Georgia Tech Football: GT vs. Clemson Advanced Stats Review

Final Score: Clemson 41-10

Model Prediction: Clemson by 25.5, Clemson to cover —> correct

Projected EPA (Offense and Defense) Margin of Victory: Clemson by 18 (before garbage started with 6:30 left in the fourth quarter)

GT Win Probability (Based on Success Rate, Yards Per Play, and EPA): 10%

Georgia Tech kicked off the season with a valiant two and a half quarter effort before succumbing to the superior talent and depth of the Clemson Tigers. Overall, the defense assuaged some fears that the malaise of last season would carry over, while significant questions persist regarding the offense. Of course, those observations must be made in the context of Clemson having a much better defense than offense at this point, as well. There was enough improvement from the disastrous end of 2021 to maintain intrigue with what this team will turn out to be, but there was also enough of the same kind of failures to prevent optimism from rising too far.

Just how competitive was the game? Clemson took a 24 point lead with about six and a half minutes left in the game; at that point Clemson had an 18 point advantage in offensive and defensive EPA; the two blocked punts contributed about 7.8 total EPA to Clemson, so we can fairly say that 2/3 to 3/4 of the margin of the game had nothing to do with those two failures by the GT punt team. That means Tech played just ahead of expectations looking at the more predictable aspects of offense and defense. Overall, that tracks with my sentiment about this game; there was some improvement, but it’s also not nearly enough for this coach’s fourth year at the helm. Let’s dive into some of the finer details from this one.

Advanced Stats Box Score

Clemson solidly outplayed Georgia Tech, but this was nothing like the blowout experienced in the Notre Dame and Georgia games to close 2021. Looking across all of the statistical categories we track with our in-game charting, Clemson held an advantage in 17 out of 23. That is actually a better result for GT than what we saw in our advanced stats preview, where GT held a predicted advantage in only one out of 21 categories based off of numbers from last year. Again, we can see evidence of GT running slightly ahead of expectations, while still being far from a good football team.

When GT Had the Ball

The yards per play, yards per pass, total EPA, and explosive play rate numbers on offense are especially concerning. Clemson will continue to have an excellent defense and restrict opposing offenses, but performances with single digit percentiles do not augur well for GT’s offensive prospects.

Georgia Tech came out taking a shot, calling the same play that got Malachi Carter a long touchdown in the 2020 version of this rivalry. Unfortunately, Luke Benson was left to block KJ Henry, by himself. That didn’t go well, and the subsequent pressure led to an inaccurate throw intercepted by Andrew Makubu. From there, the GT offensive plan shifted considerably, relying almost exclusively on quick-game passes and conservative runs. GT wouldn’t turn the ball over again, but it also wouldn’t find much in the way of explosive offense the rest of the game either.

Quarterback

Jeff Sims received lots of accolades from the Tech fanbase for his performance against Clemson. I understand the sentiment, as the offensive line and wide receivers certainly didn’t do very much to magnify what Sims could do. On the positive side, his CPOE (completion percentage over expectation) on throws of 10+ yards was magnificent. These were the kind of throws that led to the key fourth down conversion to Malik Rutherford, to the near breakaway touchdown by Malachi Carter, and to the gorgeous touchdown pass play with EJ Jenkins. Without a gamebreaking threat at wide receiver, these throws will have to be foundational to the offense going forward.

On the other hand, Sims wasn’t great. His EPA/pass when not pressured was negative. That can’t be pinned on the offensive line, and that number is generally much stickier game to game than the with pressured EPA/pass number. Further, Sims tried three throws 20 or more yards down the field, and none were completed. That’s worrisome. If this offense can’t force defenses to respect a deeper threat, it’s going to be hard to continue finding that intermediate space. Overall, Sims earned a 60 grade from PFF, which is barely passing. Beyond the opening interception, he had two more turnover worthy plays that revived concerns about questionable decision making. Sims battled hard, but there’s still plenty of room for skepticism here.

Rushing

As we’ve repeatedly highlighted leading up to this season, Jeff Sims is currently the best rushing threat on this team. On Monday, there was no rushing game to speak of outside of Sims.

To find more dynamism on offense, Chip Long has to call more than six run plays for Sims. Going forward, we will learn more about the tailbacks against non-Clemson defenses, but the early results are worrisome. Per PFF, Dontae Smith graded out at 67, while Hall and McDuffie were both in the mid-50s. Jeff Sims looked to be the only runner capable of creating behind the porous offensive line.

Receiving

As expected, the slot receivers were clearly the best options amongst the receivers, the tight ends still aren’t a viable threat, and the outside receivers struggle to consistently get open and catch the ball.

The distribution here was a bit confusing: in the first half, Nate McCollum was excellent, catching six passes on 8 targets with 5 of them going for successful plays. In the second half, he had zero targets. Beyond him, Malik Rutherford was the highest graded receiving target for GT, but he had only the one target on the impressive fourth down conversion. Malachi Carter continues to frustrate, occasionally making a play but then also missing opportunities, as when he fell down on the play that could have been a touchdown. EJ Jenkins had high highs and low lows, averaging out to only a 53 PFF grade. Luke Benson certainly didn’t look like the dynamic receiving threat that was promised and didn’t help much as a blocker. He had no successful catches on 5 targets and checked in at a cool 36 per PFF. The threat of Leo Blackburn is badly needed, but who knows when or if that happens this season?

Offensive Line

Now, everyone’s favorite topic: the offensive line. Four new starters lined up compared to the opening game last year. It didn’t help. Other than Jordan Williams, every lineman who played significant snaps had a failing PFF grade. The guard position is an abomination. Joe Fusile had no chance against Brian Bresee, but Quick and Techie graded out even worse. As the coaches continued to talk about trying to find the best combination leading up to the opener, many of us were concerned that there simply weren’t five good options, and that clearly turned out to be a well-founded fear. This offense has some interesting pieces, but there’s no way to ever be consistent or avoid the disaster plays with a line this bad.

Also, Luke Benson gets to be an un-honorary member of this group. A note to Coach Long before calling that game-opening play again: please don’t put non-blocking tight ends in man to man blocking situations with KJ Henry. That’s all.

When Clemson Had the Ball

Now, surprisingly, we move to the brighter half of the game. The defense mostly held Clemson to average or below average results on offense. The defensive line looked stout compared to recent iterations, the linebackers were in the right places, and the secondary often had the receivers blanketed. There were impressive building blocks in this performance!

Rushing Defense

In our preview, we had identified one area of statistical advantage for the GT defense in last year’s numbers, which was EPA/rush. Unfortunately, that didn’t carry over into Monday night’s contest. Clemson ended up with about an average rushing performance, largely thanks to the second half work of Will Shipley and DJ Uiagelelei.

The Georgia Tech rush defense did a fantastic job of limiting big rushing plays. The longest carry of the night went for only 15 yards. But the defense wasn’t quite disruptive enough against the run. Clemson managed an enormously high 55% opportunity rate (the % of carries that go for four or more yards when that many are available) and had only 13% of their runs stuffed by GT. The Georgia Tech safeties did a good job in run support; the defensive line largely did a good job fitting their gaps; the linebackers were ok, although the absence of Charlie Thomas after his ejection saw the running game tip decisively in Clemson’s favor. Thomas’s PFF grade ended up at 75, while his backup checked in at a mere 57. With the starters in, this should continue to be a good group against the run, although the backups look like they may struggle to hold up.

Defensive Disruption

The run stuff rate was low for the defense, but when DJ Uiagelelei attempted to pass, Georgia Tech did manage to bring relatively consistent pressure. PFF charted significantly more pressures than I did on first watch, and I include those numbers below.

Keion white was everything I hoped he would be, turning four pressures into two sacks. After the game, he expressed dissatisfaction with his performance. I can’t wait to watch him continue to unleash that hunger on opposing offensive tackles. He’s a player that Georgia Tech has been desperately needing for years. At the other spot, Kevin Harris brought the most pop off the edge, while Sylvain Yondjouen was excellent in his snaps in support of White. There were signs that this will be a more disruptive GT defense than what we have seen over the past three years.

Pass Coverage

In another area of encouragement, there were none of the absolute coverage busts that became such a hallmark of the 2022 campaign. Again, there was improvement.

Myles Sims was the most consistent player in coverage on Monday night. As we expected, there was a significant drop-off when the backup corners were in the game. Clemson’s second touchdown drive came almost entirely at the expense of Kenan Johnson. The defensive coaching staff is going to have to search for some schematic answers when some of the starters need a blow; there’s still a significant lack of playable depth at all three levels of the Tech defense.

EPA Highlights

EPA calculates the expected number of points added (or lost in the case of a negative number) on a particular play based on the down and the location on they field.

As always, we’ll take a look at the most helpful and hurtful plays for GT.

Most Helpful Plays

1. 4.49 EPA – DJ Uiagelelei fumble recovered by Georgia Tech on 1st and 10 at the GT 22.
2. 3.38 EPA – Jeff Sims’s third quarter touchdown pass to EJ Jenkins on 3rd and 7.
3. 2.53 EPA – Jeff Sims’s fourth down completion to Malik Rutherford from the Clemson 37.
4. 2.47 EPA – Jeff Sims’s two yard keeper on 4th and 1 at the GT 37.

It’s encouraging to see Jeff Sims’s name appear throughout this list; I look forward to seeing more explosives against lower-caliber defenses.

Most Hurtful Plays

1. -4.63 EPA – The first blocked punt, setting Clemson up at the GT 5.
2. -3.35 EPA – DJ Uiagelelei’s 28 yard completion on 3rd and 5 from his own 27 early in the fourth quarter.
3. -3.20 EPA – The second blocked punt, setting Clemson up at the GT 15.
4. -2.90 EPA – DJ’s 27 yard completion on 3rd and 3 from his own 40 in the middle of the second quarter.
5. -2.60 EPA – Defensive Pass Interference on Zamari Walton to convert 3rd and 10 in the middle of the third quarter.

Punt blocks and pass defense allowed the biggest plays of the night for Clemson. The punt protection should be easily fixable, while GT fans wait with baited breath in hope that the pass defense does not revert to last year’s levels. These plays were nothing compared to what we saw last year, but the concern isn’t gone.

Tracking Season Goals

*I set these goals for the 2022 season in some of my offseason preview work. We will be tracking them as we go this year.

Six out of seven metrics fell below the goal for the game, although almost all of them were close. The continued reliance on run plays in second and long situations was discouraging, as was the insistence on throwing at or behind the line of scrimmage. It’s possible to get the ball out quickly and push it a bit further down the field, which is especially important when there aren’t tons of broken tackles happening from the pass catchers. Chip Long can continue to push the offense forward in those areas. On defense, I’m largely encouraged, as performances at a similar level to this one will produce even better results against less talented opponents.

Takeaways

1. It’s time to talk about the failures of game management. With about a minute left in the first half and Clemson facing fourth down, Geoff Collins accepted a ten second runoff and then allowed Clemson to exhaust the entire play clock before punting…while holding three timeouts in his back pocket. I’ve heard his halftime and postgame explanations, and neither holds up to the reality that he sacrificed a possession. When David’s facing Goliath, David needs to throw every rock at his disposal, even if it is risky. Then, in the second half, Collins managed to exhaust all of his timeouts before the fourth quarter started, including one where Dabo baited him by shifting formations on a punt and then laughed when Collins called the expected timeout. Of course, the timeouts didn’t end up mattering in this game, but they sure might in another one. Then, on the last possession of the third quarter for GT, with the score still only 24-10, the drive starts with a two yard run for McDuffie, followed by another two yard run for McDuffie, then a dropped pass by Benson. It felt like going through the motions, if not giving up. I will give Collins credit for the gutsy (and correct) decision to go for it on 4th and 1 from the GT 37, although this was paired with the curious decision to later try a 50 yard field goal with a kicker who had no previous career makes. The game management decisions appear just as misguided as they did in 2019, and that’s tough to stomach for GT fans.
2. The combination of a weak offensive line and a lack of game-breaker on offense is a tenuous one. Last year, Jahmyr Gibbs papered over some of the offensive line failures with many an explosive touchdown run or catch. Where will those explosives come from this year? Jeff Sims needs to be freed up to run more, and the slot receivers need a higher share of the receiving targets. As the five false starts showed, this isn’t likely an offense that is going to play mistake-free, efficient football drive after drive. Explosives have to help bring the points.
3. The defense put up a great first showing. Can the starters continue to play at that high level, while the coaches find ways to cover up the lack of depth? Clearly, backups have to be mixed in with starters within defensive levels. When the backups play together, the drop-off is dangerous. Keion White looks the part of the game-wrecker we need; the linebackers played solid before Thomas’s unfortunate ejection, and the secondary looks at least adequate. That’s a recipe for an order of magnitude jump up from last year.
4. From this one game, I don’t think my outlook on the season has changed. The offense looked a bit worse and the defense a bit better than I expected. The game management and special teams issues are liable to cost GT a game or two that should be winnable. This still looks like a three win team, but we won’t really know until the Ole Miss-UCF stretch. Buckle up.