Dalvin Cook planning on holding out

While Over The Cap founder Jason Fitzgerald said that Dalvin Cook could command around $13-14 million a year, he thinks the Vikings would be crazy to offer more than $10 million for their starting running back.

“They’d be crazy to go above that. They’d be insane,” Fitzgerald said. “I don’t think he has any legs to stand on here. I don’t think it should turn nasty. I think people just have to be realistic with this. It’s the unfortunate case of a running back.”

Fitzgerald also mentioned Melvin Gordon in relation to Cook’s situation.

After holding out on the Chargers and not playing until Week 5, Melvin Gordon received $8 million this offseason with the Broncos.

I’ll say this about Gordon: He was not coming off a year as good as Cook’s in 2019. Cook ran for 1,135 yards and 13 touchdowns in 2019, while Gordon was coming off a 2018 where he had 885 yards and 10 touchdowns before holding out.

“I think they’ll make him a legitimate offer, just like the Chargers made Gordon a legitimate offer, reportedly,” Fitzgerald said. “But you just can’t sink that kind of money into that position.”

I’ll start out by acknowledging just how good Cook is at running the football. He’s undoubtedly a top-five running back right now, and that’s easy to see any time you pop on his tape. The 24-year-old is particularly good at executing the zone running plays that make up the bulk of Minnesota’s run game, which the entire system is seemingly based around.

Here comes the but: Cook was far and away the Vikings’ best runner in 2019 based on expected points added, but his running plays weren’t overly efficient. The Pro Bowl back averaged 0.03 EPA per attempt; compare that to the 0.16 Kirk Cousins averaged per play in 2019, according to Pro Football Focus. For the millionth time: Runs just aren’t as efficient as passes.

Of course, Cousins’ numbers were inflated by Minnesota’s high usage of play-action, and old school football types will tell you without a productive run game, those play-action fakes aren’t as effective. That’s when the nerds will tell you there is actually no evidence suggesting that’s the case. And I guess I’ll include myself in that group. I came to the same conclusion when I ran the numbers myself, finding running performance has little-to-no effect on a team’s play-action passing game.

The Vikings aren’t some exception to rule, either. I charted all of Cousins’ play-action dropbacks in 2019, taking note of which running back was on the field. Though Minnesota did produce a higher EPA on the plays that included Cook (0.46 to 0.32 per dropback), the success rate was actually higher for the plays he was on the sideline (53.5% to 50%). Given all that we know about play-action passes, the difference in both categories is probably just noise. Vikings offensive coordinator Gary Kubiak has been producing dominant play-action pass games for two decades now, and he’s done so regardless of the talent he has at running back.

Contrary to popular belief, a running back’s presence on the field doesn’t seem to have any effect on a defense or how it lines up. You know the theory that having a strong running game forces a team to load the box? Yeah, that’s pretty much nonsense, as I covered in an article looking at Ezekiel Elliott’s value to the Cowboys offense.

A team’s willingness to drop an extra player into the box has more to do with the structure of an offense more than anything else. If the offensive formation creates an extra gap by adding a tight end or full back to the equation, the defense is going to have to add a defender in the box to account for it, regardless of the running back’s reputation.

The offensive formation largely dictates how a defense will line up. That was certainly true for the Vikings in 2019. It didn’t matter which player was in the backfield, the number of defenders in the box was pretty much the same for each of them.



Maybe Cook didn’t have an effect on the passing game in that regard, but what about his contribution as a receiver? He caught 53 passes for 591 yards, both numbers were near the top of the league for running backs. But unlike the Christian McCafferys and Alvin Kamaras of the NFL, Cook is not a back who threatens a defense downfield. That’s no exaggeration, either. His deepest target last year? It traveled a whopping FOUR yards through the air. In fact, his average depth of target was -1.39 yards.

Cooks isn’t a player you line up out wide to create a mismatch, either. He was targeted exactly once when lined up outside of the backfield. The play went for three yards.

So where did all of those receiving yards come from? Well, Cook was dominant in the screen game, leading the league in total EPA (18.7) on 24 plays, per Sports Info Solutions. But we probably shouldn’t expect that to continue. Over his first two seasons in the league, Cook was not very efficient in the screen game, compiling -1.5 EPA on 20 targets. It turns out that performance on screen passes is extremely volatile year-to-year, which makes sense given the sample sizes we have to work with. The correlation between a player’s average EPA on screen passes in 2018 and 2019 was non-existent when I ran the numbers…



Cook’s performance on screen plays in 2019 tells us nothing about what he’ll do in 2020. If his receiving numbers are your argument for paying him, you’ll have to find another argument.

Cook is a brilliant runner and not much more. But there are a lot of talented running backs out there and they will come at a more reasonable price. And chances are, those other running backs haven’t missed 40% of their games due to injury as Cook has through the first three seasons of his NFL career.

(Minnesota already has one of those talented running backs on the roster. While Alexander Mattison’s efficiency numbers don’t touch Cook’s, he did outperform his more heralded teammate in Sports Info Solutions’ Total Points metric on a per-snap basis. Total Points appears to be more predictive than EPA.)

As the Vikings front office sure knows, thanks to the new CBA, Cook’s holdout threat is completely hollow. As Brad Spielberger of OverTheCap.com points out, a player who fails to report to training camp now loses an accrued season. So if Cook showed up after camp, he’d only receive a credited season, meaning he’d be a restricted free agent next offseason. The Vikings could slap a first-round tender on him, which would automatically give Cook a 2021 salary of about $3.5 million, and force another team to give up a first-round pick in order to sign him. That isn’t going to happen.

Minnesota has all the leverage and, based on all of the evidence suggesting running backs don’t matter, it should have little motivation to cough up the money Cook is looking for. But the same has been true in just about every situation in which a team has decided to pay its star running back (hello, Carolina Panthers). I’d like to think NFL teams have learned by now; but, if that were actually the case, I wouldn’t have had to write this same article again.