Georgia Tech went up to Blacksburg and continued its streak of beating Virginia Tech in its home stadium with a backup quarterback at the helm. The Yellow Jackets haven’t lost in Blacksburg since 2012, and in that time, they’ve started at quarterback:
- 2014: Justin Thomas
- 2016: Matthew Jordan (in relief of Justin Thomas)
- 2018: Tobias Oliver (for Taquon Marshall)
- 2020: Game cancelled
- 2022: Zach Pyron (in relief of both Jeff Sims and Zach Gibson)
It’s very interesting (to say the absolute least) that a program that’s dabbled with national relevance in recent memory has had a “Georgia Tech problem” at home in the last decade, especially with the most recent performances coming against shorthanded GT teams. One can only hope that more teams with national relevance develop the same problem.
But enough wishcasting — onto the stats:
As usual, the charts first:
Note: GameOnPaper.com did not have a valid box score for this game due to underlying issues with ESPN’s feed.
Here are three major themes from the weekend we should keep an eye on:
1. Offensive Line Play
To set the stage, here’s what I said about Tech’s offensive line play last week:
Tech’s offensive line has struggled in pass protection in the past two games — per [Sports Info Solutions (SIS)], Tech has allowed 12 sacks and pressure rates of 50.0% (vs UVA) and 36.1% (@ FSU), and what’s more concerning is that these numbers are coming off low blitz rates: FSU only brought an extra rusher on 22.2% of snaps, while UVA blitzed on 14.0% of snaps.
How is Tech’s OL doing on the ground? Well, things aren’t much better: per SIS, Tech rushers averaged 0.2 yards before contact versus UVA and posted an even lower 0.0 mark at FSU. The Cavaliers stuffed Tech on 27.0% of rushes last week, while the Seminoles did the same on 36.7% of rushes on Saturday.
How did Tech do in these metrics this week?
- Sacks Allowed: 3
- Pressure Rate Faced: 51.2%
- Blitz Rate Faced: 41.5%
- Yards/Rush Before Contact: 2.1
- Stuff Rate: 33.3%
A bit of a mixed bag, all considered: Tech did a considerably better job blocking for its rushers upfront (judging by yards before contact), but it still suffered a concerning stuff rate. Tech allowed fewer sacks this week (down from eight and four in the last two), but it still struggles to maintain the pocket for its quarterback, even if the defense isn’t bringing an extra rusher (though VT certainly turned that latter dial up compared to Tech’s past two opponents).
It’s unclear how this gets resolved, especially with three games remaining on the schedule and for this staff as a whole. It’s possible the key is somewhere in offensive coordinator Chip Long’s bag of formational tricks: relying more on jumbo/multi-TE packages might help pass protection, but it would come at the cost of route tree diversity and the amount of space fewer receivers can create within the defense. Unfortunately, we don’t have the metrics available to us (at least at this point) to determine whether moving to those packages has generated better-than-average (or even better-than-current) performances for Tech this season.
2. Depth of Target
Our discussion on offensive line play dovetails nicely into some trends we’re now seeing with Tech’s passing game with Zach Pyron under center. As noted last week:
[P]er Sports Info Solutions (via Bill Connelly), Pyron’s air yards per attempt versus FSU sat at 8.1, a tick below Gibson’s mark (9.9, with Sims at 9.8) last week.
This week, Pyron’s air yards per attempt (also known as average depth of target) dropped again, now down to 7.0 (once again, per SIS). Purely based on the numbers (and without the NFL’s fancier time-to-throw stats on hand to truly confirm), we can intuit that Tech’s asking its quarterback to get the ball out a bit quicker — if we assume that Tech’s receivers run 4.65-second 40-yard dashes (and ignoring a bunch of other factors like defenders), a difference of 2.9 yards in depth of target could result in Tech’s quarterback getting the ball out approximately 0.34 seconds faster.
This 0.34-second difference is fairly significant: if we use this chart of NFL quarterback EPA/play versus time to throw as a basis (note: there is an estimate of college QB time to throw there, but it’s unclear what the basis for that metric is so I’ve chosen to ignore it), that small of a timeframe accounts for nearly 12% of the average time to throw (approximately 2.8 seconds). At that scale, we could be talking about the difference between a pass attempt and an outright sack, an emergency scramble out of the pocket (given enough quarterback athleticism and evasiveness), a batted ball interception, or even a strip-sack. At the very least, it’s potentially one fewer read the quarterback has to make and and one fewer moment the offensive line has to hold against oncoming rushers. If intentional, this is a very good change from the Tech offensive staff: targeting shorter routes accounts for the difficulties the offensive line has had this season and simplifies the passing game for a true freshman’s first-ever start.
However, it’s entirely possible that this is entirely unintentional and driven by Pyron rather than the staff. It’s certainly possible that he feels the pocket collapsing before it actually collapses (given the strength of the offensive line) and makes a quicker decision to throw to a shorter route. We’ll be able to evaluate the nuance between these two possibile realities more effectively as Pyron plays more snaps towards the end of this season, and given his performance versus VT, I would be shocked if he doesn’t play the remaining two games on his redshirt.
3. Sustainability and Consistency
Per SIS, Tech averaged 6.0 yards per play and put together 463 yards of offense — pretty good, right? Well, yes but remember that averages (and totals, for that matter) can be misleading: Tech’s success rate was only 36.4% (well under the national average of ~42%) and punter David Shanahan trotted out onto the field after a three-and-out six times (37.5% of the time). Given SIS’s report, the average Tech drive started at its 24 and petered out somewhere around the VT 46, with nearly 60% of Tech’s plays coming inside its own 40. For the most part, Tech (to borrow a soccerism) struggled to get out of its own end.
BUT the Tech offense was buoyed by a 18.2% explosive play rate (defined as the rate of rushes of 12+ yards and passes of 16+ yards), driven primarily by a 55.6% explosive play rate and 12.0 yards per play on third-and-long (seven yards to gain or more). While this level of execution is admirable (very much so, in fact), this kind of performance is not sustainable given the normally low conversion rates of these types of situations. Tech got lucky five different times on third-and-long; that’s not going to happen every game and heck, it might not even happen again.
In a number of its wins this season, Tech has nailed the explosive plays, but can’t quite dominate consistently over the course of the entire game. To start making progress offensively, Tech has to chain together successful but smaller-chunk plays throughout the game — it can’t bank on getting this lucky every game.
As a final note, let’s talk about explosive plays a bit more.
It’s true that football is an inherently random sport with a notoriously small sample size, and it would be irrational to throw out results that are fueled by explosive plays. However, when we look at a struggling team’s underlying numbers, we’re looking for reproducible and sustainable through-lines of performance: things we can point at to say “look, they’re doing things that correlate highly with wins consistently and doing those things will eventually bear out on the scoreboard”. By their very nature, explosive plays are unsustainable, so it’s difficult to recommend that a program key on them as signs of progress. Ultimately, if you hit on your explosive plays, you’ve gotten lucky — but we’re not looking for luck here: we want proven, consistent results.
