Welcome to Fourth and Short: Thank You for Not Coaching, Part 2
Although you would be forgiven for forgetting, there were actually brief moments of football that snuck between the blown calls and “bullshit” chants that seemed to dominate this past weekend of NFL action. Some of that football inspired very interesting coaching decisions, and while the best and worst of those decisions normally come up in the “Thank You for Not Coaching” section in the Monday column, there were so many decisions worth discussing this week that it’s going to serve as the basis for this edition of the newly named Tuesday column, “Fourth and Short.” (Thanks to reader Josh Dixon for giving us a better name than “The NFL Thing We Used to Call ‘Fabs and Flops.'”)
The important thing to remember with coaching decisions this week, as it is every week, is simple: Process, not outcome. There were some notable choices made this week that were clearly suboptimal and still led to victory, and there were others that were wise that eventually resulted in defeat. One correct decision going the wrong way isn’t a sign of failure, and if a coach continues to make smart moves when he has the opportunity to affect a game, it’ll produce positive outcomes more often than not. And speaking of scenarios where poor processes produced victories, let’s start by talking about the Jets!
In the interest of our shared writer-reader sanity, let’s skip past the first four hours of this game and start in the fourth quarter, where the Jets faced a difficult decision. Down 17-10 and with the ball on the Miami 1-yard line on first down, the Jets failed to punch the ball in. They were stuffed on a run play, threw an incomplete pass, took a penalty, and then scrambled the ball back to the 2-yard line. With 13 minutes left in a tight game, the Jets chose to kick a field goal and make the score 17-13.
Think about that one for a second. What sort of argument can you make for the field goal being right? Do you think that the Jets were so dismal near the goal line that they had no chance of scoring a touchdown and needed to settle for three? Well, considering that they were still going to need a touchdown (or an unlikely two additional field goals) to win, that doesn’t offer much consolation. Brian Burke’s fourth-down calculator suggests that the Jets should have gone for it seeing as they had a 31 percent chance of converting the fourth-down play. The league average for success on that fourth-and-2 play since 2000 is 44.2 percent. And while the Dolphins do have an excellent run defense, there’s some added value in giving the ball to Ryan Tannehill inside his own two-yard line and forcing him to do something. The Jets should have gone for it.
The Dolphins ended up losing the tilt because of several field goal fiascoes. You have to accept that some field goals just aren’t going to go through, but the Dolphins made two tactical mistakes that fall on head coach Joe Philbin and offensive coordinator Mike Sherman. One was a typical blunder that popped up in overtime. After the Dolphins picked up 41 yards on a pass to Brian Hartline, they picked up a first-and-10 on the Jets’ 35-yard line. That’s a 52-yard field goal, which is at the edge of Dan Carpenter’s range, but still an uncomfortable attempt that only goes through about 55 percent of the time. The Dolphins really needed one more first down to get in Carpenter’s range, but the Fins didn’t give much of an effort to pick it up, including a two-yard run from Daniel Thomas on third-and-7. That left Carpenter with a 48-yarder to try for the win; easier than 52, but not by much, as 48-yarders will go through the uprights about 65 percent of the time. As it turned out, Carpenter missed.
Then, Philbin managed to look like a fool on the final series of the game. When the Jets lined up for a game-winning 33-yarder, Philbin iced the kick with a timeout moments before the Dolphins burst through the line and blocked it. Loyal readers will remember that the topic of icing has come up before in this space: Icing the kicker does not work, according to a study from the book Scorecasting.
Of course, coaches don’t ice the kicker because they have empirical evidence that it works. Coaches ice the kicker because it’s a play that only has upside for them. If a coach ices the kicker and it works, he’s hailed as a game-saving genius. If it doesn’t work, nobody notices. The only time he looks bad is when a situation like the one Philbin created pops up, but even that’s not enough for coaches to stop icing kickers in late-game situations.
One of the simplest decisions of the day was Minnesota’s successful decision to go for it on fourth-and-goal from the 1-yard line at the end of their opening drive. When there’s still 53 minutes left to go in a football game, your only goal should be to accrue as big of a gap between your score and the opposition’s score as possible. You don’t give your team any comfort by kicking a field goal and getting a zero off the board, because you’re not winning games 3-0.
If we assume the Vikings convert on their field goal try 100 percent of the time, the choice to kick is worth three points. A touchdown and extra point are worth seven points, so the math would suggest that the Vikings should go for it if they think they can convert for that one yard 43 percent of the time. Since 2000, teams have converted that fourth-and-1 51.5 percent of the time.
The threshold is even lower because of what happens when you don’t make it. Just as the Jets could have trapped Tannehill on his 2-yard line, the Vikings would have started Alex Smith on his 1-yard line had they been stopped for no gain or thrown an incomplete pass. There’s value in that because it creates the distinct possibility that your next drive will have excellent field position, along with a decent shot of forcing a turnover that would produce a touchdown. Obviously, the Niners are less likely to provide a takeaway and have a hellacious goal-line run defense, but it’s hardly enough to overcome the huge statistical advantage provided by going for it. The Vikings were right to do so and were rewarded for their decision.
With six minutes left in the game and a 10-point deficit on the books, the Browns faced a fourth-and-9 from their 36-yard line. It was admittedly a dire situation. With three timeouts left, the Browns chose … to punt? Pat Shurmur’s decision was baffling because it seemed to suggest that the Browns were giving up on the game. How could the Browns realistically win by punting?
Fourth-and-9 is tough, but by punting, Shurmur was suggesting that his team was going to get a three-down stop while using all of his timeouts (which he did), receive a punt, and drive down the field and score. Then, the Browns were going to force another three-and-out (or successfully employ an onside kick), receive a punt, and drive downfield again to kick a field goal.
Meanwhile, if the Browns just bit the bullet and tried to go for it, their path to winning would have been a lot easier. A conversion moves the ball to at least the Cleveland 45-yard line, putting them within about 25 yards of Phil Dawson’s range. Then, if the Browns were stuffed and kicked a field goal, they could choose to kick the ball deep, use their timeouts on one stop of the Bills, and then drive down the field for a possible game-tying touchdown. You tell me which sounds more plausible.
The Buccaneers had an opportunity to employ one of football’s most bizarre statistical gambits in the fourth quarter of their loss to the Cowboys on Sunday. They might have even been right to call it into action.
The scene wasn’t pretty. Tampa Bay was down 13-7 with five minutes to go while facing a fourth-and-18 from their own 1-yard line. The only bright sides were that the Bucs had played solid defense all day and had all three timeouts in their pocket. Greg Schiano chose to punt the ball away, but Dez Bryant returned the ball to the 6-yard line and the Cowboys eventually kicked a field goal that made it a two-score game, basically ending it as a contest with 2:47 to go.
Was punting Tampa Bay’s only option in that scenario? Absolutely not. There was one other strategic choice that the Buccaneers could have opted for: the intentional safety.
By punting, Tampa created a situation that was going to be downright impossible to overcome. Even a good punt by Michael Koenen from deep in his own end zone was only likely to get to midfield, and if we assume that Bryant will produce an average return, the Cowboys would be taking over somewhere around the Tampa Bay 40-yard line. That puts them one first down away from field goal range, and a field goal makes it a two-possession contest. With such little time left, Tampa’s goal needs to be keeping the game within one score while preserving as much time as possible.
That’s where the intentional safety comes in. The two points give Dallas an eight-point lead, leaving the Bucs with a chance to tie the game on one drive. The ensuing free kick likely gives Dallas inferior field position, forcing them to pick up multiple first downs to get into field goal range and end the game. A stop would also give the Buccaneers better field position than they would have had through punting, too.
It’s a similar decision to the one Bill Belichick famously made against the Broncos in 2003. That was for a one-point lead as opposed to a six-point one, but it was also with half the time to go. If anything, a team needing a touchdown craves the field position provided by the free kick even more than one needing a mere field goal. The call’s debatable because it would have prevented the Buccaneers from likely winning in regulation, but I think an intentional safety would have improved their chances of getting to overtime and winning. At the very least, it’s a conversation worth having.
Oh, and Jason Garrett’s surprise onside kick? Don’t knock it based on the fact that Tampa Bay recovered. The numbers say that surprise onside kicks work about 60 percent of the time, making it a good idea to throw into the playbook. Of course, you can’t just throw it out there every single week and expect it to work; you need an opponent who truly isn’t expecting it, one whose players are turning away from the ball before it’s kicked to try and get downfield and set up blocks.
Garrett presumably called for it because he saw something on film suggesting that Tampa was turning away too early. On the kick itself, it didn’t look like the Buccaneers were turning around before the kick, because it was an easy recovery. That’s the reason why you should be suggesting it was a bad kick, not merely because Garrett tried it and the Cowboys failed to recover.
Finally, let’s finish up with the decisions made by each of these teams in their epic overtime encounter. The Titans made a likely wrong call and won; the Lions (accidentally) made the right call and lost.
With Tennessee facing fourth-and-2 from Detroit’s 8-yard line on the opening drive of overtime, they were put to a difficult choice. If the Titans would have gone it, succeeded, and scored a touchdown, they would have ended the game without handing the ball back to the Lions offense. By kicking a field goal, they took the lead, but returned the ball to the Lions, who had marched down the field on them for most of the fourth quarter. There’s not enough of a sample with the new overtime rules to get a good idea of the specific probabilities for this scenario, but it’s hard to imagine that the Titans wouldn’t have benefited from holding onto the football and scoring a touchdown without giving the Lions an opportunity to touch it.
The clearer example came on the opposite side of the field, when the Lions faced a fourth-and-1 on the Tennessee 7-yard line. Although Detroit apparently “accidentally” snapped the ball while trying to draw the Titans offside, they were probably right to go for it anyway. This is a more traditional overtime situation, so it’s somewhat easier to break down.
The math is slightly different depending on which Expected Win Probability model you use, but either way, the Lions were in the same boat. By kicking a field goal and tying up the game, they were giving the ball back to the Titans with under seven minutes to go in sudden death. In other words, there was a distinct possibility that they would end up losing to a field goal without ever touching the ball again themselves. ESPN Stats and Information estimates that Detroit’s chances of winning after a successful field goal were just 38 percent.
If the Lions did choose to go for it, their odds of success were a lot higher. ESPN Stats and Info also notes that teams who went for it on fourth-and-1 between the 4- and 10-yard lines finished their drives with touchdowns 54 percent of the time and field goals an additional 8 percent of the time. With an exhausted defense facing a running game that had been effective throughout that overtime drive, those figures might be conservative. Regardless, this doesn’t require a math degree: 54 percent is significantly higher than 38 percent. The Lions were better off going for it, even if they didn’t mean to.