Rethinking late game strategy in the NFL

NFL teams are engaged in an understandable long term battle between statistics and tradition. For example, there is a small statistical advantage to going for two after every touchdown, but nobody does it because if you buck tradition like this and lose — you don’t have a job. The risk is too high to justify the .02 expected points per attempt.

But, in one limited circumstance, the traditional way of thinking appears to be considerably inferior, and it is incredibly rare to see any team try the opposite.

The scenario: your team trails by two touchdowns (14 points) with roughly four or fewer minutes left in the game. Your goal seems pretty straight forward, score a touchdown, kickoff, get a defensive stop while burning through time outs, and then get one more chance to drive down the field and tie it up. If it works you take your chances in overtime.

The most likely outcome of course, is that you aren’t able to get both touchdowns. But, what if you do?

Traditional strategy would suggest you kick the extra point on the both touchdowns and go for the tie. But, in a world where people obsess over the tiniest edge, a relatively huge advantage can be gained by a team with the guts to go for the two point conversion on the first touchdown.

Let me start with the details of each plan — if you don’t care about them, scroll down to the summary in the big letters.

Kick it old school:

There’s about a 94% chance you make each individual extra point, which works out to an 88% chance you make both. This gets you to overtime, where you have a 50% chance of winning. Combine it all, and you have a 44% chance to win the game.

There’s another scenario in which you can win. If you miss the first extra point, (a 6% chance), you would then go for two on the second touchdown. If you succeed, you would also go to overtime. This would result in a win in an additional one percent of scenarios.

Overall chance of winning kicking the extra points: 45.6%

Go for two:

If you go for two on the first touchdown, you only have to make one of the two attempts to tie. Let’s say you miss the first attempt, you would have a 48% chance of making the second to tie the game, sending you to overtime. This would lead to a win in a little over 12% of scenarios.

But, the real advantage comes if you are able to make the first attempt. Then, you would then only trail by six points, and you would kick the extra point on the second attempt. Do the math, and this scenario would happen about 45% of the time. That’s slightly more than the success rate of tying the game by kicking the extra points. However, in this scenario you are ahead, not tied. There is no risk of going to overtime and losing.

The one remaining possibility is making the first two point conversion, then missing the extra point on the second touchdown. But, in that instance you would still be tied, leading to a win a little over one percent of the time.

Overall chance of winning going for two: 59.0%

To summarize:
Overall chance of winning kicking the extra point: 45.6%
Overall chance of winning going for two: 59.0%

It’s difficult to overstate how huge of an advantage this is in modern sports. This is not some minuscule theoretical edge that would play out only over multiple decades. This is a no brainer.

It’s such a no brainer that it was even true by a large margin when extra points were being kicked from the two yard line, before the 2015 rule change. It’s even more true today.

It’s a beautiful day for a ballgame … Let’s go for two.

Some disclaimers:

Yes, this is just back of the envelope math.

Real world variables would change the numbers, but not enough to overcome a double digit percentage.

Smarter people will surely tell me why all of this is completely wrong.

Yes, I’m a loser for obsessing about such a stupid thing, but it drives me nuts.

Here are the breakdowns:

Extra Points:

Scenario 1: A) Make XP1, B) Make XP2, C) Win in OT = 44.2% chance of win

Scenario 2: A) Miss XP1, B) Make 2PT2, C) Win in OT = 1.4% chance of win

Total chance of winning: 45.6%.

Go for two:

Scenario 1: A) Miss 2pt1, B) Make 2pt2, C) Win in OT = 12.5% chance of win

Scenario 2: A) Make 2pt1, B) Make XP2, C) Win in regulation = 45.1% chance of win

Scenario 3: A) Make 2pt1, B) Miss XP2, C) Win in OT = 1.4% chance of win

Total chance of winning: 59.0%.