Advanced Stats Check In: Bye Week

After three games, Georgia Tech sits at a slightly disappointing 1-2. Pregame win probabilities pegged us to go 1.3-1.7 in these three games, and postgame win expectancies predicted the exact same number based on underlying success rates, yards per play, and EPA/play.

I’ve been disappointed at some of the areas where we have struggled so far this season, but it’s important to step back and let the numbers help guide our thinking. The first bye week is a good time to do an initial progress report. Where is growth evident? Where are we slipping? Let’s dig in.

*2020 numbers come from my play by play charting. 2019 numbers come from @CFB_Data

Offense

We’re improving! This is good! The only category with a significant downturn in 2020 is offensive EPA/play, and this decline is almost completely attributable to the turnovers committed by the offense. In some ways, that’s good news, because baseline efficiency numbers tend to be more predictive of future success than turnover margins and explosive plays. In the efficiency metrics, the offense is better across the board. Though they aren’t displayed here, both passing and rushing success rates are improved. The quarterback play has ups and downs but has had more consistency, and the skill players are improved.

When people see the high turnover numbers, almost all of the blame tends to go to either the thrower or carrier of the football. But we also have to consider the role of the offensive line. I started tracking pressure rates this past Saturday, and I’ll be doing that throughout the season. The average rate during a game is to create pressure on about 30% of the opponent’s passing plays. GT gave up a 45% rate this past weekend, which is awful and helps explain the 5 turnovers. Pressure rate is a strong predictor of turnovers created, so that will help us see how we are doing at improving one of the key underlying issues leading to this early rash of turnovers.

Defense

On the defensive side, the results are more of a mixed bag. We’re giving up a slightly higher success rate and achieving a slightly lower havoc rate. Given that we have played two of the worst offenses we will see all year, this is concerning. Missing several defensive starters in each game certainly hurts, but this side of the ball returned almost everyone of note from last year, so the lack of progress is discouraging.

On the other hand, we’re giving up less yards per play and less EPA/play, run stuffs have increased slightly, and third down defense is better. The defensive end position in particular has been a bright spot. The two defensive ends who started against Syracuse both currently rank in the top 25 nationally at their position according to Pro Football Focus’s proprietary grading system. That’s hugely encouraging for Marco Coleman’s group

Verdict

I’m coming away from this analysis feeling more optimistic than I did after the last two disappointing Saturdays. We’re not yet where we want to be, but we’re not still in the same place we were a year ago. I’m thankful the numbers helped bring me some clarity and grounding when I was starting to panic; I hope they help you look at the rest of the season with more clear eyes as well. Looking forward to taking on Louisville!