Back to the Drawing Board: Using EPA to Fix the Hokies Offense

Well, at least that's what it looks like right now after the Hokies grinded their way to a 3-1 (1-0) record heading into the bye week. After an offseason of hope and promise for the offense, many Hokie fans have been disappointed to see lackluster performances against weaker opponents coupled with inconsistent execution in the most crucial of scenarios. Still, it's hard to fault a team that sits in the driver's seat in their division and whose only loss is by a possession on the road, against a rival, against one of the best defenses in college football. And maybe that's all we should've expected in the first place: just enough to win (most of them).

During this bye week, I want to share with you some findings on how the Tech offense has performed using the top analytical measures in the game, compare that nationally, and see if we can project what will happen for the rest of the season. Spoiler Alert: I have no idea.

I told fellow Sons of Saturday contributor Robert Irby, who wrote a phenomenal piece on Tech's QB development, that this was going to be the sassy sister article to his. That's mostly true. The majority of this article is cold hard facts, but this is the sassy part:

If you think that Justin Fuente, who is hanging onto his current job by the tiniest of threads hanging from Whit Babcock's tailored maroon blazer, cares about anything but wins this season you are vastly mistaken. His job at Virginia Tech is to win football games at a high rate. He hadn't been doing that. His offenses from 2016-2020 had set school records, been top tier nationally, and even sent guys to the NFL. None of them resulted in the consistency and relative dominance that Virginia Tech demands from it's football team.

At Memphis, Justin Fuente and his staff had two jobs.

• Win a fair amount of games and compete for a group of 5 conference title
• Score a billion points

Schools like Memphis have to do that because no one cares about Memphis unless they're scoring 60 a game. Do you really want to watch a grudge match between Memphis and UCF that finishes 17-10? Of course not. Now what if both of those offenses are scoring 50+? Now you have a game that people will pay attention to. But that's not the job description at Virginia Tech. What is?

• Win at a high level
• Compete at a national level in the post season

Virginia Tech's entire football history is based on terrifyingly good defense, elite special teams, and discipline. That's the way to win at Virginia Tech. The issue with the current Tech offense is that it doesn't seem good enough to third-wheel the power couple of defense and special teams. Combine that with the fact that Hokie fans have been promised a more electric offense for years now and it hasn't been delivered, and it's not hard to understand why so many in Orange and Maroon are calling for a regime change.

But that regime change is out of the picture for at least a few more months. Until then, let's look at what the Hokie offense has been doing well, what they've done poorly, and what they need to do capitalize on a great opportunity to compete for a conference title.

Fellow contributor Shelton Moss wrote a great 'Inside the Numbers' article Tech's offensive struggles so far this season. And it's clear, things need to change. The question of what and how things need to change is the stuff that OC Brad Cornelson gets paid to answer.

Firstly, you need to take what you do well and try to do it more often. Sounds simple, right? Well, not exactly. When players don't execute simple things like hitting wide open wide receivers, blocking the guy directly in front of them, or simple miscommunications, it's hard to tell whether or not your team has the ability to pull off that specific play. Sometimes, guys just mess it up. For instance, someone watching Tech's offense through four weeks would say the team is not very good at passing the ball. Contrary to popular belief, that's not the case.

A measure I'm going to refer to for the rest of the article is one of the leading analytical tools in football. It's called "Expected Points Added" or simply "EPA". What EPA measures is pretty self explanatory, actually. EPA is a measurement of how many points you can expect to score after each play as compared to what an "average" team would score. The average EPA would then be a 0.00, and teams can either have positive or negative EPA's (just so we're clear, positive is good.) The opposite is true for defenses. A negative EPA is good for a defense because that means that the offense is scoring less than an average team would on a per play basis.

EPA can be broken down into passing and rushing plays. This is useful in determining whether a team is better or worse at passing or rushing the football. Obviously, a team would want to have as higher scores for offensive EPA, and as lower scores for defensive EPA. Let's take a look at how the Hokies stack up nationally. A few data filters:

• Data only consists of teams with quarterbacks that have logged >100 drop backs this season in order to eliminate teams with QB changes/injuries, triple option teams, and account for smaller data samples for teams that are overly reliant on running (looking at you, Michigan)
• Data only consists of FBS teams

I divided this graph into quadrants. Here's what the quadrants tell us:

Top Left: Good at rushing, bad at passing

Top Right: Good at rushing & passing

Bottom Right: Good at passing, bad at rushing

Bottom Left: Bad at rushing, bad at passing

EPA tells us that the Hokies are a good passing team (0.248 EPA/pass), but a below average rushing team (-0.102 EPA/rush). That's interesting, because almost every Hokie fan would think the opposite of the program's offensive identity over the past few years. Last season, the Hokies were one of the best rushing teams in the nation. This year? The Hokies are actually losing expected points every time they run the ball. That's not good.

This makes us ask a new question: why does the passing game look so bad then? Well, there's two things. Firstly, it's much easier to notice a bad passing play than it is a bad rushing play. Think about it, every time you see an quarterback miss an open receiver, the crowd does a collective sigh of frustration. Contrarily, when the running back goes up the middle for a two yard gain, most fans don't blink an eye.

Secondly, the Hokies simply don't throw the ball enough. Top notch passing offenses need two things: effectiveness and volume. The latter is where the Hokies are falling short. To show this, let's look at all of the offenses from our previous visualization that have positive EPA/pass measurements and chart them by the percent of plays they run that are passing plays.

Similar to above, this graph is divided into quadrants. Here's what they tell us:

Top Left: Decent at passing & throws the ball a lot (should throw less)

Top Right: Really good at passing & throws the ball a lot (should keep throwing)

Bottom Right: Really good at passing, but doesn't throw the ball a lot (needs to throw a lot more)

Bottom Left: Decent at passing, but doesn't throw the ball a lot (needs to throw a bit more)

Virginia Tech falls into the quadrant of teams that need to throw the ball a lot more. In fact, the Hokies are only throwing on about 41% of plays. That's a very low amount for a spread offense. If you combine that with the fact that the team is not good at running the ball (again, -0.102 EPA/rush) then you can see how the offense looks absolutely abysmal.

I've done a lot of research into how QB runs can impact the game of football, and I'd highly recommend looking into that because the information in that article is very pertinent to the current situation with Tech's offense. Currently, the leading rusher for the Hokies is QB Braxton Burmeister. He's averaging a solid 5.2 yards per attempt (which includes sack yardage) and has a PFF rushing grade of 71.5, which is the 15th highest grade for a QB in the country with a minimum of 100 drop backs. And anyone who watches the Hokies play can easily see one thing: Braxton Burmeister is the best athlete on the Virginia Tech offense.

If we take a look at all of the quarterbacks in our data set and their PFF rushing grades, we can maybe find the missing ingredient to the Hokies offense.

Let's explain these quadrants like the last couple graphs:

Top Left: Bad rushing team, but good rushing QB (should run QB more)

Top Right: Good rushing team & good rushing QB (keep rushing QB)

Bottom Right: Good rushing team, but bad rushing QB (rush QB less)

Bottom Left: Bad rushing team, bad rushing QB (LOL stop running the ball)

Coming into the season, we knew Burmeister was going to have a heavier weight of the rushing load. That's proving to be true in a production standpoint, but after four games the athletic QB only has 18 designed rush attempts that have gone for almost 200 yards. In terms of both effectiveness and efficiency, he's the guy in the Hokies backfield. Getting him more caries is really the only hope for the Tech running game.

It's also imperative that the Hokies throw the ball more often. As you can see from the below tweet, when visualized the margin between Tech's passing and rushing EPA's is astonishing. Simply put, the Hokies are doing more of what they're bad at and less of what they're good at.

What's the reasoning behind this? Simply put: depth. The Hokies can't afford an injury to Braxton Burmeister. If he's injured, then Virginia Tech will be a 0.500 team, the chance of an ACC Coastal title will be gone, and Fuente will most likely be hitting the job market. No one in Blacksburg wants that to happen. Virginia Tech is in this mess because the offensive staff has been targeting running quarterbacks in both the transfer portal and its recruiting classes. The team's offensive success is tied more to a running QB then almost anyone in the country. (Again, you can article on this for a deeper dive.)

So, running Braxton Burmeister at a larger clip is not entirely out of the question, but it's not a fool proof solution either. The other solution is throwing the ball at a much higher rate. Regardless of how good Burmeister is at throwing the football, he has a knack for not throwing interceptions and the wide receivers have proven that for all their flaws, they can go up and get contested balls as well as be electric in the screen game.

I would like to reiterate, the Hokies don't need to score 35+ a game this season to be successful. So far, this is one of the best scoring defenses the Hokies have had in a while. Combine that with top notch special teams and winning the "hidden yards" battle, and the Hokies have the old-fashioned formula for success.

But then there's the hard part of this whole equation. You can say "throw the ball more", but that means that someone has to throw an accurate pass and another player has to catch the ball. Sounds simple, but clearly it's not for the Hokies. Don't even get me started on things like blocking. Truthfully, I'm not sure this staff has faith that the offense can be successful throwing the ball 40 times a game. Why? I have no idea. To their credit, however, in the few times the Hokies have thrown the ball there's a real reason to doubt. Still, with all of the missed opportunities the Hokies have still managed to score more points than that other team. Outside of a few boneheaded play calls and some bad luck with a couple official reviews, the Hokies have been better on offense than their opponents. Obviously, there's more work to be done than any of us imagined at this point in the season. Will the coaching staff be able to work things out? I'm not sure. But the way forward, at least schematically, seems simple enough.

Going forward, the Hokies need to give guys like Tre Turner, Tayvion Robinson, Raheem Blackshear (who is the #2 graded receiving RB in the country by PFF), and even young guys like Dawain Lofton and Jaden Payoute a chance to make plays. When the Hokies do that, you will see a much more productive offense that can, and probably should, compete for an ACC title in 2021.