vikings

Previous Story Back Page: Deadspin needles Vikings with commemorative Love Boat brick Next Story Zulgad: Mike Zimmer’s steadying hand is instrumental to Vikings’ success

Arif’s deep stats dive: Vikings can be proud of win, but big holes still exist

Sep 18, 2016; Minneapolis, MN, USA; Minnesota Vikings wide receiver Stefon Diggs (14) catches a touchdown pass past Green Bay Packers cornerback Damarious Randall (23) in the third quarter at U.S. Bank Stadium. The Vikings win 17-14. Mandatory Credit: Bruce Kluckhohn-USA TODAY Sports

BY ARIF HASAN

Follow Arif on Twitter

Each week, we’ll be breaking down advanced statistics to help give us an even better picture of the Vikings’ games beyond the traditional box score. Last week, looked at deeper statistics at the team level and individual level, and that piece should provide a good primer for any statistics discussed here that might need further explanation.

Team-Level Statistics

We know that volumetric measures, like total yards and points tends to quash nuance, like pace and the overall ability to convert first downs. Teams with good offenses that tend to operate at a slow pace, like Seattle’s last year, don’t quite get the full credit they deserve.

The lifeblood of any offense is first downs, and the goal is touchdowns. A statistic like Drive Success Rate accommodates both while giving credit to offenses that convert a new set of downs outside of the traditional third down conversion rate statistic.

Football Outsiders keeps track of the statistic and you can see season totals stretching back to 1997 here, but for now we’ll look at the totals in each game.

The Vikings’ offense had had 11 drives, 13 converted first downs and two touchdowns. That means they had 24 opportunities to convert a first down or touchdown and made good on 62.5 percent of them. That’s identical to last week, and it ranked 27th among offenses last week.

The Packers had 12 drives, 18 converted first downs and two touchdowns. That means they had 30 opportunities to convert and did so on 66.7 percent of their attempts. That would have ranked exactly 24th.

Does this mean the Packers offense was better than the Vikings offense? Not quite, because DSR does not take into account turnovers, which have a strong impact on field position. But it does put into light the significant difference that special teams had on the game, which lies beneath the surface of all of these statistics.

The Vikings did end up having to punt seven times, and that’s not a sign of an efficient offense. It helps that the Vikings averaged more per kick return (by ten yards, a significant amount). The Packers also only pinned the Vikings inside the 20 yard line once on punts, while the Vikings did the same to the Packers five times.

Still, a negative 0.042 drive success rate is definitively below average, and would have ranked 22nd among teams last year.

All of this culminated in the Vikings scoring 1.55 points per drive, though 1.70 seems closer to accurate as they did not attempt to score on the final drive. The Packers fared worse, naturally, with 1.17 points per drive (1.27 points per drive when excluding their meaningless end-of-half drive).

Offensively, 1.55 points per drive typically ranks in the bottom five, and 1.70 is only a little bit better. But that does mean that defensively, the Vikings were pretty on-point—any team average between 1.17 to 1.27 points per drive would have been the worst in the NFL last year.

Netting 0.38 points per drive is pretty good and would have tied for 10th overall last week.

Overall, the team-level success seems to imply that a largely lackluster offense (likely due to its running game) is being boosted by excellent field position as a result of hidden yards on special teams and short fields created by the defense.

Passing

We’ve got a good bead on why the offense is putting up such lackluster numbers, and it certainly seems like the second consecutive game with a sub-2.0 yards-per-carry rusher must be at fault—but we can’t really check that out unless we strongly evaluate the passing game.

Sunday night looked like a spectacular coming-out party by Sam Bradford to pay the Vikings back nearly immediately for the draft assets they gave up to get him. Bradford’s passer rating certainly fit that mold, and with a rating of 121.2 (125.1 when excluding that final pass designed to waste time) ranking fifth of quarterbacks this week.

But we looked at more than just passer rating last week. We also looked at ESPN’s Total QBR, which has the strongest track record of predicting success, adjusted net yards per attempt—easy enough to calculate on our own—and implied yards per attempt, which corrects for drops and throwaways.

Again, a full primer will be found at the original Week One article, but suffice to say that adjusted net yards per attempt (ANYA) corrects some of the crazy values found in passer rating while also incorporating sacks. Implied yards per attempt looks at accuracy (any pass that results in a completion or drop is “accurate” while throwaways, spikes and batted passes are excluded from the “attempts” in the denominator) and adjusts for the average distance downfield a quarterback attempts a target.

Bradford’s 8.4 ANYA is excellent and though it ranked ninth among passers in Week 2, would have (like his passer rating) ranked first in 2015 over the course of a full season.

That’s commensurate with the grade he received from Pro Football Focus, which made him the second-highest graded quarterback of the week behind only Ryan Fitzpatrick.

His “implied” yards per attempt was pretty fantastic, too. After combining his accuracy with his depth of target, the former Oklahoma Sooner ranked fourth among all quarterbacks with a functional YPA of 8.71. Of those above him, two threw multiple interceptions.

But Bradford’s ESPN QBR is much lower (64.3), ranking a mere 16th among passers this week and nearly identical to Shaun Hill’s mark of 63.0 from last week.

What gives? Bradford ranks second, fourth, fifth and ninth among a few metrics, and the ninth-ranked ANYA looks much better in the context of a whole season.

There are likely a few reasons. The first is that the quarterback seems to take a harsher penalty for sacks in QBR than they do in ANYA, relatively speaking. Beyond that, Stefon Diggs grabbed the second-most yards-after-catch among receivers with at least five catches (fifth among those with at least three receptions). That’s 78 yards of YAC, per PFF.

That shouldn’t have too much an impact; the Vikings’ passing offense ranked 17th in percentage of yards that came after the catch.

But there is a chance that because ESPN uses an “expected points” formula (as determined by how many points an average team scores from that down, distance, field position and time remaining on the clock) that Bradford’s QBR and passer rating don’t match because he had three pass attempts when the Vikings were at 3rd and 12+, a situation where defenses willingly give up passing yards and offensive coordinators do not generally scheme for conversion, setting for field position gains.

Without his three completions on third-and-unmanageable, which produced yards and completions but accomplished nothing, Bradford loses 40 yards. But even after eliminating those yards and completions, Bradford’s passer rating drops to a “mere” 119.0 and his adjusted net yards per attempt is at 7.94, an elite number that wouldn’t rank No. 1 last year, but No. 3.

It’s also not as if down-and-distance have a large impact; he averaged 9.7 yards per attempt on first down and 10.4 yards per attempt on second down.

In all likelihood, this measure is giving us poor information relative to all the other statistics about Bradford’s outing. Here’s how he ranked in the statistics mentioned above among Week 2 passers and how those ranks would have done if measured against the 2015 season:

Metric Weekly Rank Rank vs. 2015
PFF 2 8
Implied YPA 4 3
Passer Rating 5 1
ANYA 8 1
ESPN QBR 16 14

As Bradford accumulates more starts, we’ll get year-to-date rankings going, too.

Also notable in the passing game: the Vikings offensive line allowed four sacks while ten players logged quarterback hits on Sam Bradford. In general, Bradford was pressured on 48.6 percent of snaps, per Pro Football Focus. That’s remarkably high; the Vikings had the highest pressure rate in PFF’s tracking history (a decade) last year with a rate of about 46 percent over the course of the season.

Receiving

We isolated two receiving stats last week. The first was drops, which is pretty subjective. Kyle Rudolph, for example, could have had a drop on Sunday night but Pro Football Focus did not log that incompletion as a drop.

There were no drops logged at all by the folks at PFF and the folks at Football Outsiders do not have drops logged quite yet. STATS, Inc. records no drops from last night as well.

The other statistic we looked at was yards per route run, which looks at efficiency from an on-field opportunity angle, instead of a per-target angle. Below are the yards per route run receivers for the Vikings produced, how that ranked among their positional peers last week and how it would have ranked in 2015.

Player YPRR Rank Rank vs. 2015
Stefon Diggs 5.52 1 1
Adam Thielen 1.38 40 67
Charles Johnson 0.45 73 126
Cordarrelle Patterson 0.00 81 127
Kyle Rudolph 0.94 35 69
Adrian Peterson 0.23 18 54
Jerick McKinnon -0.13 62 59
Matt Asiata 1.10 44 17

Below are the year-to-date yards per route run for receivers the Vikings produced and how they rank.

Player YPRR Rank
Stefon Diggs 4.19 1
Adam Thielen 1.77 36
Charles Johnson 0.30 92
Cordarrelle Patterson 1.20 64
Matt Asiata 0.69 34
Adrian Peterson 0.12 61
Jerick McKinnon 0.10 62
Kyle Rudolph 1.5 19

Stefon Diggs ranks first in total yardage and yards per route run in the NFL so far. There’s probably reason to be concerned about the secondary receiver position, but once Treadwell cracks the starting roster, the problem may be resolved.

Running

Last week, we looked at several running back statistics to give us a bigger picture on how they performed and what kind of running backs the Vikings currently have. While none of the statistical tools are definitive, they each give us a different picture.

Once again, some of these statistics are dependent on the offensive line. Success rate, for example, can be pretty high for a running back with low yards per carry and that might be the result of an offensive line that consistently creates two yards of room from the line of scrimmage—though decisionmaking and quickness are a big part of that as well.

Some clearly have mixed responsibility. Yards before contact seems like an offensive line statistic, but running backs that switch teams often have similar yards before contact despite the ability of the offensive line in front of them. Like success rate, yards before contact can be dependent on the offensive line as much as it is on vision, speed and agility.

We also discussed yards over expectation, which is a metric that corrects for the fact that it is easier to get running yards on some down and distances than others.

Below are the ranks for the running backs relative to the rest of the league this week in those statistics. We’ll compare to running backs with at least ten carries (27 this week).

Player Success% Rank Yds vs Expctd Rank Yds Before Cont Rank Yds After Cont Rank
Adrian Peterson 50.0% 4 -2.69 27 -0.2 27 1.8 26
Jerick McKinnon 0.0% 28 -1.18 23 0.5 21 0.5 28
Matt Asiata 16.7% 26 -1.27 24 0.4 22 1.8 26

And compared with running backs year-to-date with at least 15 carries (40 running backs)

Player Success% Rank Yds vs Expctd Rank Yds Before Cont Rank Yds After Cont Rank
Adrian Peterson 35.5% 34 -2.52 40 0.1 38 1.5 40
Jerick McKinnon 0.0% 41 1.21 8 1 27 1.9 36
Matt Asiata 30.0% 40 -1.16 36 0.7 34 1.7 38

The returns aren’t good on McKinnon and Asiata, but they have only had 13 carries between the two of them. Still, they don’t have great yards after contact and only McKinnon ran beyond the expected yardage for anyone running in his down and distance situations.

It’s pretty clear that the Vikings offensive line failed the running game yet again, and Peterson did happen to look a bit better this week than he did in the previous week, but not “better enough” to overcome the run blocking ahead of him.

The Vikings run game is clearly the issue with the offense right now, which is a very unusual thing to say. Some of that is certainly on the blocking—generally speaking, blocking seems to account for half of the performance of a team’s run game—but the Vikings running game isn’t being advanced by the runners, either.

Generally speaking, Peterson has been able to overcome poor blocking by generating additional yardage through power or speed. He rarely has done that over the past two weeks. The difference between the high success rate in Week 2 and his yards over expectation means that his successes didn’t get too many yards and the unsuccessful plays lost quite a few yards.

With Peterson injured for potentially the next several weeks, the Vikings will ask more of McKinnon and Asiata. While the two of them might unbelievably offer more consistency and upside, they haven’t been crushing it, either.

Coverage

We won’t have specific, double-checked coverage stats handy for each game, but we do know that before Trae Waynes’ interception, he gave up seven receptions and three penalties. The receptions amounted to 98 yards and a touchdown.

Waynes had 11 targets, making him the most targeted defensive back of the past two weeks, 24. Giving up 15 receptions means he’s also given up more receptions this year than anyone else. With 185 yards given up on 24 targets, plus a touchdown and an interception, he’s been much more of a liability than a help. Passers passing to Waynes over the past two weeks earned a passer rating of 82.3 and an adjusted yards per attempt of 6.5 while passing to anyone else on the Vikings earned a passer rating of 73.6 and an AYA of 4.6.

Waynes, targeted on 28 percent of snaps, is an outlier defensively. Only four defensive backs were targeted on a higher percentage of snaps in the NFL in week two, and Waynes has been more targeted than any other Viking for two consecutive weeks.

Positively, Anthony Barr and Eric Kendricks were only targeted on 8.3 percent and 10.8 percent of their snaps in coverage. Munnerlyn was targeted on 20 percent of his snaps last week, but did extremely well in those situations. Terence Newman was targeted 11.6 percent of his snaps, Andrew Sendejo 2.4 percent of his snaps and Harrison Smith wasn’t targeted at all.

Run Defense

Eddie Lacy didn’t get much in the way of yardage, but that had more to do with his down and distance running situation than it does with the Vikings defense. His success rate of 50 percent, identical to Adrian Peterson’s, ranked fourth for the week. Unlike Peterson, Lacy consistently overperformed vs. average running backs in the same situations—with a yards over expectation of 1.82 per attempt and ranked sixth among running backs.

But which Vikings did the best in run defense specifically? These statistics don’t account for defenders out of position or anything, so it’s not a complete accounting of run defense, but it’s good to see which defenders had their tackles matter and which cleaned up their tackles.

Stop rate looks at which tackles accounted as a loss for the defense, and have been tracked differently by Pro Football Focus and Football Outsiders. Here’s my accounting after adding up the play-by-play:

Player Total Tackles Sacks Stops Stop Rate
Linval Joseph 6 1 4 66.7%
Danielle Hunter 4 1 2 50.0%
Brian Robison 2 1 1 50.0%
Tom Johnson 2 1 1 50.0%
Chad Greenway 2 0 1 50.0%
Anthony Barr 2 0 1 50.0%
Captain Munnerlyn 5 0 2 40.0%
Eric Kendricks 5 0 2 40.0%
Shamar Stephen 3 0 1 33.3%
Harrison Smith 7 0 1 14.3%
Andrew Sendejo 11 0 1 9.1%
Trae Waynes 4 0 0 0.0%
Everson Griffen 2 1 0 0.0%
Terence Newman 2 0 0 0.0%
TEAM 57 5 17 29.8%

Unlike last week, these should not include any special teams tackles. One of the reasons the stop rate is high despite Eddie Lacy’s high success rate is because James Starks didn’t record one successful run against the Vikings.

Pass Rush

The Vikings pressure rate wasn’t as gaudy as their sack total. With five sacks, the Vikings certainly got home. But they only created pressure on 25 percent of snaps, a league low. Situational rushers Tom Johnson and Danielle Hunter created the most pressure (as is typical) in passing situations, getting into the pocket on 16.7 and 14.3 percent of their pass rushes, respectively. 16.7 is a strong rate for a 4-3 defensive end, much less a defensive tackle.

Brian Robison’s 9.4 rate is respectable, but Everson Griffen’s 5.1 is lacking.

And of course, Linval Joseph’s 7.7 percent pressure rate is astounding for a nose tackle.

Joseph’s four stops, sack and fantastic pressure rate makes for a phenomenal game. While Diggs may have had a near-record setting game as a Viking, we shouldn’t forget how well Joseph played.

The Vikings should be proud of their win, but there are a lot of holes to be revealed in their game that appeared before Peterson went down. If they want to sustain their success, they’ll have to shore up the run game on both sides of the ball and figure out a solution to Trae Waynes (or wait for Xavier Rhodes to come back).

  • Politicalpuck

    Very good stuff. Going to start checking back here every week.

  • sandbun

    “There’s probably reason to be concerned about the secondary receiver position, but once Treadwell cracks the starting roster, the problem may be resolved.”

    Could this not be also be fixed by elevating Thielen to the #2 spot instead of Johnson? Granted, if AP is out both are going to see a lot of time with the Vikes probably going to more 3 WR sets, but sure seems like Thielen is making more of his opportunities than CJ, and the stats above bear that out.

    • Brian_pdx

      Thielen should stay playing and Johnson sit if Treadwell comes up and this would be a good game to give that a try.

  • Brian_pdx

    My head is spinning. MY MBA program was a bit less complicated at times but it is interesting. The OL is a train wreck and how many times has a unit been improved from least or nearly least effective OL after two games to even league average in the NFL?





vikings

Previous Story Back Page: Deadspin needles Vikings with commemorative Love Boat brick Next Story Zulgad: Mike Zimmer’s steadying hand is instrumental to Vikings’ success