it may be "6 of one, half dozen of the other."  Either way, you need two scores.  Either way, you need to recover an onside kick with enough time for the second score. I think these are the major factors that say kicking a FG late in the 4th quarter is the barely better strategic decision: You can score a TD from any down and distance, but you have to be somewhat close to score a FG.  That means that scoring a TD with tens of seconds may mean you can't win, because you need to advance the ball after the onside kick into field goal range.  Whereas scoring a FG with only tens of seconds left means you still have a chance to score from far out. You need to leave enough time for the second score.  That suggests that if you're close enough to make the first score, and time is running out, you need to take it.  Otherwise, you're going to use up a variable amount of time getting that first score, leaving possibly not enough time for the second score.  How long would it take to kick that FG after an onside kick?  Well, the onside kick: ~5 seconds, and a 15+ yard pass play out of bounds (or with a timeout) to get within FG range from midfield?  Another ~5-10 seconds, if you get it on the first try.  So at least 15 seconds, and more likely about 30 if you want more than one shot. How long would it take to score a TD after an onside kick?  Lots of possibilities here, but a Hail Mary play is at least 10-15 seconds, and with enough time there are opportunities to get some of the yardage in pieces so that you can leave a higher percentage chance to win at the end (think Tony Romo's tying throw against the Saints with no time left in Week 16, after driving to the NO19: http://live.advancednflstats.com/weekly.php?gameid2=55731&week=16) So look at the NE-SF game again.  Belichick and Brady were driving for a TD, but once they got in range and realized that they only had 40 seconds left, they realized that they couldn't afford to keep trying for a touchdown if they wanted to be sure to have time for a second score.  They might be forced to be trying for a TD after the onside kick anyways because they would be too far away for a FG with the time remaining.  So they took a high-percentage FG instead, giving them 20+ precious extra seconds to attempt to get that TD after the onside kick.    Coach Philbin made a similar decision playing the Pats in Week 13, kicking the FG with 41 seconds left before failing to recover the onside kick. The primary counterargument to this is that if you score a touchdown, there's still a chance you can actually win the game, instead of forcing overtime, by scoring a second touchdown.  But again, the likelihood of that happening is so small that it's almost not worth considering in the analysis. The other counterargument is that you need to score a touchdown at some point anyways, aren't you better off trying to get it from your FG-range field position than from the 50 after an onside kick?  Again, the question is time.  At some point the coach has to make sure there's reasonable time for a second score, and the break even point for today's coaches seems to be with ~40 seconds left. Other discussions I found online about this topic: http://www.thebiglead.com/index.php/2010/09/29/kicking-the-late-field-goal-down-by-two-scores/ http://forumserver.twoplustwo.com/47/science-math-philosophy/dolphins-kick-field-goal-2nd-10-a-1273777/