INDIANAPOLIS — Sometimes jokes hit too close to home.
At the podium inside the Indiana Convention Center at the NFL Combine, Minnesota Vikings head coach Mike Zimmer was asked about using analytics and he elected to answer with a crack.
““Yeah, [GM Rick Spielman] loves it. Analytics is a tool. He likes to use it. It’s a good buzzword for him,” Zimmer joked. “For me, tendencies about the other team and things like that, we look at tendencies, we could say they’re analytics. But I have a hard time with somebody telling me to go for it on fourth-and-5, and we’re up by two scores, and they don’t know the team that they’re playing against. If you do go for it, and you don’t get it, they don’t get fired, I do. So that’s my take on it.”
The back-and-forth that Zimmer referenced was apparently against the New York Jets in 2018. The exact situation he described does not match up to one in that game but the analytics department may have seen reason for a different decision in the third quarter when the Vikings were up 10 and kicked a field goal from the opponent’s 4-yard line to increase the lead to 13 midway through the third quarter.
They might have been suggesting that a lead of 17 vs. 13 would be worth the risk of failing on fourth down. Using Pro-Football Reference’s “Win Probability Calculator,” the Vikings would have had a 98% chance to win had they scored a touchdown. With a 13-point lead they were still overwhelmingly favored to win at around 94%.
Ultimately it was a small difference and the Vikings won easily that day. No harm, no foul, right?
Multiple sources in Indy said that Zimmer’s pushback against some uses of analytics is a point of contention within the organization. At the podium Zimmer more or less laid it out: The front office is among the most progressive in the NFL and he sees the role of numbers as less useful. One person with knowledge of the Vikings’ analytics usage said that they are in the upper tier of teams using data to evaluate players, especially at the NFL Draft. Spielman has gone into detail in the past about how the team uses Combine scores to “clone” players. One example would be the number of late-round defensive ends who are tall with long arms and strong three-cone drill scores that the Vikings have picked over the last number of years.
“We go all the way back to 2005 now, and what we graded them as college, what they ended up being with their pro grades, how they ended up performing and you take those two and then you start, that’s when the analytics come spinning back in,” Spielman said in 2018. “And we’re to the point now where you can say… ‘out of the 2,500 traits that we looked at from effective commitment in a psychological score to his arm length to a short shuttle, those were the three most important things at the offensive center,’ OK, they spin it through their algorithms and bam, ‘This guy is similar to that guy that came out in red,’ or bam, ‘This guy is similar to that guy that failed.'”
But when it comes to suggestions on the coaching side, the analytics department’s advice is falling on deaf ears more often, sources said.
As it pertains to fourth down calls, there seems to be an ongoing battle between the numbers and coaching philosophy. Zimmer’s fourth down decisions have also largely been good by the data. Even a call to go for it against Washington last year (that Zimmer called his “worst decision”) in which the Vikings failed on fourth-and-1 was a statistically defensible call. He’s been one of the more aggressive coaches on fourth down, going 8-for-15 last year, ranking in the top third of the league in attempts. The NFL’s most aggressive playoff teams — Philadelphia and Baltimore — each went for it on 24 fourth downs.
But the difference in a handful of game management decisions can swing a game or season. For example, the Vikings punted in fourth-and-9 at the Kansas City 45-yard line at Arrowhead Stadium in Week 9. The numbers likely pointed to going for it but instead the Chiefs quickly turned it into a 91-yard touchdown run from Damien Williams and ultimately beat the Vikings — a win that helped KC earn home field advantage.
Zimmer said last year he’s put forth great effort in training camp to improve his own situational performance. Still there’s concern that teams using the data rather than instinct might get it right more often.
Aside from in-game decisions, another debatable point is the value of the running game in a league that is turning more toward passing.
Zimmer has clarified repeatedly that he wants a strong running game to set up play-action passes, which were highly successful last year using Gary Kubiak’s long-proven system. Under John DeFilippo in ’18 the Vikings were 19th in play-action percentage and last year they jumped to sixth and Kirk Cousins ranked fourth in play-action quarterback rating (per PFF).
It has been proven by data analysts that an outstanding run game isn’t required to succeed with play-action passes but it certainly doesn’t hurt. Dalvin Cook became the center of defenses’s attacks against the Vikings last year and it doesn’t take a deep All-22 study to see it in action. And by PFR’s “expected points added” the Vikings gained an estimated 22 points by improvement on the ground from ’18 to ’19.
But some viewed the commitment to the run game as leaning too heavily on plays proven to be inefficient. The Vikings passed on 51% of plays. Only three teams passed at a lower rate and two of them — Baltimore and San Francisco — had more plays when leading.
Running, opponent tendencies, draft comparisons and fourth downs are hardly the extent of analytics in today’s NFL.
The league is in the midst of a revolution with numbers. PFF and player tracking with NFL NextGen are creating new analysis tools for teams behind the scenes that some teams are using a great deal with their decision making process. Player health data (which is being used by the Vikings too) is also another frontier that has not fully been explored yet.
If the NFL follows baseball’s footsteps, analytics departments are likely to grow and the Vikings — as they are in many other ways — want to be a modern franchise at the forefront of these developments. The hiring of COO Andrew Miller, one person said, is likely to continue the push toward modernity because of his background in Major League Baseball.
So even if Zimmer’s decisions are often matching up with the numbers, the question is whether he will come along to the new frontier. Because that is yet to be seen, some have openly wondered how things might be different had the Vikings kept Kevin Stefanski, who was hired as the head coach of the Cleveland Browns.
Whether the conflict in philosophy ends up playing into a contract extension for Zimmer is also yet to be seen. Opinions in Indy were widely varying, with some believing that the team’s win against the New Orleans Saints gave ownership every reason to believe in Zimmer and others thinking they might be hesitant.
No doubt his record is impressive. Zimmer has made the playoffs three of the past five years with different quarterbacks each time. His defenses have consistently ranked in the top five and the organization has taken huge steps under his watch from a team struggling for relevancy to a top-tier franchise with Super Bowl expectations. There is no debate over whether he’s a good NFL head coach, it’s whether they will fall behind if the numbers aren’t embraced.