Revisiting the Extra Point: Should Teams Be Going For Two?

With a year of extra points taken from the 15-yard line, does the data suggest teams should be going for two?

With the 2015 regular season coming to a close, it's time to revisit one of the biggest rule changes to the game of football in recent memory. 

Before the season, we wrote about whether we thought teams should still be kicking the extended extra point after every score. The analysis used a model to predict average field goal accuracy based on distance, adjusted for the ever-increasing accuracy of today's kickers.

We projected kickers would make the new 33-yard extra point at a rate of 93.58%, making a team expected to earn 1 * 93.58%, or 0.9358 points, per attempt. With a historical two-point conversion rate of roughly 48%, that makes the expectation on a two-point conversion 2 * 48%, or 0.96, which is higher than the expectation on an extra point. 

All said, we posited that, while the decision was close, going for two is the optimal choice in most cases. But this can obviously change depending on the efficiency of both the team's kicker and offense. 

Based on this analysis, though, we should expect the average team to be going for two the majority of the time. Yet coaches see an extra point as automatic, and a two-point conversion as "risky," so how would decision-making change?

With a full year under our belt, we no longer have to use a projection for the extra point conversion rate. In 2015, there were 1,217 extra points attempted, 1,145 of which were made, good for a 94.08% rate of success (just above our projection). 

Compared to 1,217 extra points attempted, there were just 91 two-point conversions attempted. That means coaches decided to go for two only 7% of the time (up slightly from 4.6% last year). Of those 91 attempts, 45 were successful, or 49.45% of them (just above the historical conversion rate). 

So based on these numbers, the expectation kicking an extra point is 0.9408 compared to 0.9890 when going for two -- a team going for two every time would earn 0.05 more points per touchdown. On average, teams scored 41.19 touchdowns this season. So, the average team would have scored approximately two more points on the season with optimal decision-making. 

What could those points have done for your team?