120 Comments

2019 Bears: “No Reason to Think They Won’t Be Really Good Again”

| July 29th, 2019

I recently looked at the history of teams to make a significant improvement from one year to the next and found that many of them win fewer games the year after their breakthrough season. This suggests that the Bears might be due for a bit of a letdown from their 12-4 record in 2018 (though they do match the profile of teams that generally stay good after making the jump.)

Today, I want to look more closely at Chicago’s underlying performance in 2018 to see if there’s anything there to suggest they are a team poised for a fall. This is closely modeled after work Bill Barnwell does every off-season, where he uses three factors to identify teams who are likely to improve and likely to regress.


Pythagorean Expectation

The first factor is called the Pythagorean expectation, and it is a measure of how many games a team is expected to win based on how many points they scored compared to how many points they allowed. The exact formula can be seen here, but the general idea is that truly good teams score a lot more points than they give up. Teams that win a lot of games without a large difference in points scored/allowed were considered more lucky than good and are likely due for a fall.

  • 2018 Bears stats: 421 points scored, 283 points allowed, 12 wins
  • 2018 Pythagorean expectation: 11.5 wins

The Bears didn’t significantly outperform their Pythagorean expectation, which means they won a lot of games because they were legitimately good, not lucky. So far, there is no reason to think that significant regression is coming.


Record in Close Games

The 2nd factor looks at how well teams performed in close games, which Barnwell defines as having a final scoring margin within 7 points. I think 8 points makes a lot more sense given that’s still a one possession game, but in this case it doesn’t change anything for the Bears, so we’ll stick with 7.

Read More …

Tagged: ,

54 Comments

Possible Regression in 2019: A Mathematical Analysis

| July 22nd, 2019

For the first time in a long time, the Chicago Bears were legitimately fun to watch in 2018. Following years of terrible, boring teams, they went 12-4, scored some big man touchdowns, had plenty of awesome celebrations, and started the most exclusive club in the country to celebrate their wins.

But 2018 was last year, and now I’ve seen some worry that it will prove an aberration. They point to the 2017 Jacksonville Jaguars, who made a similar jump from years of awful to a division win and playoff berth before falling back to Earth in 2018, as a sign of what is to come.

While I’ve been on the record going back to 2017 that this is the earliest year when their title window will fully open, I still wanted to take a realistic look and see if there might be reasons to expect regression in 2019 instead. Accordingly, I’m looking at recent NFL history to see how teams similar to the 2018 Bears followed it up the year after. Since the NFL switched to its current 32 team, 8 division format in 2002, that serves as a nice starting point for this study. I looked at wins per year for all teams from then to 2017 (the last year in which we can track how teams did the year after), and identified teams similar to the Bears in a variety of ways. Full data can be seen here.


12+ Wins

To start out, I looked at teams that won 12 or more games in a season, as Chicago did in 2018. That data can be briefly highlighted like so:

  • Teams with 12+ wins: 74
  • Average # wins: 12.7
  • Average # wins next year: 9.6
  • Net change: -3.1 wins

The average team that won 12+ games decreased by just over 3 wins the following year, which makes a lot of sense. It’s hard to win 12 games in a season, which is why fewer than 5 teams per year, on average, do it. Remaining one of those top 5 or so teams for a 2nd year straight is no small feat.

Read More …

Tagged: ,

100 Comments

Will the Bears Defense Regress in 2019? History Tells Us…Not So Much.

| July 15th, 2019

Chicago’s defense was really, really good in 2018. They led the NFL in points allowed, turnovers forced, touchdowns scored, and passer rating against, and finished 3rd in both yards and sacks. They finished as the runaway best defense in Football Outsiders’ DVOA, which is intended to be an all-encompassing metric, and even finished as the 8th best defense ever in DVOA’s database, which runs back to 1986.

Now as we head into 2019, fans are rightly wondering if Chicago’s defense can repeat that performance. While I won’t pretend to be able to predict the future, I can look at the past to see what it might have to tell us. So I looked at top defenses in recent NFL history and measured, through a variety of metrics, where the 2018 Bears excelled. Then I looked to see how they followed that up in the next season. Full data collected can be viewed here for transparency’s sake.


DVOA

The DVOA system is set up such that an average defense gets a score of 0, with negative numbers indicating you are better than average (the farther from 0 the better). The Bears finished with a final score of -26.0, so I looked at other teams in the last decade (2008-17) who finished at -20 or better. This was quite a small list, as it featured only 10 teams. Here’s how they fared in the season following that dominant performance:

  • Average DVOA: -25.1%
  • Average following DVOA: -8.8% (8th in NFL)
  • Change: 16.4%
  • # teams with better DVOA following year: 0
  • # teams top 5 in DVOA following year: 5
  • # teams top 10 in DVOA following year: 8
  • # teams below average in DVOA following year: 1

First, notice that none of these defenses were as good the following year. This isn’t surprising; there were only 10 teams in 10 years who achieved this caliber of DVOA. The odds of doing that twice in a row are very low.

Read More …

Tagged: , ,