During the fourth quarter of Sunday’s game in Philadelphia, the Eagles scored a touchdown to cut their deficit to 20-12. Everyone assumed they’d kick the extra point to make it 20-13.
But coach Doug Pederson did what no one expected: He went for two, and it worked, and the Eagles cut the deficit to 20-14.
When Pederson made the highly unusual decision, the TV announcers questioned it, and social media lit up with the question: What is Pederson doing?
What Pederson was doing was showing he has a far greater grasp of analytics and game theory than any other NFL coach. Because while every NFL coach has always kicked the extra point in that situation, the analytics community has steadfastly insisted that going for two is the better option.
Kevin Cole, the Director of Data & Analytics for RotoGrinders, spelled out the reasoning last year: If you’re trailing by 14 late in the game, and you score a touchdown to cut the deficit to eight points, when you kick the extra point to cut the deficit to seven you’re hoping that you’ll later get another touchdown and another extra point, which will send the game into overtime. In overtime you’ve basically got a 50-50 shot of winning. But if you go for two after that touchdown in an attempt to cut the deficit from eight to six, now you’ve changed the odds from 50-50 to more like 62.5 percent in your favor: 50 percent of the time you’ll make the two-point conversion, in which case you’ll win in regulation with a touchdown and one-point kick. But there’s also the 50 percent of the time in which you’ll miss the two-point conversion, and in those cases after you score another touchdown you’ll still have a 50-50 chance of making another two-point conversion, followed by a 50-50 chance of winning in overtime. And 50 percent (that you miss the first two-point conversion) of 50 percent (that you make the second two-point conversion) of 50 percent (that you win in overtime) is 12.5 percent, so add that 12.5 percent chance of winning to the 50 percent chance of winning from making the first two-point conversion and you have a 62.5 percent chance.
The Eagles ended up losing Sunday’s game, but that’s not the issue. The issue is whether Pederson made the correct strategic decision at the time, and he definitely did: He went from his team’s best-case scenario being another touchdown and a 50-50 chance of winning in overtime, to his team’s best-case scenario being another touchdown and a game-winning extra point in regulation. And even if the Eagles had failed on that two-point conversion, they still could have gone for two after their next touchdown to get back to that 50-50 chance of winning in overtime.
This might all sound like a justification that only sounds good after the fact, because the Eagles got that two-point conversion. But that’s not what this is. People who have thought through the scenarios have been saying for years that some coach ought to do what Pederson did. During the fourth quarter of the Steelers’ playoff loss to the Jaguars last season, I tweeted when the Steelers scored to bring the deficit to eight that they should have gone for two. I also added that “no NFL coach would even consider it.”
As it turned out, there was one NFL coach who would consider it, and it was the same coach who won the Super Bowl last season while embracing analytics. Pederson made a great call.
