This article is an installment of a series in which I use my Observational Rushing Numbers (ORNs) to shed light on just how good the 2017 rookie running backs were at carrying the ball. You can find each previous article about these numbers on the series’s hub.
For different reasons, plenty of 2017’s top-drafted rookie backs struggled. Christian McCaffrey took a half-season to adjust his vision to NFL speed. Perhaps due to fatigue from a huge workload, Leonard Fournette wasn’t any more than average. Joe Mixon was simply kind of disappointing. Dalvin Cook, though, was just fine.
A couple weeks ago, I proclaimed Kareem Hunt as the best running back (nuance that was shed when the article had its title changed) of the 2017 rookies. I probably should’ve added even more nuance to that take, because Cook has a case too.
(I’ve taken these tracking stats from Weeks 1-14 for each rookie running back with at least 75 carries. If you’re wondering what any of the stats mean in a fuller sense, check the series glossary.)
At a glance
Four Big Stats and Class Radar
It’s a doggone shame that Cook went down so early in the season. He put up the class’s second-best stats through four games, and his supplementary rates suggested that his production was more sustainable than first-place Hunt’s. But let’s step away from the Hunt comparison.
The two graphics above show that Cook was awesome in short yardage (what SGY/C measures) and possessed reliable vision and agility (what Success Rate measures), but was disappointingly average in the open field (what HGY/C measures).
He still finished above-average in that final category, and was one of just backs (Hunt and Alvin Kamara were the other two) to achieve that mark of consistency. These stats could change if we got to see 200 or so more carries out of him, but we didn’t. So, I’ll be working to balance these stats with the limited sample size they occurred in.
Cook’s vision (measured by his Identification Rate) was nearly the best of the class, and against all odds, he overcame poor Combine performances in the shuttle and 3-cone drills to attain above-average levels of quickness and agility (measured by Hit Rate):
It turns out that Dalvin Cook‘s agility is, in fact, not bad. Huh. (The next dumb Combine idea floating around? Calvin Ridley, who ran a 4.43 40-yard dash, above many other things, is not athletic. This is a scientific fact.)
Logically, with two solid sub-stat measures, Cook was one of the class’s best at hitting holes overall.
Hit Generated Yards
Seventh place in HGY/C is a disappointing finish for a guy like Cook, who was so tantalizing in the open field at Florida State. For whatever reason, he just didn’t work himself into the open field very much, in a limited sample size. Looking at his HGY/C plotted against his Gash Rate (how often he generated at least 10 yards on carries in which he’s hit a hole) shows the issue:
Cook actually produced much more in space than we’d expect, given how often he generated 10-plus yards. That’s because he was, unsurprisingly, one of the class’s best tackle dodgers with a 0.39 Broken Tackles per Hit (self-explanatory) mark, good for fourth in the class.
Perhaps, with more carries in 2018, Cook can increase that Gash Rate and thus his HGY/C, since he was just so promising in college. Regardless, 5.21 Hit Generated Yards per Carry is a good mark for any back (above-average for the entire pool of veterans and rookies I’ve sampled thus far). So, there’s still no cause for actual concern here.
Stuff Generated Yards
After watching Christian McCaffrey finish second in SGY/C, I’m not surprised to see Cook take first place. While he isn’t known for his strength, the Viking’s built well and has mastered the little things (namely, low pad level, getting downfield quickly, and driving feet at contact). Most of all, he shines on stuffed perimeter carries, where he really gets to leverage his special athleticism.
Despite the small sample size, I’d feel confident in him keeping up this limited-space excellence, too; his Chunk Rate (how often he generated at least three yards on runs where there was no hole to run through) indicates that, if anything, his SGY/C should’ve been higher:
Of course, both of those measures could fall, but even if they do, he’s still likely above average. Also, he’s just as likely to rise in open-field play (via HGY/C) as he is to fall in limited-space play (via SGY/C).
Three Big Stats
(Note: I pooled the tracking stats for Jerick McKinnon together with those of Cook in order to get a more complete, 212-carry sample size for the unit.)
Intuitively, you’d expect a team to be about as good at generating baseline yardage (what OL-GY/C measures) as it is at creating opportunities for the back to gain even more (what Hole Rate measures). But that isn’t really the case in practice — r-squared coefficient between OLGY/C and Hole Rate is under 40%, and thus, the relationship is pretty weak:
The Vikings, for example were much better at creating holes than they were at moving the line of scrimmage forward, due to a leaky offensive line. We can deduce that permeability from zooming further into the stats. The proportion of carries in which the offensive line generated at most (x) yards show that Minnesota allowed much more penetration than the typical offensive line:
To put those numbers into words, defenders had a chance to tackle the Vikings RBs two yards (or further) behind the LOS on more than 13 percent of carries, and almost five percent more often than average. Cook and Jerick McKinnon had to beat tacklers before they got started far more often than most backs. The good news is that when there wasn’t that sort of penetration, everyone did their job pretty well — hence the strong Hole Rate.
When a line gives up a significant amount more penetration than the average offensive line, I suspect that its weakest link(s) are particularly weak, as I explained in my Marlon Mack review. Again, I will leave up the matter of identifying those links to a bigger Vikings fan than myself, though without much of a hunch this time — they enjoyed relatively little turnover and lost just one Week 1 starter (for just six games, at that) in 2017.
And though I’ve begun this section by fixating on the unit’s one flaw, this line was perfectly fine in aggregate. You can also make a strong case for them improving in 2018. Of the nine contributors who saw snaps in the regular season, just two depart: Joe Berger — whom I hadn’t of until five minutes ago, but who has apparently played in the NFL for 12 seasons and is now 36 — and Jeremiah Sirles — who is playing on his third team in four seasons and didn’t play in 2015. Berger has logged 46 starts in the team’s last 48 regular season games, but it’s hard to see him being a huge loss.
Advanced analytics show that when two players leave a group of nine, there are seven players remaining, which means that the Vikings have both continuity and experience on their side. Furthermore, second-rounder Pat Elflein and fifth-rounder Danny Isidora won’t be rookies anymore. And to top it off, Minnesota added Piesman-winning dynamo Brian O’Neill in the second round of the 2018 draft. There isn’t a ton of upside, necessarily, but this is still one of the more promising run blocking units I’ve reviewed thus far.
The Vikings allowed a surprisingly high level of penetration into the backfield in 2017, but that didn’t stop Dalvin from Cooking (pun not originally intended, but now intended). Otherwise, the run blocking collective performed pretty well, and basically all of the arrows are pointing up for Minnesota’s run blocking in 2018 and beyond.
Meanwhile, Dalvin Cook’s season-ending torn ACL was a dadgum disappointment, both because it robbed us of the joy in watching one of the league’s most fun backs and because it robbed me of a greater sample size to evaluate him on. The early signs, given that small sample size? Predictably, he’s one of the class’s absolute best rushers, and if I expect anyone to unseat Kareem Hunt as The Best Class Of 2017 Ballcarrier™, it’s Cook.