NFL broadcasts and coaching news conferences are full of football proverbs. Very often, these are simply explanations for a why a situation demanded avoiding risk, or at least delaying it. And very often, the numbers -- while not perfect -- tell us otherwise.
In many cases, these unwritten laws of coaching and game management in football are more a function of history and tradition without a space for self-reevaluation or change. That's not acceptable. Teams who spend all week looking for the tiniest little competitive advantages abandon them when given the opportunity to impact a game. That's a waste, and it's time for a change.
Let's run through some of these close scenarios and explain why the traditional method of thinking about them is antiquated. Starting with a classic ...
Never take points off the board.
Points are valuable! Of course we want points! And when you have to battle for 60 yards, then turn to a questionable kicker who narrowly sneaks one through the uprights, the last thing you want to do as a coach is repeat this process while running the risk of coming away with no points. The possibility of scoring seven points, though, should make three seem much less valuable.
There are obvious situations in which teams should keep their points -- to tie or take the lead in a close game, for one -- but early in a game, when the only goal should be to score as many points as possible, coaches need to be open to the idea of leaving their offense on the field to score a touchdown. Since kickers are better than ever before and turnover rates are at their lowest since 1932, the chances are slim that an offense will take three off the board and end up with zero.
Let's use the expected points model that underpins ESPN's QBR metrics. Here's a simple scenario: It's early in the second quarter of a 7-7 game with league-average offenses and defenses. Your kicker hits a field goal on fourth-and-2, but the defense is offside, giving you a first down if so inclined. Here's how many points your team would expect to score with a new set of downs from each given yard line:
These numbers account for the risk of turning the ball over or not scoring at all versus the reward of scoring a touchdown. In a vacuum, when you expect to score more than three points from a given yard line, you should wipe the field goal off the board and go back out on offense. Just about every feasible situation in which you would be kicking a field goal seems to suggest that taking the points off of the board is the superior option. Even if you are conservative and have an subpar field goal kicker, a new set of downs would most likely get you inside the 30-yard line. Send your quarterback back out there, coach.
Never throw when you're running a four-minute offense to kill clock.
Thirty years ago, when quarterbacks were throwing farther downfield and completing a far lower percentage of their passes, it made sense for teams to strictly limit their quarterbacks to handing off the football in situations where running clock was more important than picking up a first down. Quarterbacks simply couldn't be trusted not to screw up, and most receivers weren't good enough to regularly win one-on-one matchups.
Today's game is different. Quarterbacks routinely throw bubble screens and other short passes designed to get the ball out quickly, and their success rate on those throws is remarkable: They complete 70.9 percent of their passes within five yards of the line of scrimmage and throw interceptions 1.1 percent of the time. That's similar to the fumble rate on rushing plays since 2012 (1.2 percent, although only 0.7 percent are lost to the opposition).
Repeat: Many passes are high-percentage plays.
In situations where teams are one or two first downs away from ending the game, passing simply has to be part of the equation, if only to prevent teams from teeing off on your running game. The screen Dak Prescott set up with Cole Beasley to seal the game against the 49ers is a perfect example. And, with run-pass options, you can ask your quarterback to make a simple check at the line of scrimmage and either hand the ball off or make a pass that is likely to be completed.
In situations where teams simply want to run as much clock as possible and punt without any real concern about getting a first down, they're probably better off kneeling than running offensive plays, if only because of the risk of a fumble. There's little logic behind teams "half-trying" to succeed on offense. Another example of that ...
Run to start your two-minute drill, because if it fails, just run out the clock.
Many NFL teams are fond of starting their final drives before halftime with a draw or another sort of running play to try to test the waters. If the play goes well, they'll kick it into second gear and start sprinting down the field to try to score. If the play fails, they'll slow down and waste time before hitting the locker room for halftime. In a league where teams constantly talk about dictating the game and imposing their will upon the opposition, the halftime draw is weirdly passive.
But the halftime draw doesn't suit either master. If the play works and gains 10 yards, you're now sprinting up to the line of scrimmage to run your next play while wasting precious seconds or you're burning a timeout. Given how far these plays likely are from the end zone, they're the most likely passes on your drive to be completed and the mostly likely to end with a free pass out of bounds to stop the clock. Teams are afraid of throwing incomplete passes and being stuck punting to the opposition, but if that's such a concern, you're better off kneeling and avoiding the risk of fumbles.
The classic example of end-of-half clock mismanagement came in Week 3 from the Titans, who ran a draw on first down for 8 yards from the 25-yard line with 33 seconds left in the first half. The opposing Raiders were down to one timeout. Oakland didn't call a timeout after the play, so Tennessee could have let the game go to halftime or called one of their own remaining timeouts to try to set up the next couple of plays. Instead, they rushed to the line and threw a pass with 11 seconds left, a meaningless 3-yard in-route that was telegraphed and nearly intercepted. On the next play, with eight seconds left, Marcus Mariota threw another pass over the middle that was tipped and intercepted by Reggie Nelson, who stepped out of bounds during his return with no time left on the clock. It was the polar opposite of how to manage a late-half or late-game situation.
Teams can get in trouble throwing the ball in these spots, although it's often with low-reward decisions; think about the Cowboys throwing a checkdown with time running out in the half against Washington in 2010 and having Tashard Choice's fumble returned for a touchdown -- which ended up as the margin of victory in a 13-7 loss. And there are times when the draw works, too. But that's not the point. It's better to have a plan and go all-out in attack or time expenditure without letting the opposition decide what to do on your behalf.
When teams are dealing with this decision in the fourth quarter of a tie contest, chances are it's better to be aggressive and try to win games. There are too many teams throughout history who have sat on the ball after allowing a late score and regretted the tale. The flip side of that, quite famously in opposition to John Madden's commentary, was the 2001 Patriots. They allowed a back-breaking touchdown to Ricky Proehl with 1:36 left in the Super Bowl to tie the score at 17, and while Madden suggested the Patriots kneel on the ball, Bill Belichick rightly realized he was a massive underdog and would only be running the risk of giving Kurt Warner the ball in overtime.
Tom Brady promptly drove the Patriots 53 yards for a game-winning Adam Vinatieri field goal.
Never go for two before you have to.
The rule differs around the league, but there are a fair number of NFL playcallers who don't go for two until the end of the game is in sight. For some, you start at the beginning of the fourth quarter, while others might not even think about their two-point plays until there are seven minutes or less remaining in the contest.
The argument says you shouldn't chase the score until there's a good chance it might be the final score, which makes some sense, even if many of those same teams and commentators ignore that logic in more conservative situations. (Many of them will argue how teams should kick a field goal to tie the score or make it a one-possession game at similar times of the contest.) It's true that teams shouldn't treat the numbers on the scoreboard as if they're guaranteed to be the final score, but it's also naive to suggest that scoring is entirely random from that point forward.
It's impossible for humans to imagine all of the possibilities in their head in real time, but it's far from impossible to imagine them with a computer, which is why coaches should look to models. The one created by Football Commentary is outdated because it doesn't account for the new extra point rules, but it gives us some broader insight into how teams should think about these sorts of decisions.
Their model suggests going for two isn't as meaningful or clear-cut early in the second half as it might be later in the contest, but there are still situations where the choice to go for two should be quite obvious. Teams should basically always be going for two when they are up by five or down by five. The same is true when they are down by two, eight, nine, 13 or 15, which leads to another unwritten rule ...
Kick the extra point when down nine to make it a one-score game.
The Browns ran into the furor surrounding this unwritten rule a couple of weeks back, when Hue Jackson decided to go for two down 28-19 with 2:10 left to try to make it a one-score game. He failed. This essentially ended the contest: The Browns recovered an onside kick and scored a touchdown, but that was only enough to make it 28-26. They failed on a second onside kick try and subsequently lost.
Let's flip that scenario and say the Browns kick the extra point first to go down 28-20 with 2:10 left. Then, let's say that the same stuff happens. The Browns recover the onside kick, score another touchdown, and then go for two and fail. They're in the same situation. They're down 28-26, but now, it's with 30 seconds to go. There is no time left to overcome their failed two-point play.
The outcomes are exactly the same. If you get the two-pointer, you need only one score to tie, regardless of whether you get it early or late. And if you don't get the two-pointer, you need two scores to tie, regardless of whether you miss with the first attempt or the second one. The only difference between the two plays, as Chase Stuart first pointed out to me, is that teams who go for two and fail on their first drive have more time to adapt their decision-making for the fact that they've failed by getting more aggressive with blitzes or offensive playcalling. The team who goes for it late and fails has spent the preceding few minutes assuming they were going to get the two-pointer to tie. (A team who gets the two-pointer on their first drive also can win by going for a two-pointer again on their second touchdown drive, but that is an avenue of aggressiveness coaches are unlikely to pursue.)
Coaches don't go for two on the earlier drive for a couple of reasons. 1) They rarely want to do something that will influence the game or take it out of their players' hands, and 2) the decision-making is based on putting off losing for as long as possible as opposed to attempting to actively win, which is why they pass up fourth-and-short early in the fourth quarter and are often stuck going for fourth-and-12 later.
I've also heard the argument that players will be demoralized by failing to go for two early and then get shut down, but I'm more skeptical there. The Browns had every reason to turn off the switch after failing to score their two-pointer, given the general hopelessness of the short-term situation in Cleveland, and they promptly played their butts off to get the ball back and score again.
The Browns' situation wasn't the classic example, if only because they didn't really have a ton of time to optimize their decision-making either way. It makes going for two on the earlier drive less meaningful, but there's still no added argument toward going for two on the later drive. You're fooling yourself if you think getting the early extra point does anything more to make it a one-score game.
Kick the field goal in a low-scoring game because points are at a premium.
I've never been able to understand this one. When scoring is low, some points -- any points -- could be the difference, so take 'em.
Problem: Isn't the flip side of that argument even more compelling? If you're not expecting to move the ball and come away with a drive deep into opposing territory, shouldn't you take advantage of that rare opportunity to score a touchdown? If three points are valuable, seven seems significantly more useful, given that it would take the opposing team a touchdown drive of its own (or three field goals) to match what you've accomplished on one drive.
Here's the benefit that comes whenever you miss a fourth-and-short deep in opposing territory: field position. In a low-scoring game, field position is critical. Teams punt in their opposition's side of the field and play it safe on third down to try to ensure a field-position advantage. If you kick a field goal, you're giving the other team the ball back after a missed field goal or an ensuing kickoff. If you go for it inside the 5-yard line and fail, you leave your defense in an advantageous position. That sort of field position advantage will add up, both over that drive and in the one to come.
There are more maxims to delve into while talking about football coaches and commentators. I still don't understand why teams on the road have to take the points, or why NFL teams try to ice kickers when the evidence suggests it has no meaningful strategic value (and can often allow a practice kick). But we'll keep hearing these, because sometimes, superstition and tradition rule the way over reality and logic.