When NOT to score - Pittsburgh at Green Bay - December 22, 2013

Outside of sportsmanship purposes, is it ever right to NOT go for the most expected points?  Should there be a time when you tell your team to NOT score a touchdown?  If that's true, then wouldn't the defensive team WANT you to score a touchdown?  This scenario came up again in Week 16 of 2013 with the Pittsburgh Steelers visiting the Green Bay Packers at Lambeau Field.

The Situation

Pittsburgh and Green Bay are tied up, 31-31 in the 4th quarter.  On the first play after the 2 minute warning, Steelers safety Troy Polamalu forces a Matt Flynn fumble, recovered by Brett Keisel who returns it to the Packers' 17 yard line.  

The Steelers now have the ball with 1:51 left in the game.  The Packers have all 3 timeouts remaining.  Knowing that a score is highly likely, the Packers proceed to use their timeouts in an attempt to get more time on the clock for a comeback once the Steelers score.

The situation is a favorable one, as they force a 4th and 3 at the 10 yard line with 1:35 left in the game and the field goal team trots out to attempt a 28 yard field goal.  A field goal here would give the Packers over a minute left to tie or win the game and the Packers still have 1 timeout left.  However, an encroachment penalty gives the Steelers an automatic 1st down at the 5 yard line.

At this point, the question is: Should the Steelers intentionally not score until the last second, so they prevent any chance at a Packers comeback?  If this it true, then should the Packers intentionally allow the Steelers to score?  Remember, what's bad for my opponent must be good for me.

What should the Steelers do in this case?

Not Scoring to Win the Game

In order to see what the best course of action is, we need to make a couple of assumptions to simplify the calculations.

1. If the Steelers intentionally NOT score from the 5 yard line, they will kneel down three times and attempt a field goal try.  Alternatively, they may choose to kneel down twice in case of a bad snap on 3rd down and can retry the FG attempt.  The Packers have 1 timeout left, so the three-kneel sequence would take off roughly 90 seconds from the 1:35 clock, giving the Steelers enough time to try just one play.
2. If the Steelers do try to score a touchdown, they will do so on the next play, giving Green Bay the most time possible to mount a comeback.  The ensures that scoring a touchdown early will give the Packers the best case scenario as far as time is left on the clock.
3. If the game somehow goes to overtime, both teams have a nominal 50/50 chance at winning the game.

Kneeling Down then Kicking a FG

Kneeling down and settling for a FG attempt would take the clock down to basically no time.  This would ensure a victory.  But wait!  Didn't we just talk about FG attempts being just that, attempts, and not guarantees?  

That of course is true.  There are no guarantees for field goal tries.  There may be a bad snap, a gust of wind, poor blocking, or simply a kicker shanking a kick, but it is NOT 100% that a kicker will make a field goal.  We will avoid binary thinking and instead assign a probability to this kick.

We will assume the Steelers will lose a yard on each kneel down, so the try will be snapped from the 8 yard line, making this a 26 yard field goal attempt.  A field goal try from between 25 and 29 yards is good 94.4% of the time on average.  The conditions at Lambeau Field were poor for kickers, with swirling winds and snow falling, so let's say it's a 90% chance of making it.  We will also look at changing this number later.

Also, even if the kick is missed, the game goes into overtime, where the Steelers still have a chance of winning the game.


The Win Probability (WP) therefore is:

Probability of making the kick * 100% + Probability of missing the kick * 50%

WP(kneeling then kicking) = (.90)(1) + (.10)(.5)

WP = .9 + .05

WP = .95

The Steelers still have 95% chance of winning this game if they kneel down 3 times then kick the field goal.

Trying for a Touchdown

Scoring a touchdown would force the Packers to score a touchdown of their own just to tie the game and send it into overtime.  If the Steelers score right away, and leave 90 seconds on the clock, Green Bay will get the ball back after a kickoff with 1 timeout remaining.

Assuming a squib kick and a generous return to the 35 yard line, the Packers have an 11% chance of winning the game according to the Advanced NFL Stats WP calculator.  This number does take into account the 50/50 chance of winning in overtime, which is basically the Packers' only hope.  This leaves the Steelers with an 89% chance of winning the game if they score a touchdown right away!

Scoring a Touchdown after taking more time off the clock

Taking more time off the clock exacerbates the problem for the Packers.  With 60 seconds left and a 7 point deficit, they have only a 7% chance of winning.  Clearly, the more time off the clock is better for the Steelers.  Also, if they score with no time remaining, it doesn't matter whether the score was a touchdown or a field goal.

Conclusion

It's clear in this particular situation with these particular numbers that kneeling down to take time off the clock is more beneficial than scoring a touchdown.  The difference in win probability was "only 6%" in this case.  However, if you extrapolate 6% out over the course of a season, that works out to 1 extra win per year, something the Steelers certainly could use right now.

In the actual game, the Packers did nearly come back to tie up the game with a 1:25 drive which fell just short.  It looks like the Packers did in fact allow the Steelers to score intentionally to get the ball back with as much time remaining as they could.

When asked about this situation later, coach Mike Tomlin replied, "I'm not into that. Given an opportunity to score, we are going to score." 

Alternate Situations

Let's look at a different situation to see at what point would it be right to score the touchdown here.  At what probability for a field goal try would it be right to score the touchdown rather than kick the field goal in a kneel down situation.   

Recall the WP for the Steelers if they scored the touchdown was 89%.  The WP for the Steelers if they tried a field goal was found with the formula: P(FG)*(1) + P(FG missed)*(.50).  Making the unknown variable the chance of a field goal and then setting the two formulas equal to each other to find the break-even point gets us this equation:

x + (1-x)*.5 = .89

x + .5 - .5x = .89

.5x = .39

x = .78

In other words, for it to be right to score a touchdown in this spot instead of kicking the field goal, the kicker would have had to make the kick less than 78% of the time.  NFL kickers are simply much more accurate than this at the short distances we're talking about.

Comments?

Feel free to leave your comments.  Think I'm wrong?  Put in your own numbers and let's see what we come up with.