Advanced Stats 2020 Season Review Part 1: The Big Picture

In the midst of a rebuilding process that can feel long and discouraging, we want progress. As we trudged through a season that was disjointed for fans and brutally hard for players, we all want to know: was the 2020 football season for Georgia Tech a step back, a step forward, or a step in place?

2019 was a massive step (or three) back. But there are compelling statistical reasons to think 2020 was a pretty big leap forward.

Let’s take a look first at how some of GT’s raw advanced stats compare between 2019 and 2020.

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

2020 Offense vs. 2019

With the addition of Jeff Sims and Jahmyr Gibbs, the Georgia Tech offense rose from the dead in 2020. The previous year’s offense was historically bad, but this year showed improvement across the board except in allowing runs to be stuffed for no gain or a loss. Despite frustrating struggles with turnovers, penalties, and pass protection, the offense was definitively better than last year. In particular, the gains in success rate and offensive EPA/play show a jump from the very bottom of the country to something close to average. Seeing this kind of one season jump in offensive production while starting three true freshmen and still working to get the offensive line in place is hugely encouraging. I’m happy to admit that some of my preseason criticism of CDP may have been misguided, although there is plenty of room left for improvement in play calling and creativity. The persistent struggle with run stuffs, as well as the high pass pressure rate allowed that we will look at more next week, point squarely at the offensive line as the unit still in greatest need of improvement.

2020 Defense vs. 2019

Just as the offense improved almost across the board, the defense declined in almost every significant metric. Yards per play was about the same as 2019, while defensive success rate and defensive EPA/play performance regressed. Defensive run stuffs improved slightly, while the already below average havoc rate took a worrisome turn downward.

We expected growing pains to continue in the front 6 in the positions that lacked talented depth. We did not expect the secondary to fall apart as it did in 2020. The secondary returned nearly all of the personnel from 2019 but did not look like it. The defensive coaching staff, which looks like it will return next year unchanged, has its work cut out to develop and deploy personnel in more effective ways in 2021. Sophomore Zamari Walton came out as PFF’s highest graded member of the secondary, followed by Junior Tre Swilling (who battled an injury for much of the year) and Freshman Wesley Walker. That bodes well for improvement next year, but we need to see it.

Were we really better in 2020?

We’ve seen how 2020 GT compares to 2019 GT in these raw metrics. What about how 2020 GT looks nationally to how 2019 GT looked nationally? Here, the improvement becomes very clear. The following series of charts takes three important metrics: yards per play, EPA/play, and success rate, subtracts each team’s defensive number from its offensive number, and gives us a net performance in each category. We will show charts for both 2020 and 2019 in all three areas. When you look for GT, you’re going to be looking pretty far to the left in the 2019 charts and a lot closer to the middle in 2020.

Putting it all Together

The creatively and non-narcissistically named Binion Team Strength Index combines the three above metrics into one overall measure of team strength. We will be working on developing this metric more in the offseason to turn it into a true power ranking, but for now, it gives us a very insightful picture into the overall standing of GT in 2019 vs. 2020.

In 2019, GT graded out as the fifth worst team in all of the Power 5, almost 1.5 standard deviations below average. This year, GT lands almost exactly in the middle, just a hair below average. In case you’re wondering if this is significantly affected by a difference in schedule strength, it’s not. I ran an average of four different ways of calculating strength of schedule: GT came in at 47th in 2020 and 51st in 2019, which is not a statistically significant difference.

3-7 wasn’t fun. But the underlying performance was miles better than the performance underneath 2019’s 3-9 record. Take heart. We did take a step forward this year.

In the coming weeks, we will look more closely at the offense, then at the defense, then at our 2020 predictions and goals before finishing with a look at some of the hidden factors like special teams, penalties, and turnovers to help us begin to get an idea of what to expect in 2021.