Simulating the 2019 NFL playoffs under the new playoff format

The coronavirus offseason is in full swing. The XFL came and went like a flash, and while we were lucky enough to see the 2020 NFL Scouting Combine and NFL free agency come to fruition, things like OTAs, pro days and the like have almost all been canceled. There are, however, plenty of things to bet on during this time, like win totals, futures and draft props.

In this article, though, we aren’t going to look to the future. We’re going to look at the past and take a look back at what could have been during last January’s playoffs.

The recently approved collective bargaining agreement, followed by the subsequent owner’s vote, has increased the NFL playoffs from six teams in each conference to seven, expanding the wild-card playoff round to three games with the rest of the playoffs going as usual (with the second seed possibly replaced with the second-best-remaining seed getting a home game but not a bye).

This is the NFL’s first change to the playoff structure since 2002 and the first increase in the size of the playoffs since 1990. Prior to 2002, when there were three divisions per conference, the wild-card team with the best record was granted a home game against the fifth seed, while the three division winners got the first through third seeds.

Before 1990, the playoffs had five seeds, with all three division winners earning byes, and there was only one wild-card game per conference during wild-card weekend. 

A big question is what the impact of the second seed losing the bye will be. Before 1990, only three teams in NFL history made the Super Bowl without first securing a division title and hence the bye that goes along with it (the 1985 New England Patriots, the 1980 Oakland Raiders and the 1975 Dallas Cowboys).

Lower seeds had something of a successful run of it for a while after expansion, with the 1997 Denver Broncos, 2000 Baltimore Ravens, 2005 Pittsburgh Steelers, 2007 New York Giants and 2010 Green Bay Packers all winning the Super Bowl outright as wild-card teams, with a host of other teams winning without a first-round bye (the last being the 2012 Ravens).

However, since 2013 no team has even made the Super Bowl without the benefit of the bye, leading many to question whether granting one to only one such team per conference bestows way too much of an advantage to that team.

To test this, we ran our playoff simulation for the 2019 playoffs and compared them to what we had going into this January. The seventh seeds for the NFC and AFC would have been the Los Angeles Rams and the Pittsburgh Steelers, respectively. Here are a few nuggets from this exercise:

[Editor’s Note: PFF’s simulations and projections are powered by AWS machine learning capabilities.]

Even with a Weak Seven Seed, the Two Seeds Get Hurt the Most (as Expected) While The Third Seeds Benefitted

Kansas City and Green Bay both took a hit to their Super Bowl and Super Bowl champion odds, as you would expect since both teams went from 100% to make the second round to substantially less than that. 

Kansas City went from our second-most-likely team to win the Super Bowl, at 19%, to a longer shot of 15%. Green Bay fell even further, as we gave them an 11% chance at the beginning of January, whereas adding another game to their path dropped it to 7% since we would have them beating the 9-7 Rams just under 65% of the time.

The two biggest beneficiaries were understandably the three-seeded Saints (11% to 15%) and Patriots (7% to 10%). Pittsburgh, having to start a backup quarterback, had roughly a 0.5% chance to win the Super Bowl in the new system, while the defending NFC champion Rams had just over a 1% chance, which was higher than Houston and Buffalo out of the AFC.

The Ravens Wouldn’t Have Benefitted Much From the New System, and the 49ers’ Chances Got Worse

The Saints’ wild-card loss to Minnesota obscured the fact that they were very likely not the third-best team in the NFC, so a circumstance where they were on equal footing with Green Bay actually had a negative impact on the NFC’s highest seed, the San Francisco 49ers. This is what we see, as the 49ers’ probability of making and winning the Super Bowl dropped from 35% to 33% and from 15% to 14%, respectively. We would have made the Packers less of a favorite to beat the Rams on wild-card weekend than the Saints a favorite to beat the Vikings.

While Kansas City was a better team than New England in 2019, the Patriots would have been a decent substitute for them should the Chiefs have failed to make it past the first round (which happened in only 23% of simulations). As we saw when the real playoffs took place in January, the favorites are a lot less likely to make it through lower seeds than many think, curbing some of the perceived benefits of being the only team with a bye. 

Conclusion

Concerns about a lack of parity in this new system are probably a bit overblown, as our simulation suggests that, at least for the 2019 playoffs, the effects of taking away a bye from one team in each conference has more of a negative effect on that team than a positive effect on the team above them in the standings.

The 2019 season was a unique season in the sense that the third seed for the NFC was, in almost every way other than head-to-head matchups with the Minnesota Vikings and a random tiebreaker, better than the second seed, and leveling the playing field greatly benefitted the former. 

While a full exploration of relative team strengths and their impacts on this new format was beyond the scope of this exercise, it appears at least from 2019’s composition of playoff teams that the consolidation of power to the one seed is less of an issue than many believed it to be. Instead, parity between the second and third seeds increases, which — in the case where those teams are equally yolked — may not be a bad thing.