Georgia Tech Football: Advanced Stats Preview - GT vs. Clemson

Saturday afternoon in Death Valley pits together two teams who have not gotten off to the starts they wanted in 2021. Clemson’s defense under Brent Venables looks as formidable as ever, but the offense has looked glitchy after enjoying the worry-free life of having Deshaun Watson or Trevor Lawrence start at quarterback for 6 of the past 7 years. D.J. Uiagalelei (hereafter, DJU) boasts outstanding talent but is not a finished product. On the flip-side, Georgia Tech fans have ridden the roller coaster of a Jeff Sims injury, a brutal loss, and an outstanding follow-up performance from Jordan Yates, while entering Saturday without knowing who will start at QB. (The social media tea leaves seem to be pointing towards Sims.)

Geoff Collins has not had a good run of it against the Tigers. His first two matchups against the I-85 nemesis have left GT on the bottom of a 125-21 scoring margin. Understandably, there is trepidation in the fanbase heading to Clemson this Saturday. Should we expect more of the same in this series, or is GT primed to put up a different kind of outcome on Saturday?

We’ll dig into the numbers, showing the 2021 year to date stats but adding come context as well, as 2 games is still quite a small sample size.

When GT Has the Ball

*GT numbers come from my play by play charting. Opponent numbers come from @CFB_Data and teamrankings.com

After starting the season against two relatively weak opponents, GT’s numbers are strong across the board on offense. Unfortunately, even with one elite opponent already played, Clemson’s defensive numbers are out of this world. Other than a few GT numbers that are artificially encouraging because of the KSU matchup, Clemson has the advantage across the board. That’s a story that should feel familiar to GT fans, as Brent Venables has largely shut down the GT offense for the past - checks notes - 7 seasons. Yikes.

Throughout the offseason and early part of the season, I’ve been trumpeting the need for GT’s offense to pass more, specifically on 1st down. This may be the one week of the season that I do not endorse that plan. Clemson’s pass defense has been nearly perfect. JT Daniels ended up with an average depth of target of just 3 yards against the Clemson defense, due to the pressure from up front and the swarming coverage on the back. With GT’s offensive line struggles and relative inability for pass catchers to work themselves open, it’s hard to see GT being consistently efficient in the passing game. If CDP asks the quarterback to drop back 40 times, it’s not hard to see Clemson having 8-10 sacks.

Instead, this is the game to lean into the run and then look to hit a few chunk plays like the nearly perfect call that led to Jeff Sims completing a long touchdown to Jalen Camp against Clemson last year. The quarterbacks will need to be part of the running game, and further, this is a game that would lean more into the strengths of Jahmyr Gibbs and Dontae Smith in the running and short passing game. Look for Gibbs to increase his workload and for Smith to get equal snaps to JP Mason this week.

The goal: slow the game down, minimize possessions, and create some positive value with the running game (EPA/rush >0). If GT can do that, things will will look a lot different than last year.

When Clemson Has the Ball

The recipe for a close game combines the above offensive formula with what, so far, has been a bad Clemson offense. In previous years, GT would have had the advantage in exactly 0 categories in the above table. This year, GT leads in 6 out of 10 when comparing its defense to the Clemson offense. Of course, Clemson’s numbers are skewed by having played the undisputed best defense in the country, but they also played a bad FCS opponent and weren’t particularly efficient. DJU completed only 58% of his passes for 7.1 YPA, and he threw an interception for the second consecutive game. In ESPN’s opponent adjusted QBR, he currently ranks 96th nationally.

Between what looks like a relatively weak Clemson offensive line, breaking in a new QB, and a large overhaul at the skill position, this is likely GT’s best chance to hold Clemson in check since Deshaun Watson was hurt early in the 2014 contest. What would that look like?

Obviously, the GT front won’t provide anywhere close to the disruption created by UGA, but the front showed flashes against KSU of what will be necessary against Clemson. It’s disappointing that Keion White does not look on track to play this week, but Jordan Domineck, Jared Ivey, and Charlie Thomas provided consistent disruption last Saturday. The three of them need to play as many snaps as possible. Djmon Brooks created three QB pressures this past week; if he could produce even 1 or 2, that would go a long way in making DJU uncomfortable and keeping him from stepping straight up in the pocket.

On the back end, you can’t give up explosives. DJU has not shown the ability to be consistently efficient, and he will be looking for chunk plays against a GT defense that has often been susceptible. If GT can keep Clemson from completing 5 passes worth more than 1.5 EPA, that would be a huge victory.

So here it is: Pressure up front + no lapses deep on the back end = GT can keep this thing close.

Prediction

The consensus Vegas line has bounced between Clemson by 28.5 and 29, which translates to a 98% win probability. I’m under no illusions that GT will win this game. But I’m also convinced that it will look a lot more competitive than the last few years have in this series.

Hopefully, you’ve had a chance to check out the Binion Index on FTRS by now. Clemson has lost several ratings points since the start of the season, and the model projects a 25 point Clemson victory.

The relative weakness of the Clemson offense to open the season points me in that direction as well. I think GT can muck things enough to keep it from getting out of hand. Clemson 35-14.

Vegas: Clemson by 28.5

My Pick: Clemson 35-14

The Binion Index: Clemson by 25

Year to Date Against the Spread: 50-41-5 — 54.7% (Goal: >=55%)

Average Absolute Error: 13.95 points per game (Goal <= 12.5 points per game)