0 Comments

Data Entry: Turning Over the Turnover Problem

| March 13th, 2017

This is the 3rd installment of a monthly offseason piece I’ll be doing here at DaBearsBlog, helping fill the content void of the long offseason. Each one will be a numbers-crunching look at something Bears related in which I attempt to earn the “Data” moniker so kindly bestowed on me by the comments section regulars and, more importantly, answer a Bears question that I’ve been wondering about. If you have anything you’d like me to look into, let me know in the comments or email me at woodjohnathan1@gmail.com and I’ll see what I can do.


Chicago’s defense has significantly improved in the last two years from the disaster that was the Mel Tucker era, but there is one area where they have actually regressed: forcing turnovers.

Tucker’s defenses in 2013 and 2014 actually forced turnovers at a slightly-above average rate (Tucker can probably thank the leftover Lovie Smith-trained players for that), while Vic Fangio’s defenses have forced fewer turnovers in the last 2 years than any other NFL defense. In fact, 13 defenses have forced as many turnovers in one season (28) as the Bears’ defense has the last two seasons combined.

The problem was particularly pronounced last year, when the Bears forced a measly 11 turnovers, tied for the fewest by any defense in the NFL in the last 10 years.

Given the strong and well-established relationship between winning the turnover battle and winning football games, this is a real problem for Chicago. All of this research looks at turnover differential, not just turnovers forced. But forcing turnovers is half of turnover differential and it’s the part I want to focus on today. Avoiding turnovers is largely a product of your quarterback (and luck for fumbles/fumble recoveries). That’s a separate issue that has already been discussed on here at length.

Setting it up

Here’s my question: What is the history for teams the year after they have forced as few turnovers as the Bears have recently? Does the defense continue to struggle generating turnovers, or does it improve quickly?

Here’s how I approached the study:

  • I looked at turnovers forced by all teams each year for the last decade, so dating back to 2007 (full data here). The only factor considered was turnovers forced.
  • Turnover rates have generally decreased over the last 10 years. Specifically, the average varied from 22 to 28 turnovers forced in a season, with highs ranging from 33 to 48 and lows ranging from 11 to 19. Thus just looking at raw numbers isn’t that helpful, so I looked at z-scores instead. Basically a z-score is a measure of how far from average you are for that data set (that season). More negative z-score = fewer turnovers forced relative to league peers.
  • The Bears’ z-score was -2.0 in 2016 and -1.3 in 2015. Thus I looked at two data sets: teams with a single season -1.5 or lower and teams with 2 consecutive seasons -1.0 or lower.

One Terrible Season

There were 15 instances between 2007 and 2015 where a defense’s turnover z-score for a single season was -1.5 or worse (I didn’t include the 2016 season because we don’t have follow-up data yet). In the seasons following that turnover-deficient season, the average z-score improved to +0.5, meaning that teams, by and large, forced slightly more than average turnovers the following season.

Of course, the z-scores for the following season varied from -1.6 (remaining awful at forcing turnovers) to +2.1 (among the league leaders in turnovers forced), so this in and of itself is no guarantee that anything will change. Still, it is encouraging to see that a rapid recovery from a poor turnover year is not only possible but has a large amount of recent precedence.

Given the relationship between turnover differential and win %, it is unsurprising to see that this improvement in forcing turnovers brought with it an improvement in winning football games. These 15 teams improved from an average of 6.0 wins to an average of 8.4. Again, there are a variety of outcomes, from 9 win improvements to 4 win regressions, meaning there are certainly no guarantees, but it is encouraging to see that improvement is likely. To be fair, there’s not much room to get worse from 3-13–these aren’t the 2008 Detroit Lions, after all–so expecting a modest win improvement from the Bears next year isn’t exactly a bold statement.

Of course, it could be argued that much of this change in turnovers and success is due to a change in scheme. Many teams fire head coaches and/or defensive coordinators after a terrible season, and 8 of the 15 in this sample did just that. The results are slightly less positive but still promising for the 7 teams that kept both their head coach and defensive coordinator like the Bears did this offseason. The average z score was still slightly above average at +0.2, indicating teams basically forced turnovers at an average rate the next year, albeit with a wide swing in outcomes. These teams only increased their average wins from 8.0 to 8.4, likely because the average wins initially was much higher (and thus teams didn’t feel the need to fire their coaches).

Consecutive Bad Seasons

Of course, it might look different for a team that has had two straight poor seasons at forcing turnovers. That might indicate more of a personnel deficiency than a single season of bad luck. With that in mind, I looked at teams who had 2 straight seasons with a z-score of -1.0 or less; this set up a sample size of 6 teams between 2007 and 2015, showing you just how bad Chicago’s defense has been at forcing turnovers the last two years.

Looking at averages in a 6-team sample size is kind of pointless (even the 15 above is too small to draw definite conclusions from), but every team except 1 in this sample had a defense in year three that forced turnovers at an average rate or better. 4 of the 6 teams forced turnovers at basically an average rate (z-scores between -0.1 and +0.3), 1 had a great turnover-forcing defense (z-score of +1.5), and one stayed pretty poor at forcing turnovers (z-score of -1.1, with an improvement to average in the 4th year after 3 straight bad years).

Again, this is too small of a sample size to say anything definitively, but I think the main point to make here is that it’s hard to be as bad as the Bears have been at forcing turnovers for too long. There’s a large amount of luck involved (though certainly plenty of skill as well), and sooner or later bounces start to fall your way a bit more. Talent also gets upgraded before too long if players aren’t performing, which can then lead to more turnovers as well.

And again coaching changes could factor in, with 3 of the 6 teams changing coaches after year 2. All 3 who kept their coaches saw the defense around average at forcing turnovers; while the 3 who changed coaches saw their results much more spread out.

We also see general improvement in wins as more turnovers are forced, with teams going from an average of 6.3 wins in the 2 seasons with few turnovers forced to 9.2 when the turnovers start to come in year 3. Those numbers change to 6.3 and 10.7 if you only look at teams who kept their coaching staff after 2 years of bad turnover-forcing defenses.

Conclusions

It’s hard to draw any definitive conclusions from small sample sizes, so I’m not going to say “expect the defense to force turnovers at an average rate in 2017” or anything that specific. Instead, the main conclusion is this: there are very few teams in the last 10 years that match the Bears’ level of turnover-forcing ineptitude from 2016 or the last 2 seasons combined. That tells you just how hard it is to be that bad, and therefore it’s logical that the trend won’t continue forever.

There are no guarantees that the Bears force more turnovers in 2017, but it’s highly likely they will. Only one team in the last decade has had a turnover drought like the Bears’ last for 3 years. The combination of improved luck and better talent (started in free agency, will continue in the draft) should lead to the Bears forcing more turnovers in 2017 than they have either of the past two years. Logic dictates that more wins should follow. How many more turnovers and wins remains to be seen.

Tagged: , , , ,

© Da' Bears Blog