What Are the Odds the Vikings Defense Regresses?

Mandatory Credit: Brace Hemmelgarn-USA TODAY Sports

It’s no secret that the Minnesota Vikings have been driven by the strength of their defense since Mike Zimmer took the head coaching job — and honestly, several years before that as well.

Despite a persistent emphasis on defense, the Vikings reached heights last season that they hadn’t seen sight of in decades. In total points allowed, for example, they ranked first. The last time they ranked in the top three was 1988, where they ranked second. In Football Outsiders’ defensive DVOA, they ranked second.

The last time they ranked in the top three was 1993.

While the team has invested heavily in offensive upgrades, their choices in this year’s NFL draft spell out their priorities: win with defense.

With that in mind, it’s worth asking whether or not the Vikings can maintain this incredible defensive performance — will the Vikings defense regress?

There are a dozen ways to dissect this question, and it’s worth going over the major indicators of regression. How do defenses typically do after hitting a peak like the Vikings?

The 2012 Seahawks ranked first in points allowed and maintained their supremacy for three additional seasons before dropping from first to third, then in 2017, hitting 13th — high-level defensive play for five years. On the other hand, the 2009 Cowboys hit second in points allowed, then followed that up by ranking 31st.

They didn’t crack the top 10 again until 2016.

Since 2008, generally speaking, 40.6 percent of top three defensive teams stayed in the top three. Just over half of the teams in the top three (55.6 percent), maintained a top-five status the following year.

On the other hand, the teams that fell out, fell out hard — 37 percent fell outside the top 10 and 18.5 percent fell outside the top 20.

An even better measure of defense than total points is Football Outsiders’ DVOA metric, which combines the concepts of success rate with explosive playmaking capability and adjusts for down, distance, time remaining and point differential and then once again adjusting for opponent strength.

There’s a lot that goes into their calculations, but the gist of it is that it ends up being a pretty great predictor of how teams will do in-season — more than most other advanced approaches — and even better than point differential and past wins to predict future wins.

A solid ranking there usually tells us that a team is genuinely performing well in a sustainable manner. Moreover, it’s a decent guard against the concept of regression — a team doing well in DVOA one year should, generally, do well the following year.

The Vikings defense happened to do extraordinarily well in DVOA; they finished second overall in 2017 to the Jacksonville Jaguars — marking theirs as an elite defense. There are a lot of reasons to believe that it should maintain that ranking. The defense has improved every year under Zimmer, and they haven’t hit a ceiling. Though one of the two best defenses in 2017, it hadn’t hit the heights of historic defenses like the 2015 Broncos or 2013 Seahawks.

The problem is that it is pretty rare for a top defense to stay at the top and Football Outsiders regularly projects high-level defenses to fall off — much more than offenses. Defenses tend to be more unstable from year to year than offenses, as much as people view defenses as rocks of stability from which to build their teams off of.

That said, the Vikings were not the kind of extremely good defense that should see a dramatic regression to the mean by a simple matter of course. “Great” defenses tend to regress towards becoming “good,” while “good” defenses tend to stay “good.”

From the perspective of DVOA, the Vikings shouldn’t expect much regression — though we’ll need to look at the most common indicators of regression to see whether or not the Vikings are susceptible.

Injury Luck

There is virtually no relationship, year-to-year, in injury luck for teams, and that specifically carries over when looking at injuries across units in the same year — teams who benefitted from injury luck on defense did not do so on the other side of the ball, and vice versa.

The Vikings defense had the best injury luck of any defense over the past four years, per Football OutsidersAdjusted Games Lost metric, which evaluates missed games as well as games impacted by injury-ridden play, like players on the Questionable or Doubtful list who nevertheless play. The four years mark might be an underestimate, as Football Outsiders doesn’t seem to have their AGL data split into different units earlier than 2014.

While this doesn’t fully capture the unique injury situation of Xavier Rhodes, who started every game in 2017 but missed 10 percent of team snaps, per Pro Football Focus.

On the other hand, after excluding games that starters were benched — Week 11 against the Rams and Week 15 against the Cincinnati Bengals — Rhodes only missed 66 snaps, which is about one game.

In all likelihood, the Vikings defense will suffer unpredictable injuries. Some of the players on the defense have had injury issues in the past, but generally, it will be difficult to determine which players will miss playing time.

The top three defenses in AGL in 2014 ranked seventh, eighth and 17th in 2015. The top three in 2015 ranked 18th, 20th and 22nd in 2016. The top three in 2016 ranked ninth, 10th and 19th in 2017. No top-three injury team finished in the top five the following year while more finished in the bottom half of the league than the top half.

Turnover Rate

What’s unusual about discussing regression is that the things teams are lauded for during the season can often be “warning signs” in the offseason. Injury rate is one example, and turnover production is another – teams should produce turnovers on defense, but a high turnover rate can often be a worrying sign for predicting future performance because they are largely unsustainable.

Turnovers statistics do an excellent job providing us with information about how teams won or lost games, but they don’t do well predicting future outcomes, especially from year-to-year.

In 2014, the top three defenses in takeaways finished 13th, 15th and 32nd in takeaways the following year. In 2015, the best takeaway teams finished sixth, ninth and 13th next year. In 2016, the top three defenses in takeaways finished seventh, eighth and 31st in takeaways.

Over that time span, the nine teams that finished in the top three followed up their performances with four top-10 finishes and four finishes in the bottom half, with two of the teams ranking last or near last in the category.

Mandatory Credit: Brace Hemmelgarn-USA TODAY Sports

Because of the outsized effect turnovers have on games, a sudden decrease in turnover production can result in massive swings on the scoreboard. The 2006 Bears had one of the best turnover rates of all time, and finished third in points allowed as a result. The following year, their turnover production dropped to about league average – and they dropped to 16th in points allowed.

The same things have happened to the best takeaway defenses over the past 15 years. The 2012 incarnation of the Bears went from third in points allowed to 30th. The 2007 Chargers went from fifth to 15th. The 2003 Rams went from 17th — already a poor mark — to 25th. The 2006 Ravens went from first to 22nd.

Luckily for the Vikings, they managed to dominate offenses without the help of gaudy turnover production, finishing 23rd in total turnovers. Because turnovers can be just as much about opportunity, it’s worth noting that they also finished 22nd in turnovers per drive and 22nd in turnovers per play.

If anything, they’re likely to get better in the turnover department as they regress towards the league mean and what is likely an even higher “true” mean.

Turnovers aren’t random — they are a product of both skill and luck — but the fact that they are such low-frequency events means that they don’t reflect true skill over the course of one season. Better defenses will tend to produce more turnovers over time, and the safe bet is to assume that the Vikings will improve in this department, not get worse.

Success Rate

The other side of “fluky” turnover production is the consistent and sustainable ability to stop offenses from moving the ball. A great way to measure that is the ability to stop first downs.

The Vikings defense set records during the regular season at preventing third down conversions — better than any team since 1991, when it was first recorded as a statistic — and that ended up getting reflected in their first down rates overall across all plays.

A good way to measure first down stoppage is to look at something called “Drive Success Rate,” which measures how often teams score first downs or touchdowns and divide that number by the number of down series they had.

As an example, a team that receives a kickoff and converts three first downs before punting will have had a drive success rate of 75 percent because they had four down series — the possession after the kickoff, as well as a new set of downs every time they converted one.

Had that team scored a touchdown on that drive, they would have had a drive success rate of 100 percent.

This is a good measure of year to year stability, as much as any exist. Of the nine teams that finished in the top three between the years of 2014 and 2016, four repeated their appearances in the top three, with another two teams finishing fourth and sixth in their following years. Only one team fell out of the top half.

That’s excellent news for the Vikings, who made up for their mediocre turnover production by preventing first downs; they finished second in the league in drive success rate, only getting bested by the Jacksonville Jaguars.

This is also reflected in other drive statistics, like score percentage (the Vikings ranked third in having the fewest opponent drives end in scores) and touchdown percentage (second in having the fewest drives end in touchdowns).

For the most part, this is somewhat strong evidence that they should be able to carry their success over from year to year.

Strength of Schedule

There’s a decent argument to be made that if the Vikings have to play against a tough schedule this year and their schedule last year was fairly weak, that they could be due for some regression. After all, the schedule matches them up against other division winners.

This likely won’t be a significant problem, as the Vikings rank in DVOA — second — proves that even after accounting for opponent strength, they did well. The Vikings also ranked first in opponent-adjusted points allowed, per Pro Football Reference.

It still may be worth examining whether or not the upcoming schedule could represent a tough reality check.

Opponent strength is notoriously difficult to predict going forward. Analysts often use the win percentage of a team’s opponents when attempting to forecast a schedule’s strength, but those don’t bear out because of how imprecise “wins” are at measuring true quality.

Opponent point differential does a much better job, and in that measure the Vikings look good; their schedule is a bit easier than average, ranking 18th overall instead of the 8th overall ranking they get just from looking at opponent wins.

Unfortunately, that doesn’t give us much insight into the defense itself. If the Vikings offense played easy defenses while the defense played tough offenses, then the defense had a difficult strength of schedule, even if the team overall did not.

Mandatory Credit: Geoff Burke-USA TODAY Sports

Instead, we should look at how many points teams on this year’s schedule produced last year.

The Vikings have the fifth-most challenging schedule from that perspective, with high-powered offenses like the Eagles, Saints, Rams and Patriots. It should help that the Vikings played two of those offenses during the regular season last year, but this is a more difficult regular-season schedule from that perspective than last year, where the Vikings ranked 12th.

It’s not an enormous spike in difficulty, but it might have an impact. On the other hand, the Vikings schedule from a DVOA perspective looks better than it does from a points perspective, and we know that’s more predictive of future seasons.

Last year, Minnesota’s defense had the 13th-most difficult schedule in the league last year, which is similar to the ranking produced through total offensive points. Next year, the total offensive DVOA of opponents only ranks 20th.

This difference is largely due to several high-powered offenses overperforming from last year, teams that Football Outsiders may expect to regress next year. The biggest offender is San Francisco, but there should be some minor regression on its way for the Los Angeles Rams and Philadelphia Eagles. Even underwhelming offenses overperformed; Arizona ranked 25th in points but, even worse, ranked 30th in DVOA.

With all of that in mind, strength of schedule should not be a concern for Minnesota when projecting their defense for next year.

Roster and Staff Losses

Many teams that suddenly drop off in defensive performance do so as a result of coaching changes or roster shakeups.

The 2014 Buffalo Bills ranked second in defensive DVOA but fell to 24th after overhauling their coaching staff and switching defensive systems. The 2013 Chicago Bears improved significantly on offense but took a beating defensively after switching from Lovie Smith to Marc Trestman — dropping from first to 25th.

The 2015 Detroit Lions lost both starting defensive tackles in free agency, including Ndamukong Suh, to drop from third to 16th. Houston’s defensive performance seems to track with J.J. Watt’s presence, and Seattle took a sharp turn after losing Richard Sherman and Earl Thomas to injury.

Going further back, the 2008 Tennessee Titans fell from second to 28th in part because of losing defensive coordinator Jim Schwartz to the Detroit Lions and Albert Haynesworth to Washington.

Barring injury, this isn’t an issue for the Vikings. Despite fielding one of the top defenses in the NFL, they didn’t suffer coaching turnover and will keep their position coaches and defensive coordinator. They also will only lose one starter on defense — a player who the Vikings upgraded from with Sheldon Richardson taking over for Tom Johnson.

They even returned a good chunk of their depth, with Brian Robison and Terence Newman agreeing to return for one more season. Though Shamar Stephen’s departure might concern those who worry about depth, draft picks and last year’s rookies may be able to fill in.

We can check which measures repeat from year to year and which metrics are often products of good fortune by expanding our sample to the past ten years instead of the past three.

The below table looks at the measures discussed above and breaks down how often teams in the top three of a particular metric repeat their appearance into the top three and top five of the NFL the following year.

On the bottom row, I’ve also used a random number generator to simulate a truly random statistic to compare against.

Measure Top 3% Top 5%
Points 40.7% 55.6%
DVOA 37.0% 44.4%
Drive Success Rate 37.0% 44.4%
Turnover Rate 18.5% 28.9%
Adjusted Games Lost 16.7% 27.8%
Random Number Generator 18.8% 25.0%

Overall, there are more reasons to think the Vikings defense will repeat — or even improve — on their impressive performance from last year.

They placed in the top three of every “sustainable” category of defense and don’t have to worry about roster or staff turnover. The only potential hitch is an increase in injury rates, but with the investment in depth the Vikings have made, the impact of that might be minimal.

Expect another season of defensive dominance from the Vikings.

Become a Zone Coverage Member Today!

Kene Nwangwu Will Be Something To Look Forward To After the Bye
By Eli Kuphal - Oct 22, 2021
The Vikings Can Be Sellers and Still Stay Competitive
By Tanishka Mhaskar - Oct 22, 2021

What Are Realistic Expectations For Dantzler?

Mandatory Credit: Brace Hemmelgarn-USA TODAY Sports

There was a lot of optimism surrounding Cameron Dantzler this year. He was coming off a rookie season that had its ups and downs but finished strong […]

Continue Reading