This has has been a really weird year for NFL players, coaches, analysts and fans. Home-field advantage is basically nothing, games are being scheduled and (sometimes) played on Tuesday and the Dolphins are in the playoff hunt. One positive of 2020 is that we’ve seen a number of longstanding football regularities upset. One new head coach, with his fifth-career win in the balance during Week 12, went with convention and probably regrets it going into his first career bye week…
The Carolina Panthers, 3-point underdogs in Minnesota, kicked a field goal to go up 27-21 with 1:54 remaining and the Vikings out of time outs. This was not viewed as an egregious error by most in the NFL punditry —after all, the Panthers just went from up a field goal to up almost a touchdown!
However, some of us viewed it as a bad decision at the time:
Panthers should have gone for it tbh
— Eric Eager ???????? (@PFF_Eric) November 29, 2020
And after the result failed the Panthers:
This is why https://t.co/B0kKgzzWYB
— Eric Eager ???????? (@PFF_Eric) November 29, 2020
Firstly, the argument from the win probability perspective: The Vikings’ win probability trailing 6 points and starting from their 25-yard line was an estimated 18.6 percent due to their offense being very good and the Carolina Panthers‘ defense being poor.
The Vikings’ win probability down 3 points and starting from their 3-yard line (assuming they stop the Panthers on a fourth-down attempt) is estimated at 18.8 percent. While that might seem weird given its proximity to the previous number, historically teams down 3 do not play as aggressively as teams down more than 3 do (more on this in a second), and a field goal basically puts the game in the hands of a coin flip prior to overtime. Additionally, the extra 22 yards of field position with no timeouts are worth quite a bit.
Obviously, the Vikings' win probability is essentially zero with a Panthers conversion.
With such an obvious conclusion mathematically, why do we see this decision made?
It’s Not Actually Made That Often
Sunday’s game marked the first time a team made said decision in a game this year within two minutes, and only the Chargers (of course) have made this decision with four or fewer minutes left in the game, kicking a field goal with 2:34 left against the Broncos in Denver (they would go on to lose the game by one point on a Drew Lock touchdown pass as time expired).
In 2019, said decision was made five times to kick a field goal up 3 with two minutes or less to play:
New Orleans versus Houston, Week 1:
After Wil Lutz’s 47-yard field goal put New Orleans up 6 with 50 seconds left, Deshaun Watson hit Kenny Stills for a 37-yard touchdown to put the Texans ahead with just 37 seconds to play. The Saints would use those 37 seconds to put Lutz in position to kick another field goal, this time 58 yards, to win 30-28.
Detroit at Philadelphia, Week 3:
Four-point underdogs and up 27-24, the Lions attempted a 46-yard field goal to try to extend their lead to six, only to have it blocked by Malcolm Jenkins. Luckily for Detroit, the Eagles stalled out near midfield, and the Lions prevailed.
Carolina at Houston, Week 4:
Joey Slye, whose missed field goal kept the Panthers from rebounding from this poor decision against Minnesota on Sunday, put Kyle Allen and the Panthers up 16-10 with 28 seconds left in Houston. The Texans, with three completions in five plays, would get to Carolina’s 47 before the game’s end.
Pittsburgh versus Arizona, Week 14:
On fourth and goal from the Cardinals' 7-yard line — and with a backup quarterback in Ben Roethlisberger’s stead — the Steelers kicked a 25-yard field goal to put them up 23-17. Rookie Kyler Murray was intercepted by Joe Haden on the subsequent drive, sealing the Pittsburgh victory.
Oakland versus Jacksonville, Week 15:
In what would be a rough ending to the Raiders' second stint in Oakland, Jon Gruden’s bunch led the upstart Jacksonville Jaguars 16-13 with just 1:48 left in the game. Second-year kicker Daniel Carlson would miss a field goal on fourth and 6 from the Jacksonville 27-yard line. Rookie Gardner Minshew would then drive the length of the field, throwing a touchdown pass to former Jaguars receiver Chris Conley with just 34 seconds left. While the Raiders mounted a drive after, they would only reach the Jags’ 40 and lose a game that largely made the playoffs unreachable for them.
This list, while certainly a small sample, shows all of the things that can go wrong with this approach. The team that is ahead can miss the field goal altogether, mitigating the advantage of kicking versus going for it (and in one of the cases, the kick was blocked!). The team can give up a touchdown on the following drive, leaving them having to fight back with little time left to attempt another field goal, as New Orleans did in week 1 of 2019 (and what the Panthers almost did yesterday).
The team can also get a stop, of course, which seems like the default assumption of many NFL coaches (“if we can’t get a stop, we don’t deserve to win”). But in today’s NFL, even with the number of backup quarterbacks playing and the way in which defensive penalties are being called, this is a difficult bet to make, to say the least. There have been 16 games this year where a drive started with less than two minutes left and the offense trailing by 4, 5 or 6 points. The results:
Detroit versus Chicago, Week 1: D'Andre Swift dropped a game-winning touchdown pass, and the drive stalled in Chicago territory.
Carolina versus Las Vegas, Week 1: The Panthers had just 8 seconds left when they took the ball at their own 27-yard line and did not score.
New England at Seattle, Week 2: Cam Newton was tackled short of the goal line on fourth and short inside of Seattle’s 5-yard line.
Las Angeles Chargers versus Carolina, Week 3: The Chargers drive stalled in Carolina territory.
Atlanta versus Chicago, Week 4: Matt Ryan was intercepted in Chicago territory by Tashaun Gipson.
San Francisco versus Philadelphia, Week 4: The 49ers drive stalled in Philadelphia territory.
Seattle versus Minnesota, Week 5: Seattle converted multiple fourth-down plays to come from behind and beat the Vikings on a touchdown to D.K. Metcalf.
Detroit at Atlanta, Week 7: Matthew Stafford completed a touchdown pass to T.J. Hockenson as time expired to beat the Atlanta Falcons in Atlanta.
Green Bay versus Minnesota, Week 8: The Packers drive stalled in Minnesota territory.
Baltimore versus Pittsburgh, Week 9: The Ravens drive stalled in Pittsburgh territory.
Dallas versus Pittsburgh, Week 10: The Cowboys drive stalled in Pittsburgh territory.
Chicago versus Minnesota, Week 10: Nick Foles was injured on the first play and the Bears drive, led by Tyler Bray, stalled in their own territory.
Arizona versus Buffalo, Week 10: Kyler Murray completed the “Hail Murray” pass to DeAndre Hopkins to beat the Buffalo Bills, 32-30, in Arizona.
Las Vegas versus Kansas City, Week 11: Daniel Sorensen intercepted Derek Carr near midfield to clinch the game for the Chiefs.
Minnesota versus Carolina, Week 12: Kirk Cousins‘ touchdown pass to Chad Beebe, along with Slye’s missed field goal, are enough to beat the Panthers 28-27, in Minnesota.
By my count, four of these drives ended up in game-winning touchdowns (25%), while the majority of the others ended up on the defense’s side of the field, in a few cases needing high-leverage plays to fall the defense’s way to avoid yielding a devastating loss.
The Psychology of Being Down Three Versus Being Down by 6-ish
Why are these comebacks relatively frequent? The answer is probability “they have to be.”
Necessity breeds a bunch of things, but urgency is one of them.
There is a psychological difference in a team’s approach when down by 3 points or less versus being down by 4, 5 or 6, especially given the modern day kicker’s ability to make kicks from over 50 yards with ease.
And this gets to the biggest reason why kicking the field goal up three is not the right move: The downside to failing to convert the fourth-down play is far lower than we think it is, because of the misplaced relief the opponent feels when they realize that they can extend the game (i.e. make it “less than” a coin flip) with a field goal on the subsequent drive. The relent generated is not trivial.