NFL Coaches Need to Stop Delaying Their Two-Point Conversion Tries

Cody Kessler completed a five-yard touchdown to Terrelle Pryor with 2:07 left in the game, making the score 28-19. The Browns trailed the Titans, but an extra point and a successful onside kick would give Cleveland the ball -- and some hope -- in a one-score game.

But, in a nine-point contest and one where the Browns needed another touchdown anyway, Cleveland opted to go for two. They didn't want to wait -- they wanted to convert right then and there.

The conversion failed.

The Browns then went and recovered the ensuing onside kick, and they actually scored another touchdown with 30 seconds left on the clock. The final score was almost meaningless, though, because it was a nine-point game instead of an eight-point one.

Twitter was pissed.

Why would head coach Hue Jackson go for a two-point try when they needed another touchdown to tie things up anyway? Why not just kick the extra point? Why not try to extend the game?

Well, here's the thing: this idea of "extending the game" with point-after-touchdown kicks when a team will eventually and undoubtedly need a two-point conversion doesn't make any sense. It's not logical. And, yet, color commentators and football analysts alike tell you that it's an accepted football strategy. That teams should be extending games with kicks and not runs or passes.

But they shouldn't. Teams actually should be doing what the -- gasp -- Browns did.

The Logic

The reason you, a football coach calling the shots, would want to go for two as early as you can -- assuming you know you'll eventually have to go for two -- is actually pretty straightforward: you want to have as much information as possible.

Let's say the Browns decided to kick the extra point rather than going for two in the situation described above, which happened back in Week 6 of the 2016 season. Assuming they make the kick, they'd then be down eight, or a one-score game. So, when they scored the touchdown at the tail-end of the contest, they'd have an opportunity to tie things up and head into overtime with a two-point conversion.

Their probability of converting the two-pointer late in the game versus earlier in the game, though, is no different.

Going for two earlier actually opened up the ability to know more about the game for Cleveland. In other words, if your team is going to fail an inevitable two-point try, they might as well fail earlier rather than later. They might as well have an opportunity to make up for their mistake.

What happens if the Browns, after the Terrelle Pryor touchdown, think they need to score just once rather than twice to close out the game? The latter calls for aggressive play-calling in order to score with time remaining, since you need two scores. The former says, "Hey, we just need a touchdown here, so let's score with as little time left as possible to try and force overtime with a converted two-point try."

And there's the problem. The two-point conversion probability isn't any different, so if it's assumed the Browns will fail (which they did earlier in the game), all of a sudden, the contest is over. It's done. Cleveland has no way of trying to make up for the failed conversion.

Going for it after the Pryor touchdown allowed them to score a touchdown with time remaining, knowing they still needed another score. Sure, they didn't win the game, but math tells us that they wouldn't have won the game anyway by waiting forever to go for two.

Strong Assumptions

One of the other things that drives me nuts about "extending the game" is that there's this assumption that the point after touchdown try is automatic. With the recent rule change in moving the extra point kick back, that's not the case. Take a look at the conversion rate of an extra point try since the start of the 2015 season:

A 94% rate means teams, on average, should expect 0.940 points to be added to their score when their kicker runs onto the field for an extra point try. Naturally, some teams have great kickers, while others have Roberto Aguayo. So this isn't the case for all NFL squads -- it's just the mean.

What's interesting is that, since the start of 2012, two-point tries have actually been pretty awesome in the NFL. So awesome that the expected point value is higher for them versus what we see from a point-after kick.

A conversion rate of 49.75% translates to an expected point value on a two-point conversion try of 0.995 points. (Remember, it's two points instead of one, so you've got to multiply the conversion rate by two.) And, no, converting in the fourth quarter isn't higher than at any other part of the game. Get your momentum argument out of here.

This is no end-all to the "go for two" debate. That's not the intent with these charts. Instead, it's to show that going for two instead of one actually isn't a bad move regardless of what's at stake.

Going for two in today's NFL is actually a mathematically-savvy move.

Timing the Conversion

The biggest question mark around this, to me, isn't whether or not you should go for two as early as possible, but when "early" actually is.

If a team, say, is down 12 points after scoring a touchdown in the middle of the third quarter, opting for the kick isn't a big deal because there's still a good chunk of time for the score to change. If it's good, a field goal by the opposition creates a 14-point game, which is an obvious two-touchdown deficit.

But late in the fourth quarter when a team knows they're going to have to convert a two-point try eventually? You do it as soon as you can, because you want to get as much information as you can.