A numbers perspective: Why the Steelers season wasn’t as bad as it looked

Despite lofty expectations, the 2018 version of the Pittsburgh Steelers took a step backward - by three and a half games to be exact. Following a 13-3 performance in 2017 the Steelers finished 9-6-1 and did not qualify for the postseason. There have been no shortage of narratives trying to break down why this team declined. Add together the continued impact of losing Ryan Shazier, the Le’Veon Bell drama, the perceived lack of locker room leadership, the disappearance of Artie Burns and Chris Boswell, the disruption that is Antonio Brown, and some atrocious calls by referees and you have yourselves a theory.

But let’s put aside the specific issues and attempt to put this season in a broader context. That’s likely what the Steelers’ brass, known for their long term approach, is doing anyway. How might the Steelers’ season by viewed from 10,000 feet with a handful of key numbers?

The 3.5 win gap that becomes 1

While storylines provide potential reasons that fans can grasp onto, the bottom line is that points decide games. No need for NASA here. Looking at the last 10 NFL seasons over 84 percent of the variance in a team’s regular season win total comes down to how many points they score, and how many points they allow. Here is that relationship in graph form with the Steelers’ last ten seasons highlighted.

In the last ten years the Steelers have not had a season where they allowed more than they scored, and they’ve been rewarded with seasons between 8 and 13 wins. That line through the data is the best fit line and reveals how many wins a team should expect given their point differential. In 2017, the Steelers dramatically overachieved. On average they should have only won 10.7 games given their point differential per game of 6.1. Their 13 wins was a one in twenty event. So the 13 win team must be dampered a bit because they really performed no better than an 11 win team. The team did win their share of close games last season, and that could be chalked up to good fortune or gritty experience. You can be the judge.

The 2018 Steelers did underachieve from a win point of view but not by much. They were expected to get 9.9 wins based on their 4.3 point differential per game, just 0.8 wins fewer than last year. They were notably worse in close games thanks in part to field goal kicking, indicating a loss of good fortune or said grit. Again, you can be the judge.

It’s worth noting that a 9-6-1 team would have made the playoffs in 5 of the last 10 seasons, so there too the Steelers were a tad unlucky. Would your opinion of the season be different with an appearance or win in the Wild Card round?

Nevertheless the team regressed by nearly a game, so let’s dig into what’s behind that drop.

Offensive efficiency and turnover woes

The offense put more points on the board in 2018, while the defense allowed even more. The offense scored 428 points against 406 the year prior, while the defense allowed 360 points compared to 308. But life isn’t that simple because that same offense coughed up the football 6 more times in 2018, which put the defense in more difficult situations.

Before analyzing further a heuristic needs to be established that helps us connect yardage gained to points. This analysis from Carnegie Mellon Sports Analytics provides that input.

Roughly speaking 10 yards gained is equal to about 0.7 expected points. Of course, not all of those points come to fruition on a drive by drive basis, but over the course of a season this heuristic is helpful. Below is a plot of actual points against expected points based on yards gained in 2018, and in the middle of the pack it does a decent job.

The measure greatly underestimates the efficiency of the Chiefs, Rams and Saints (the three points to the right). Giving up yards to those teams was more costly this season, while giving up yards to teams like the Cardinals was less costly. But on average gaining 10 yards earns a team 0.7 points, and that roughly held true in 2018.

With that tool in the kit we can determine the cost of the Pittsburgh’s turnovers. Here is a chart of the distribution of the opponent starting field position in 2018 versus 2017.

The Steelers allowed 5% more drives to start between their 20 yard line and the opponents’ 40 yard line in 2018. That’s 11 extra drives in total and 7 were following turnovers, so the incremental turnovers clearly put the Steelers defense in worse position. The cost of those turnovers turns out to be about 13 points to the opponent, which leaves the offense with only a 9 point improvement year over year. Given the loss of Le’Veon Bell and the embarrassment that was Chris Boswell, that is still a good story.

Pass protection and penalties

With those 13 points from turnovers moved to the offensive side of the ball, the Steelers defense still regressed to the tune of 39 points. Here’s what we can learn quickly from yards allowed.

The Steelers gave up 400+ more yards in 2018 on 53 more plays, including penalties. Some of that can be blamed on getting 7 fewer turnovers, but the rest is a function of more penalties and a worse pass defense. Penalties cost this defense 5 points year over year. A stingier rush defense which reduced their opponents’ yards per carry to 4.2 from 4.4 was a benefit that was undone by the pass defense. Here is a look at the distribution of outcomes from opponent passes that did not result in a penalty.

It’s worth noting that the Steelers pass defense did give up fewer big plays. They cut the number of passes going for 50+ yards by 62%. However, they saw a significant increase in passes resulting in 10 to 50 yard gains. That measure was up 4% year over year.

And here lies the crux of the Steelers’ woes in 2018. Instead of 30 opponent passing attempts that went incomplete or for negative yards, there were 20 more attempts that resulted in gains between 10 and 50 yards. That shift in results cost the defense 31 expected points.

Football certainly is about finding a balance. Perhaps a commitment to slowing the run and preventing the big plays that plagued 2017 allowed for that middle range pass to be more available. The bottom line is the balance resulted in a worse defense overall. Was this the cause of a linebacker group not particularly good at pass protection? Was this due to a drop-off in pressure? Was this due to the drop-off in Artie Burns and Mike Hilton’s production?

Wrapping it up

Yes, the Steelers regressed in 2018 but they didn’t fall as far as it appears. The offense modestly improved year over year even with the higher turnover amount, despite the loss of Bell, and Boswell literally and figuratively falling on his ass. The defense took a significant step back allowing 39 more points adjusting for the offense’s turnovers. Most of that regression was due to more penalties and a pass defense that allowed 20 more completed pass plays between 10 and 50 yards.

The list of offseason needs lines up with this narrative as inside linebackers and defensive backs are at the top of everyone’s wish list. But it’s worth repeating that while the team did take a step back, the performance wasn’t that much different than the 13 win season in 2017. The team didn’t change all that much. If only we could average the highs and lows of a 13 win season and the way we feel now.