FanPost

Numbers show Kareem Hunt, not Alvin Kamara, was 2017's best rookie running back

Aaron Doster-USA TODAY Sports

From the Fan Posts - PS

It goes without saying that Kareem Hunt is one of the best running backs from the 2017 draft class, but let’s recap (just for fun).

He ran for over 1,300 yards and added 455 through the air in the regular season. He torched the Patriots in his first regular-season game out of the MAC. He scored a 35-yard touchdown in one touch against the Broncos. He could have done even more if the offensive line had been healthier. We’ve known he’s special for a long time.


But after watching (nearly) every carry for every rookie running back (with at least 75 carries) in 2017, I’ll go as far as to say Hunt, not Alvin Kamara, was the very best rusher.

My methodology was simple enough but quite demanding. Over the course of every game in Weeks 1-15 for each rookie back, I took down some observational stats. I collected the stats in order to give a strong, quantitative picture of how good a back was, individually, at carrying the ball when removing blocking from the equation (with some math, we can also say a good bit about the offensive line).

The process

On each play, I noted how many defenders were in the box compared to offensive players, whether or not a given offensive line opened a hole for the running back, whether the back found the hole, whether the back hit the hole, how many yards the back generated on his own (i.e. how many more yards he gained than a replacement-level back would in the same circumstances), and how many tackles the back broke (if any).

I did this for 1,800 rookie carries, alone. I’m not crazy; I just have bad priorities.


Then, I did a bunch of math to find which combinations of what stats were meaningful in evaluating rushing (and blocking) ability. By now, I’ve figured out a pretty good way to approximate a back’s yards-per-carry mark purely using derivatives of these raw numbers.

(Note: For the mathematically-inclined, the R-squared value between actual Y/C and the combination of my numbers that project Y/C hovers around 0.95.)

The benefit is that, on a macro level, we can say just how much a given skill set (think: hole-hitting ability, open-field creativity, etc.) plays into rushing performance (i.e. a good yards-per-carry mark), and on an individual level, we can use these numbers to better evaluate a given player’s skill set. The key word in all of this, by the way, is rushing. I don’t take these stats or do the math with the passing game. I’ll leave that to the great work of others.

In time, I’ve added some supplementary stats that give more context to the bigger, overarching ones. I’ll use those later on. Now, here are the primary metrics we’ll be working with:

The stats

Hit-generated yards per carry (HGY/C): Takes the amount of yards a back generates on carries in which he hits a hole, and divides it by the number of holes he hits. Effectively, it measures how well a back creates yardage in the open field. It accounts for 20 percent of the projected Y/C equation.

Stuff-generated yards per carry (SGY/C): Takes the amount of yards a back generates on carries in which the offensive line fails to create a hole (what I call "stuffs"), and divides it by the number of such carries. It measures how well a back creates yardage in tight confines. It also accounts for 20 percent of the projected Y/C equation.

Success Rate: Takes the number of holes a back hits and divides it by the number of carries in which an offensive line opens a hole. It measures how well a back can hit holes and take what his offensive line is giving him. It accounts for 10 percent of the projected Y/C equation.

Offensive Line-generated yards per carry (OL-GY/C): Takes the average amount of yards the offensive line generates on any carry. We find that number by subtracting the amount of yardage the back generates from the overall amount gained on a given play. OL-GY/C effectively measures the amount of push an offensive line gets, as well as how bad its weakest links are. The stat accounts for 25 percent of the projected Y/C equation.

Hole Rate: Divides the number of carries in which an offensive line opens a hole and divides it by the total number of carries. It effectively measures how well the unit works as a whole and provides opportunities to the back. It also accounts for 25 percent of the projected Y/C equation.

The first three stats measure running back quality, while the last two measure blocking quality (note that each grouping accounts for exactly 50 percent of the equation), so we can isolate those two groups to calculate an overall measure of RB/OL quality (called "RB/OL Composite").

Now, we can look at how well offensive lines performed in comparison to one another as well, and the results are pretty interesting.

But I won’t be talking about the Chiefs OL today. We’ll be focused on Hunt instead.

Back to Hunt

With all that out of the way, let’s get into the fun stuff. Here’s how Kareem Hunt finished in each stat relative to the rest of the rookie class:

Stat

Value

Class ranking (of 15)

HGY/C

6.87

2nd

SGY/C

0.81

7th

Success Rate

95.9%

1st

RB Composite

0.48

1st

Evidenced by the first overall ranking at the bottom, these are excellent numbers. First, he finished above average in every category, which only Alvin Kamara -- who wasn’t as far above average -- and Dalvin Cook -- probably helped along by a smaller sample size -- could match.

Furthermore, he was clearly one of the class’s most explosive open-field runners, while his Success Rate was in a tier of its own. Compared to the rest of the class:


In a league that’s been increasingly unfriendly to the running game every year, it’s critical for a back to capitalize on every opportunity the offensive line allows. And Hunt wasted less than half as many opportunities (4.1 percent) as the class’s average back (10.6 percent).


Digging a bit deeper into the numbers, it isn’t hard to see why. A back’s Identification Rate (the number of holes he identifies divided by how many times an offensive line opens a hole) measures his vision and patience, while one’s Hit Rate (the number of holes he hits divided by the number of holes he identifies) measures his agility, quickness, functional strength and balance. Multiplied together, the two stats form Success Rate.

Unsurprisingly, Hunt is one of the class’s best with both stats:


Pretty good, right?

Another positive: While he met the baseline of not wasting opportunities, he certainly made the most of them, too. We saw Hunt break away for countless long touchdown runs in 2017, which certainly translate to the stats, as he produced the second best open-field mark in the class.

When looking at his actual skill set, there may be some reason to pause, however. In this case, the complementary stats aren’t as supportive for Hunt.


Broken Tackles per Hit (BTK/Hit -- the number of tackles a back breaks divided by the number of holes he hits, pretty self-explanatory) and Gash Rate (how often a back generates at least 10 yards on runs where he hits a hole), measure how well a back dodges tacklers and how consistently he really makes stuff happen, respectively.

HGY/C is an average, so it is more susceptible to over-representing a couple big plays (which applies in Hunt’s case), so BTK/Hit and Gash Rate can be used to see how we might expect HGY/C to look in the future. To illustrate, here’s HGY/C plotted against each stat:


BTK/Hit

Gash Rate

There’s a clear strong (but not perfect) relationship in both graphs. Looking at Hunt’s place in relation with the trendline in both graphs, we would expect him to have created a bit less, given how often he broke tackles (sixth in the class) and how often he gashed defenses (fourth in the class). Of course, even if he fell back to those rates, he’d still be well above-average and perhaps one of the best. Also, he could just get better and make this a moot point.


However, where he really has the potential to get better is in short-yardage situations. Of any skill set, Hunt would gain the most by improving his ability to contribute yardage when left to his own devices. With a SGY/C measure that isn’t among the class’s best, there’s clearly still some room for him to grow.

Just looking at him and how he plays, it’s not difficult to see Hunt being one of the league’s better short-yardage backs. Even after losing some extra bulk at Toledo, he’s got a sturdy frame. Beyond that, he runs as hard as ever and decisively makes his way upfield when nothing develops (two big parts of being a net-positive without space). The flashes are there, but the consistency isn’t.

He generated a "chunk" of yardage, on his own, at a near-average rate, illustrated by graphing SGY/C in relation to Chunk Rate (how often a back generates at least three yards on "stuffs"):

Consequently, he was about average overall on stuffed runs and produced basically exactly as much as you’d expect.

The good news is that it shouldn’t take a huge improvement, skill-wise, to make a leap here. He could raise his Chunk Rate, and thus his overall SGY/C, simply by doing more on these sorts of runs.

It’s boring and quite simple, but it makes the difference over time. In fact, it makes a pretty big difference. Using some math in the background, if he could generate yardage on stuffs as well as he did in space, he’d gain roughly 0.35 more yards per carry overall -- taking his already-awesome 4.9 Y/C to 5.2 or 5.3 Y/C. That’d be the second-best mark for any running back in the NFL.

Still, overall, the collection of observational numbers I took for Hunt are overwhelmingly positive. There are three broad categories of a back’s skills that are important to contributing to the running game: How well a back generates yardage when left to his own devices, how well he hits holes when they’re available and how well he makes stuff happen when he’s hit a hole. In the first category, Hunt is just fine. And in the last couple, he’s terrific. Those finishes were good enough to put Hunt first overall in his class. Even if he falls back to Earth a bit (after having so many breakaway runs to start 2017) in 2018, he should still be one of the best backs from the 2017 running back class.

Looking at the entire picture, we should be more than happy with where he is now. He gained nearly five yards per carry over 272 attempts as a rookie and behind quite the inconsistent offensive line (a story for another day).

The numbers I took while tracking his performance further speak to those rushing abilities. And this all comes before we talk about what he adds in the passing game. We've got a good one on our hands.

This is a FanPost and does not necessarily reflect the views of Arrowhead Pride's writers or editors. It does reflect the views of this particular fan though, which is as important as the views of Arrowhead Pride writers or editors.