A few months ago I noted Pittsburgh’s "peculiar strategy" in its week 12 loss against the Seahawks, when the Steelers’ offense threw 59 passes compared with just eight runs by a running back the whole game. In that article, I put it as: "I would say the Steelers abandoned the run, but they never even made eye contact with it."
According to Chase Stuart, it was the fifth-highest pass-run ratio (81.3 percent) called in a game in 2015. And that’s adjusted for sacks but without filtering for non-called runs so Stuart includes all 14 of Pittsburgh’s official rushes in the ratio, even the halftime kneel. Of the four teams that used a higher proportion of passes in any games last year, three of them had radically negative game scripts (more than minus-double digits.) The other was the Patriots, which I'll get to shortly.
Game scripts offer a way of looking at a game’s texture, rather than just its result, and they’re good for showing how score differential during the game dictates strategy, urgency, and other dynamics of play. Technically, a game script is simply the average lead or deficit a team faces across every second of game time. Because NFL teams throw more often and throw farther downfield when they’re behind, and run to hasten the clock when they’re ahead, high passing totals often come in losing efforts while high pass frequency often correlates with negative game scripts—regardless of quarterback performance or ability.
Indeed, teams that posted a pass-run ratio higher than 70 percent in games in 2015 lost 80 out of 85 times. (New England was the only team to win when throwing more than 80 percent of the time, having done so twice. Ugh.) It’s not because passing is a bad thing though, it’s because most of those teams were already way behind when they started upping the pass ratio.
Pittsburgh lost by nine that afternoon, but they never trailed by more than a touchdown until the final two minutes and actually led the majority of the game (Seattle led for only 16 minutes and change.) The Steelers’ game script was positive because they were leading or within a field goal for 56 of the first 58 minutes. Less than 4 percent of teams with positive game scripts in 2015 dropped back to pass on at least 70 percent of their plays and Pittsburgh’s ratio was way higher than 70 percent.
That’s one reason I call the Steelers approach peculiar. They didn’t throw a high volume of passes out of any sense of urgency or in reaction to the game getting away from them. It was the plan.
You could tell it was the plan right away.
Pittsburgh received the opening kickoff and Ben Roethlisberger threw on the first three plays. After the sides traded punts, Roethlisberger attempted four more passes in a row before DeAngelo Williams got his first carry. Williams added two more rushes on what would be a 58-yard field goal drive, still mostly featuring the pass. Seattle punted and the Steelers went right back to Roethlisberger’s arm on a nine-play, nine-pass drive that went 61 yards but resulted in a fake field goal interception thrown by Landry Jones to Jeremy Lane. In all, Roethlisberger attempted 20 throws in the first quarter alone.
In the second half, even with Pittsburgh aiming to preserve a lead, Williams had two carries. No other running back touched the ball.
The other reason I call this plan peculiar is because the Seahawks have the Legion of Boom. I mean, Pete Carroll is kind of known for having a good pass defense; Seattle was coming off two straight years leading the NFL in fewest pass yards allowed. In 2014, opponents dared fewer passes against the defending champions than any other defense in the league. Yes there had been some defections and some struggles in 2015, but the scheme was the scheme and most of the personnel were the same. It just seems odd to attack one of the league’s premier units at its strongest link.
The obvious counterpoint is that the Steelers simply felt confident in their own personnel. As I mentioned in the other piece, Roethlisberger and his receivers were at the height of their powers at the time, smashing 350-yard craters in opposing pass defenses left and right like some yellow-and-black clad conglomerate of the monsters from Rampage.
And they were without their All-Pro tailback LeVeon Bell who tore his MCL earlier in the month, so why not go with the pass-heavy approach?
The reason this doesn’t quite add up is that Pittsburgh didn’t try any such imbalanced strategy against the other defenses it played, even with Roethlisberger approaching Super Saiyan status and with Bell sidelined. The Steelers threw 58 passes against the Broncos in week 15, but they also fell behind 27-10 in the second quarter and their game script was negative-5. It’s not like Williams, a seven-year starter with Carolina, is DuJuan Harris. Apart from the Seattle matchup, he averaged 21 carries per game in his six other starts from week 9 until hurting his ankle before week 17.
So what’s the deal? Why did Mike Tomlin and Todd Haley choose to play 500 against the Seahawks?
It might have something to do with Seattle’s own peculiar strategy. However much notoriety the pass defense gets, Carroll makes no secret that his priority is stopping the run. Part of how the defensive backs work so well for Seattle is their coverage allows the rest of the alignment to swallow up rushing attempts. Remember when I mentioned that the Seahawks were tops in pass yards allowed in 2014? Well they were 3rd against the run that year and, after adding Ahtyba Rubin, arguably got better in 2015. (I say arguably because even though Seattle officially finished first in yards/game, its 3.6 yards per rush allowed was marginally worse than ’14.)
Either way, like the Big Bad Wolf dressed as granny, the Seahawks play to invite pass attempts. The better to eat you with... but maybe the Steelers figured they had a sharp enough axe. If they were going to be made one-dimensional anyway, why waste valuable plays in the wrong dimension? The Seahawks are designed to bend not break, but perhaps Pittsburgh calculated that everything breaks if you bend it enough times.
It’s a curious idea but was it a good idea? Well, no and yes.
Pittsburgh didn’t win, but they threw for 480 yards. Last time I wrote about this I concluded that contrary to appearances, the Seahawks broke the Steelers offense more times than they themselves got broken: Four interceptions compared with only one touchdown pass allowed, for an adjusted net per-play average of 5.2 yards spread across the 59 attempts (plus sacks.) Still, it wouldn’t have been enough had Russell Wilson not decided to play the Woodsman for his own offense, and cut the Steelers belly open for five crucial touchdowns.
However, I’ve been talking as if Pittsburgh’s passing volume was truly a unique outlier among performances facing Seattle. Could it possibly be a record numbers of dropbacks against the Seahawks?
Nope. According to Pro Football Reference’s database, both occasions when teams tried more passes against Seattle were memorable for different reasons. The most recent one was the first playoff game after the 2010 season in which Drew Brees threw 60 times. The other was the Vinny Testaverde helmet game, when the Jets QB chucked it 63 times.
But has anyone tried it more recently than that?
It turns out in 90 games from 2011 to 2015, including playoffs, opponents lofted as many as 50 throws against Seattle only four times. That’s not a lot but it’s also about the same rate as teams that threw 50-attempt games against the rest of the league, and you get equivalent results when you widen the filter to 45 passes or even just league-average 36 passes. If you try to factor for pass-run ratio, Seahawks opponents in that span actually dropped back more than twice as often as they ran more frequently -- 24.4 percent -- than the NFL average -- 19 percent.
Warning: This is where I start talking about the Patriots again. The one that comes closest is probably New England in 2012. The Patriots lost that day and Tom Brady tossed it 58 times, but Bill Belichick also called 26 rushes. However, Brady also tossed it 50 times in Super Bowl XLIX.
It was the start of a weird but simple trend. From that Super Bowl until the Seahawks snapped this streak against the Steelers, the four teams that threw the most passes against Seattle all won. In fact, from the standpoint of Tomlin and his staff before week 12, the Seahawks were 0-4 from February to November when the opposing offense threw anything more than a league average number of attempts. Even the Lions nearly won on Monday Night Football with a total (35 passes) just decimals below the cutoff. Maybe Tomlin had more than a hunch.
Four games is not a huge sample, obviously, but if you consider Seattle’s record since 2011 when the other team had to throw at least 36 times to be its "true" likelihood of winning such games (0.735 in 34 instances including those losses), then the chances of losing those four in a row by sheer coincidence are less than half of one percent. More factors contribute to football wins and losses, for sure -- and not all opponents are equal -- but the probability suggests that a real decline had altered the Seahawks odds.
Indeed, as John Clayton noted last week, the LOB allowed opponents a QBR of 55.3 in 2015, compared with 44.2 in 2014 and 28.3 in 2013. What happened? What changed? Did the Steelers demonstrate a blueprint -- already unlocked by the Patriots -- for thwarting the Seahawks?
Run defense is the boulder at the heart of Carroll’s gameplan. Is the best move just to sidestep the boulder and throw over it? Is the optimal strategy to just keep firing and stretch the probabilities as far as they will go?
Not quite. Seattle won the Pittsburgh matchup and the remaining two times it faced a higher-than-usual number of pass attempts and the combined score for those final two games was 71-12.
Likewise, though they all shared the coarse thread of high passing attempts, those four losses each had different scripts. Cincinnati (24-7) and Carolina (20-7) both dug out from harrowing deficits. Neither the Bengals (31 carries) nor the Panthers (33) rushed particularly seldom, nor did Arizona (33), as it led by an average of 5.5 points throughout its win in Seattle. Carolina barely passed more than it ran (36-33).
Sure enough, the Super Bowl remains the only example, dating to midway 2012 (17 games), of a Seahawks loss when their opponent held a pass-run ratio greater than 2-1. And I recall that one could have gone either way.
If the select series of games that includes Super Bowl XLIX but ends right before week 12 last year has any concrete overlap with a real-world variable, it’s that this was the period Jeremy Lane missed time for his injury and rehabilitation. And quite neatly so: Lane’s sabbatical was literally bridged by his interceptions against the Patriots and Steelers. That’s right, the nickel corner Seattle just signed for four more years, with two years guaranteed, is 19-3 as a Seahawk in games that fit this high-volume criteria. Those three losses being that Super Bowl (when he didn’t play the second half) and the first two losses of Lane’s rookie year.
Numbers are funny things. Did you know the two games in which Wilson dropped back the most last season by far were the Rams games in weeks 1 and 16? Not really the matchup I would have pursued against that front, but taking lots of sacks can separate plans from tactics in a hurry. A big raw number or even an out of whack ratio rarely tells that story.
I’ll keep my eye on teams trying a pass a lot against Seattle in 2016, but unless I see more evidence to the contrary I’ll be comfortable trusting it’s no more decisive a factor than whether the fourth best defensive back is on the field.