This past Saturday, Georgia Tech walloped Virginia 45-17, giving Tech back-to-back wins for the first time this year and bringing them above .500 for the first time since right after the first week of 2020. Needless to say, it was a pretty big win for Brent Key and the Yellow Jackets.
The full advanced box score can be found below. Today, we’ll be highlighting Tech’s rushing game.
In Tech’s game against UNC a week ago, former head coach Paul Johnson was honored for his induction into the College Football Hall of Fame. The Yellow Jackets are now 2-0 with two dominant rushing performances since they honored Johnson. Coincidence? Yeah, probably, but I’m a big fan of it.
Although the total number of rushing yards declined slightly (350 to 324), I think I can say that Tech’s rushing performance against Virginia outshone that of the performance against UNC.
Individual Rushing Performances
GT-UVA Individual Rushing Performances
| Player | Carries | Rush Yards | TD | Fumbles | Yards/Play | EPA/Play | SR | WPA |
|---|---|---|---|---|---|---|---|---|
| Player | Carries | Rush Yards | TD | Fumbles | Yards/Play | EPA/Play | SR | WPA |
| Haynes King | 8 | 84 | 2 | 0 | 14.00 | 1.29 | 83.33% | 20% |
| Dontae Smith | 15 | 88 | 2 | 0 | 5.87 | 0.36 | 40.00% | 8% |
| Jamal Haynes | 17 | 127 | 1 | 0 | 7.47 | 0.03 | 47.06% | -19% |
For the purposes of this, I’ll only be focusing on these three players, but it’s worth mentioning that Zach Pyron and Evan Dickens also had a combined 25 yards.
Obviously, these are all incredible rushing performances, but the thing that stands out to me here is that Jamal Haynes had an overall negative contribution to win probability despite being the leading rusher. After doing a bit of digging, this is a byproduct of poor play-by-play data on ESPN’s side. Let’s take a look at that series of plays.
Data courtesy of GameOnPaper.com
Notice anything that’s slightly off? The PBP data somehow forgets that UVA scored a touchdown until it’s updated after Haynes’ 5-yard rush on first down (which would be considered a successful play). That results in the run accounting for -22% win probability added. With that clear outlier removed, Haynes’ WPA moves to roughly 3%. (The kickoff return play is also heavily skewed because it’s showing the return as wiping UVA’s points off the board which would drastically swing win probability.)
The reason it isn’t higher is because his longer rushes weren’t until later in the game when Tech already had a commanding lead, so they didn’t have as much impact on the game. For instance, his 43-yard touchdown came halfway through the fourth quarter when Tech’s win probability was already at 99.8%. The game was already assuredly in Tech’s hands, so despite it being the third-biggest run from an EPA perspective, it didn’t swing the game very much.
While we’re talking about individual performances, I also want to talk about Dontae Smith. In the last two games, Smith has come seemingly out of nowhere and rushed for a total of 266 yards on 37 carries with three touchdowns. He’s averaging 0.39 EPA/play and is just playing his best football of the season. Going into UNC, Smith only had 13 carries for 62 yards and no touchdowns. That’s not necessarily bad (over 4 yards/carry), but not close to what he’s done in the last two games. I’m not sure what changed here, but if Smith keeps playing like this and Haynes continues playing like he has, Tech is going to keep beating teams with their running game.
Consistency vs. Explosion
Game On Paper defines an explosive rushing play as one that earns more than 1.8 EPA. For example, Haynes’ run I mentioned earlier that was the third-best from an EPA perspective earned 3.67 EPA. That would be considered an explosive play. In most games, teams are dependent on those explosive plays to finish with a positive EPA rushing performance.
Against UNC, Tech finished with 0.30 EPA/rush across 48 attempts. Five of those attempts were considered explosive plays by Game On Paper’s definition. Taking out those explosive plays, Tech’s EPA/rush dropped to -0.08. For taking out explosive plays, that’s not a horrible number. For comparison, Tech’s EPA/dropback dropped from 0.62 to -0.17.
Having an EPA/rush – explosive plays around 0 shows pretty remarkable consistency. This is because the rushing game is not reliant on those explosive plays to be productive.
Georgia Tech continued that trend against UVA. After having a better overall rushing performance (0.33 EPA/rush), Tech also had a more consistent one, finishing with an EPA/rush – explosive plays of 0.0. Now obviously, there is still room for improvement, but I think Tech’s rushing game is in a really good place right now, and I hope to see it continue to improve.
Read more
