Putting rookie quarterbacks in challenging situations is part of what the NFL preseason is all about. There are no real consequences for failure, and the level of competition is high enough such that the experience is both revealing and an opportunity for growth. Chicago Bears QB Justin Fields learned that in his first preseason game last Saturday against the Miami Dolphins, when his team was pinned at its own 2 on the final drive of the third quarter. While the offense produced a quick three-and-out (in what was otherwise mostly a good debut for the rookie), it gave Fields a taste of what it’s like to run plays out of the shadow of his own goal against a pro defense.
By the standards of rookie QB play in the preseason, maybe a three-and-out isn’t unexpected at all in such a situation. But it’s hard to know how we should weigh success or failure when a team is pinned near their own goal. While it feels like success should be rare when you’re stuck on your 2-yard line, is advancing the ball really that out of the ordinary? And what does success in these situations look like exactly?
First, let’s define success. Although it would be wonderful if a team could always march 98 yards down the field for a touchdown, the received wisdom from NFL coaches and analysts indicates that playing deep in your own territory is a significant challenge. Instead, a reasonable measure of success might simply be getting a single first down, giving your punter room to kick instead of punting with his heel an inch away from a safety. An extra 10 yards can also give your defense slightly better field position after the punt.
So that’s the metric we’ll use: the rate at which offenses convert to a new set of downs when they’re backed up near their own goal. To get a sense of what our baseline expectations should be, we took all plays since 2001 in the NFL regular season and looked at how often drives that started deep in a team’s own territory converted at least one new set of downs. The results were surprising:
When starting out inside their own 10, NFL teams are able to move the chains at least once the majority of the time. This is true even when they start on their own 1-yard line. It’s certainly harder to get a first down when you begin a drive so close to your own goal -- teams convert at 54 percent when starting from the 1, 8 percentage points below the conversion rate when starting at the 10 -- but it’s still shocking just how often defenses allow opposing offenses to squirm out from the clutches of horrible field position.
When you pin the other team at their 1, you might be forgiven for expecting your defense to respond with a three-and-out. Instead, more often than not, the offense marches forward for a first down. That’s tough to take, and seems like it would be soul-crushing for the punter and special teams unit. You’ve done the best job you can possibly do, and it still isn’t enough to tilt the odds of a three-and-out in your favor.
related: What If Chasing Rivals Keeps Aaron Rodgers In Green Bay? We Made A Metric For That. Read more. »
Which leads to the obvious question: Why is this happening? Why does the conventional wisdom overstate the offense’s disadvantage deep in its own territory?
What makes Patrick Mahomes so great? | FiveThirtyEight
One possible explanation might be that defenses get aggressive when teams are backed up. And perhaps they get too aggressive. If defenses blitz the quarterback hoping to force a turnover or safety, it could open up holes for an offense to exploit, leading to more first downs.
The evidence for this theory, however, is not completely clear. It turns out that teams don’t blitz any more often when their opponents are backed up than they do in the open field.1 The rate of blitzes per pass play2 since 2015 are almost identical in both parts of the field: 28 percent in the open field and 27 percent inside the 5.
It’s also true that expected points added (EPA) per play has been strongly negative when an offense is pinned inside its own 5 every year since 2015. So defenses are generally getting the better of these matchups in terms of how they contribute to net points.3
But this trend breaks down when we look at EPA per play against blitzes inside the five. It’s not just that blitzing is less effective -- the defense’s edge disappears entirely on those plays. The EPA per play on the 240 pass blitzes run by defenses inside the 5 since 2015 is a positive 0.16. This is a wild swing, especially compared to dropbacks with no blitz, where EPA per play remained negative (at -0.08).4
Perhaps there’s some game theory at work here. Backed up and on edge, offenses may expect the defense to be aggressive with the blitz. In this way, even though defenses blitz at the same rate as in other parts of the field, the offense is less surprised and perhaps better equipped to counter it.
This might also help explain Fields’s failure to convert a new set of downs in his first exposure to being pinned in the pros. On his third-quarter dropback inside his own 10, Miami didn’t try to blitz him. They played press coverage at the line and made Fields read the defense. He ultimately dumped it off to running back Artavis Pierce, who was quickly tackled shy of the sticks -- earning the Dolphins a three-and-out that’s actually rarer than we tend to think when a team is backed up near its own goal line. Once the games begin to count, perhaps other teams should follow suit against other quarterbacks facing the same situation.