The Mike Smith decision to go for the touchdown at the end of the first half last night has kind of been overshadowed by other news today — Julio Jones’ foot injury is a much bigger impact. I still wanted to re-visit the decision made by Mike Smith to go for a touchdown at the end of the first half. Atlanta was trailing 17-7, and had the ball on the goal line with time for one play. Well, two more plays, because on 4th down, the Jets mugged Tony Gonzalez, got called for the penalty, and Atlanta got another opportunity.
On the second attempt, Jacquizz Rodgers was stuffed, and the Falcons went to the half still down 10. The Jets got the ball to start the second half, which was certainly a factor in Smith’s decision. He had just seen his defense, starting two rookie linebackers due to injury, get marched on by the Jets offense.
So what do the past results tell us?
First, we can look at what actual coaches have done in a similar situation. Using the pro-football-reference play finder, there have been 23 situations since 1999 when a coach has had the ball at the opponent’s 1 yard line, with four seconds or less remaining. I would say that four seconds represents the point where, even with a timeout, the coach cannot go unless he wants to make the binary choice of touchdown or field goal.
Nine times, including last night, the team went for the touchdown. Seven were scored.
Fourteen times, the team kicked the field goal, and all were made.
That shows that this is far from a situation where coaching dogma dictates only one strategy. Trailing by 10 probably evened it up even more. The true chances of converting are not 7 of 9, though it shows that over the last decade, the coaches that have gone have gotten more points on average than the field goal decision.
To better estimate the chances of success, I looked at all 3rd and 4th and goals from the 1. Over the last five years, teams have converted a touchdown 54 percent of the time in that situation. Because seven is more than twice of three, the touchdown does yield the higher average return. (3.8 points to 3.0 points).
Forget points, though. What about looking at the chances of winning? Advanced NFL Stats has the win probability calculator for a variety of situations, so it is easy to look at the Jets starting the half with the various leads. We can thus see if cutting it to seven matters more than being at 3 vs. 10 down.
As it turns out, no. The largest jump in win probability comes from going from down 7 to down 3, rather than being within one score at down 7 instead of down 10. According to the calculator, the Falcons would have a 16% chance of winning down 10, 23% down 7, and 36% down 3.
The break even point would be a 35% chance of scoring a touchdown by going for it, in terms of comparing the effect on win probability. I think there are tangible reasons to believe Atlanta had less than a 54% chance at the goal line, but even accounting for that, Smith’s decision was certainly reasonable.
I do want to throw out another possibility. Kicking twice, to end the half, and then to start the half, means the Jets get the ball already ahead by 7. You could go for the touchdown to cut the lead, which Smith opted for. If you want to take a chance to get back in the game, you could also instead choose to start the half with a surprise onside kick. A couple of years ago, surprise onside kicks were successful 60% of the time.
Here are the chances of winning the game based on the three decisions:
- Kick Field Goal, Kickoff to Jets, down 7: 23%
- Go For It, Kickoff to Jets, down either 3 or 10: 26%
- Kick Field Goal, Surprise Onside Kick, down 7: 27%
And of course, going for it and then trying the onside kick finishes a bit higher still. But we are getting greedy here. The universe would collapse on itself if a coach both went for it on the goal line to end the half down 10, then came out with a surprise onside kick to start the second half.