2021 Offensive Woes

Perhaps no team better embodies “missed opportunities” than the Minnesota Vikings. They haven’t just missed out on a ring in four trips to the Super Bowl, but a shot at a ring in six other trips to the NFC Championship Game. The Vikings, hoping to capitalize on a missed opportunity in 2017, upgraded at quarterback from Case Keenum to Kirk Cousins — and put together an 8-7-1 season.

Since then, they’ve floated between mediocre and good and have earned one trip to the playoffs in the Cousins era. That’s not to say their missed berths are entirely Cousins’ fault, but rather that the Vikings might have correctly recognized their playoff windows haven’t yet capitalized on them.

In much the same way, the Vikings’ play against backup quarterbacks in the past three years has been a study in snatching defeat from the jaws of victory. Backup quarterbacks don’t provide free wins, but to a good team, they should be close enough. Over the past five years, teams have won 70 percent of their games against backup quarterbacks (who have thrown at least 10 passes) when not fielding a backup themselves. Teams that finished the regular season .500 or above have won 80 percent of those games. By contrast, teams that finish below .500 predictably fared worse against backups as they won 54 percent of their matches against them.

But the Vikings have been less than stellar under coach Mike Zimmer, winning just 60 percent of those games in the past three years. They beat David Blough, Nick Foles, Mike Glennon, Brandon Allen and Keenum, split a series with Chase Daniel (they beat him when he played for Detroit and lost to him when he was in Chicago) and lost to Andy Dalton, Matt Moore and now Cooper Rush.

Rush, the most recent backup quarterback the Vikings have faced, is also the most embarrassing loss. Moore and Dalton have significant starting experience, and Daniel had spent nine years — 65 games — on an active roster before he played against the Vikings with Chicago. Prior to his game against the Vikings, Rush had been active for six games and thrown just three passes. While Daniel moved from team to team to escalating salaries, Rush had recently been cut by the Cowboys and the Giants before he returned to Dallas to back up Dak Prescott. Rush, who played in college at Central Michigan, finished fourth in the MAC in adjusted passing yards per attempt behind Logan Woodside, Zach Terrell and Brogan Roback.

For most of the game, the Vikings defense wasn’t bad — it allowed 1.81 points per drive and forced two turnovers. It had its fair share of catastrophic faults, and it should have held Rush to fewer points, but given the Cowboys were giving up 2.01 points per drive heading into the game, that should have been enough for the offense to win.

But instead of scoring 24 points in 12 drives, the Vikings scored 16. They found the end zone just once. On average, the field position the Vikings enjoyed on those 12 drives results in 2.75 touchdowns, meaning the Vikings left nearly two touchdowns on the field. On three of those drives, the Vikings found themselves starting at their own 40-yard line or better — the product of two turnovers and a good punt return — and scored just three points. Even worse, the drives coming from turnovers netted a total of 1 yard.

One of those good field-position drives was part of a series of consecutive three-and-outs, and that best represents the Vikings’ theme throughout the game — missed opportunities.

The Vikings converted one of 13 third downs, a rate of 7.7 percent. The Vikings have had only three worse performances on third down since 1991, the limit of the data on third-down rates. In those three games, they failed to convert once. In games with at least one third-down conversion, their performance against Dallas was the worst. That single third-down conversion occurred on the first drive, a 31-yard sideline shot to Tyler Conklin over former Vikings safety Jayron Kearse. The next 58 minutes, 18 seconds featured failure after failure on third down.

Nine of those third-down attempts were passes, and on none of them did Cousins throw past the sticks.

It’s not impossible to convert on a pass behind the sticks, but there’s a clear pattern, as data from the past five years tells us, called “air yards to sticks.”

For passes with 3 air yards beyond the sticks, the success rate is about 60 percent. At the sticks, it drops to 50 percent, and passes in the 2 yards before the line to convert have a success rate of about 40 percent. Part of that is selection bias — quarterbacks feel more comfortable throwing deep when they have open receivers and can be assured of a reasonable chance of a completion, but that the odds aren’t even across the range tells us that offenses aren’t optimizing. That’s true of one offense in particular.

Of his nine third-down throws after the first drive, Cousins tried one to the sticks, one pass within 2 yards of the sticks and another two passes 3 yards before the line for a new set of downs. The remaining five passes were 5 yards or farther away from the yardage needed to convert.

Outside of the scripted first drive, the Vikings never called a play to convert with the pass in the air, didn’t have receivers open downfield, allowed too much quick pressure or had their quarterback misread the play.

There are times when it’s not fair to judge an offense or a quarterback for throwing short of the sticks. Tight coverage downfield or extra space for yards after the catch provide important context, as does situation: If a team is down two near the end of a game, it only needs to improve the odds of a field goal attempt. Earning 7 yards on third-and-12 can do that, even if it’s a throw 5 yards short of the sticks. Not only that, it’s not meaningful for a team to throw short of the sticks when facing extreme down and distance like third-and-24. It’s smarter to play for field position on the inevitable punt or field goal than it is to risk an incompletion and no field-position advantage.

But, for the most part, a team that consistently throws short of the sticks will fail to convert. A good example is what has happened this season to the Vikings on third down with between 11 and 13 yards to go. At first glance, that seems like a counterexample. The Vikings have one of the highest conversion rates at that distance (35.3 percent) all while throwing well short of the sticks; they have the third-shortest throws of any team with at least five attempts, an average of 5.1 yards short of the first-down marker.

Maybe that means the Vikings are right to throw short and trust their exceptional skill players to turn upfield and create a conversion. But a play-by-play look at the data tells us the opposite — the only five throws they converted in that situation are ones that were at least within 1 yard of the line to convert. The other eight throws — all 4 yards or farther behind the sticks — failed to produce a new set of downs.

That doesn’t mean the Vikings should automatically throw beyond the sticks regardless of circumstances — it’s a bit like the Randy Ratio in that regard — but it’s telling that they didn’t do it once against the Cowboys when they were out of their scripted period. More egregious, part of the plan included screen passes that generally haven’t worked. In fact, if we look at this year’s short throws — within 3 yards of the line of scrimmage — on third-and-7 or longer, they’ve converted only one out of 11 attempts on third-and-long.

On their other 23 attempts on third-and-7 or farther, they’ve converted 56 percent. Even reducing the number of designed short throws should help the Vikings offense convert in those situations, even if a 56 percent clip is unreasonable to expect.

After the game against Dallas, Cousins mentioned that a “combination of factors” contributed to the struggles on third down. He pointed to pressure and coverage contributing to his remarkably low depth of target on those plays. That explanation will always be true, but it obscures the fact Cousins was often not making his reads.

Cousins was pressured on five of those nine plays, a high rate — but only one of the plays had pressure arrive early, within the first 2.5 seconds of the play, and that was a screen pass. The other four pressures arrived late in the snap as Cousins held on to the ball despite having receivers open. There were instances when Cousins didn’t have players downfield, like on third-and-6 with 2:26 left in the third quarter.



But those instances were rare. Perhaps the Cowboys threw the Vikings for a loop by deviating from their preferred coverage — they’re one of the heaviest Cover 1 teams and rarely ran it against the Vikings — but that’s not much of an excuse. That the Cowboys showcased more zone coverage than ever might have stymied the Vikings’ plans, but quarterbacks still need to diagnose coverage as it happens and find the open player.

Throughout the game, Cousins declined shots to the intermediate zone or deep downfield even though the plays called by Klint Kubiak put receivers in a position to win. The Cowboys were able to create pressure, even when rushing just four — Cousins was under duress on a season-high 41 percent of snaps. But he also held on to the ball longer than he had all season, an average of 2.99 seconds. Half of the pressures took longer than 2.5 seconds to appear, a product of the longer-developing plays and Cousins’ unwillingness to throw sooner. While few of the pressures were primarily his fault, he exacerbated the mistakes made in front of him. In doing so, he didn’t even absorb the benefit of a longer-developing play and just chose to dump it off.

On top of that, Dallas didn’t create any tighter coverage than it had in any other game. It entered the week allowing 8.2 yards per passing attempt. The strength of its defense doesn’t come from tight coverage but from exploiting mistakes or ball placement for turnovers. The Vikings might have been too worried about that last point, as they didn’t try to throw through tight windows virtually at all.

Next Gen Stats has created an “aggressiveness” index that it defines as a metric that “tracks the amount of passing attempts a quarterback makes that are into tight coverage, where there is a defender within 1 yard or less of the receiver at the time of completion or incompletion.” For Week 8, Cousins ranked last among quarterbacks in percentage of throws through a tight window, with just 2.9 percent of attempts made to receivers while a defender was within 1 yard of the intended target.

If we combine the aggressiveness index with the average-yards-to-sticks measure, we can come up with an index that looks at risk tolerance.

Cousins already ranked pretty low compared to his peers given the Vikings’ quick-strike approach and their talented receiving corps. But for Week 8, the Vikings moved from cautious to meek, placing last among Week 8 quarterbacks with a risk tolerance score of 42.9.

The coverage might not have been what the Vikings expected, but instead of adapting and mixing in risk and reward, Cousins decided to turtle and not challenge the coverage at all.

If the Vikings had better offensive linemen, it’s likely they would have won the game. But the chances were nevertheless available to Cousins to make explosive plays without being exposed to Dallas’ dangerous turnover risk, and they could have put away a team they were favored to beat by at least four points.