Editor's note: This story originally ran on Nov. 11, 2019. We're bringing it back after the Colts tried it when they were down 8 points against the Bills in their AFC wild-card matchup on Saturday. The move raised eyebrows, but for others it was a high-stakes embodiment of a longstanding strategy in analytics circles. Going for 2 down 8 late in the game is generally the correct move. And more and more, it's a move head coaches are willing to make.
An NFL team is down by 14 points in the fourth quarter, and it scores a touchdown to pull within 8. The consensus says the team should kick the extra point, right? Not so fast.
In analytics circles, attempting a 2-point conversion when down 8 points late in the game has long been discussed as a slam-dunk decision that NFL coaches should make. Recently, it has become a choice that some coaches actually do make.
Between 2000 and 2017, there were only two instances in which a team was down 14 points, scored a touchdown to cut the deficit to 8 and went for 2 on purpose (excluding aborted PAT attempts). And one was in a snowstorm, when kicking can be adventurous. Since the start of 2018, it has happened 12 times, including the Colts' attempt on Saturday.
So what is going on, and why do advanced statistics favor this strategy? Here is your guide to a trend that is gaining steam in the NFL:
Start with the basics: Why should teams attempt a 2-point conversion down 8 points?
Because the odds of converting a 2-point conversion once in one try are higher than failing twice in two tries.
Assuming the team is going to score another touchdown -- none of this matters if it doesn't -- and prevents the opponent from scoring again, going for two now gives it an informational advantage that indicates what to do after its next touchdown.
If the team converts on the first 2-point try, it knows it only needs a PAT on the second touchdown to win.
If it fails, the team knows that it has to go for 2 again the second time to have a chance to tie.
Kicking two extra points maximizes the chance to reach overtime. Attempting a 2-point conversion when down 8 points late maximizes the chance to win the game, which is the ultimate goal.
Has this ever worked in the NFL?
Only two teams ever have gone for 2 when down 8 and come back to win the game, according to Elias Sports Bureau data. One of them was the 1998 49ers, but that 2-point attempt was unintentional. According to the play-by-play, a bad snap on a PAT resulted in a failed rush attempt. The other was a snowstorm game in 2013, when the Eagles attempted a 2-point try on every touchdown they scored.
The reasons it hasn't worked:
By definition, the team that attempts the 2-pointer is losing by 8 points late in the game, which is already a long shot.
Most importantly, there just aren't that many teams that have tried it.
That doesn't mean the numbers behind the strategy are wrong, though, and soon the strategy could pay off.
Let's look at a hypothetical situation ...
OK, let's pretend it's Ron Rivera's Panthers who are down by 14 points at the start of the fourth quarter. They score a touchdown to cut the deficit to 8 and have two options: kick the extra point to make it a 7-point game, or try for a 2-point conversion to cut it to 6. Considering league averages, let's make a few assumptions about what happens next:
The Panthers are going to score another touchdown before the end of the game, and that touchdown will be the only remaining score in regulation.
The Panthers expect to successfully kick PATs 94% of the time and convert 2-point conversions 48% of the time - which are the league average rates since 2015.
The Panthers expect to win half of the time in overtime.
So let's say the Panthers follow our advice and attempt a 2-point conversion on the first touchdown. Here are the four ways the game can go, along with the percentage chance of each occurring:
Convert on the 2-point attempt, successfully kick an extra point after the second touchdown and win: 45%
Convert on the 2-point attempt but miss the extra point and tie in regulation: 3%
Fail on the first 2-point attempt but succeed on the second and tie in regulation: 25%
Fail on both 2-point attempts and lose: 27%
That leaves Carolina with a 45% chance to win, a 28% chance to tie in regulation and a 27% chance to lose the game. Since teams win half their overtime games, we'll add 14% to both the winning and losing sides. The result? The Panthers have a 59% chance to win and a 41% chance to lose, assuming they score again and stop their opponent from scoring.
So ... is that better than the alternative -- kicking both extra points and going to overtime?
Yes, if the Panthers actually make it to overtime at all, where they'd win half the time. Let's play out what happens if they ignore our advice.
Successfully make both extra points and tie in regulation: 88%
Miss the first extra point, convert the 2-point conversion on the second TD and tie in regulation: 3%
Miss the first extra point, fail on the 2-point conversion for the second TD and lose: 3%
Make the first extra point but miss the second and lose: 6%
That's a 91% chance to tie in regulation and a 9% chance to lose. Or, after accounting for overtime, 45.5% to win, 54.5% chance to lose. What if they want to kick the extra point on the first touchdown but try a 2-point conversion on the second?
This is also not a good idea. Here are the numbers:
Successfully make the extra point, convert the 2-point conversion on the second TD and win: 45%
Successfully make the extra point, fail the 2-point conversion on the second TD and lose: 49%
Miss the extra point, convert the 2-point conversion on the second TD and tie in regulation: 3%
Miss the extra point, then fail on the 2-point conversion on the second TD and lose: 3%
That's 45% to win, 52% to lose and 3% to tie in regulation -- or 46.5% to win and 53.5% to lose after overtime.
Shouldn't we consider teams' strengths, weaknesses and relation to league averages?
That's usually a good practice when making decisions, but, in this instance, it will not swing the result.
Let's say the team expects it can convert 2-point conversions at an astonishingly low 40% -- well below league average and, perhaps, only what a team might expect in the worst of circumstances ...
Convert on the 2-point attempt, successfully kick the extra point after the second TD and win: 38%
Convert on the 2-point attempt but miss the extra point on the second TD and tie: 2%
Fail on the first 2-point attempt but succeed on the second and tie: 24%
Fail on both 2-point attempts and lose: 36%
Even in this scenario, going for two leads to a higher win rate than kicking the PAT. That's not even considering that, if the team is this bad at 2-point conversions, it probably doesn't have a full 50% shot at winning in overtime, either.
How soon is too soon for this strategy?
FiveThirtyEight has a 2-point chart (the section titled "When you should go for 2") which indicates the viability of going for 2 at every score, at every stage of the game and depending on the expected conversion rate.
The chart indicates that going for 2 when down 8 is actually the better decision for the league-average team for most of the game. In fact, starting around midway through the third quarter, it's indefensible to kick the extra point. The chart also indicates the strategy is backed by ESPN's win probability model.
These two conclusions are important because so far we have been working under the assumption that the only other score in the game will be another touchdown by the losing team. The win probability model verifies that this is the correct decision despite the fact that it is working under no such guarantee.
Which teams have tried this recently?
The Eagles and Giants gave it a shot in 2018, and the Ravens, Panthers and Lions have tried it this season. Two of the five (Ravens and Eagles) were successful in their 2-point conversion attempts, but none was able to come back and win the game. The Panthers were the only one of the four that prevented the opponent from scoring more points.
Worth noting: The Eagles and Ravens are widely considered two of the most analytically inclined teams, and the Panthers poached their director of football analytics from the Eagles this past offseason.