The Hokies took care of business in their first trip to visit an ACC member in California last Saturday in their 31-7 win over Stanford. In terms of Net EPA per play, this was the fourth best win of Brent Pry's career (behind UVA, Wake Forest, and Syracuse last season) and the strongest of 2024. It was the 9th best game of Week 6, and the 5th best among major conference schools.
Virginia Tech dominated on defense against an offensive making a change at quarterback, while posting their first game of the season with a positive offensive EPA/play in every quarter of action. Combining the results against Stanford with the results against Miami (strong offensively but poor on defensive), Virginia Tech moved (slightly) into the top quadrant of teams in unadjusted net EPA/play.
Even though the win was not overly flashy, Virginia Tech had some of their most success passing the ball. Successful plays are first down plays that gain 50% of the yards to gain, second down plays that gain 70% of the yards to gain, and third and fourth down conversions. Virginia Tech hit 50% passing success against Stanford, well above their average for the season (38.9%) and was explosive on 15.0% of their passing attempts (above their season average of 12.6%).
The Hokie took a hit against a surprisingly stout Stanford rush defense, but overall have improved slightly on their rushing explosiveness (0.5% increase) while maintaining their rushing success rate (down 0.06%) since the Week 4 recap.
On defense, Virginia Tech held Stanford to about their season average on overall success rate (41.8% in game, vs. 41.2% on the season) but did not allow a single pass play over 20 yards and cut the explosive run rate (runs over 15 yards) by 2/3rds from 8.7% to 2.7%.
The reason I like to look at success rate and explosiveness is that it allows you to look at teams' ability to move the ball both in chunk, explosive ways and methodically down the field without taking into account scoring and turnovers. Good teams manage to be both successful and explosive; and for most of the season, Virginia Tech has been explosive on offensive without sustained success, while having decent down-to-down success on defense, but allowing too many big plays. The Miami and Stanford games have so far shown a offense that is getting back to what worked on a down-to-down basis last year, while maintaining explosivity. The defense is continuing to improve on explosive plays (Miami, a top-5 offense notwithstanding) and starting to also generate turnovers. While the turnovers are great, having a sustainable defense is going to be what can take this team to reach their full potential.
Comments
Is there a link to what the adjusted net EPA/play look like? VT has played a decent schedule. Not sure that helps though, if the best was against a weak Stanford.
Parker Fleming, @statsowar on Twitter, I believe has more thorough rankings for adjusted net EPA for College Football Insiders. I made this using the College Football Database API, hence why it is unadjusted.
Fleming has VT ranked 48th overall, 51st in offense and 42nd in defense.
My unadjusted model has VT at 54th overall, 66th in offense and 56th in defense - so not that far off all things considered. We are definitely hindered in an unadjusted model.
Schedule to Date:
Schedule Remaining:
Using the adjusted model, only Syracuse is ahead of us on net EPA. Using my model, Boston College, Georgia Tech, and Clemson are ahead of us. Overall... A lot of winnable games on the schedule if we play the way we know we should play.
Here is a table from another advanced stats site, gameonpaper. I usually use this site to review match-ups post-game, but they have a leader board that tracks adj EPA/play by net stats, offense or defense. I got rid of a lot of the teams to make it more readable and left the ACC, our OOC opponents, and a few other teams I thought were interesting to use as comparison.
I'm pretty thrilled that I'm in the range of this and College Football Insiders since I'm not running a website or that strong of an analyst (in my own personal estimation of myself).
Honestly if you have a metric with VT anywhere between 30th and 70th I feel like I'd buy it. Every year it seems there's 6-12 really good teams and 40 more who could beat them on any given night.
We're rooting for 18-23 year olds who also sometimes play school. It's hard to definitively say who is "good" each week. But come up with a couple of comparison points (I like success rate and explosiveness but pick your poison) and you can start to find some insights. May not always mean something, but it might!
Yea, he's a good follow. As a stat and model nerd myself, I definitely appreciate the data and analysis you provide.
Clemson is surprising. And so is Duke. Parker's model definitely fluctuates more than Bill C's model. I wish they spoke more on the intricacies of their models, but, I guess trade secrets have to be kept secret to be intellectual property.
Speaking of business you can create custom jerseys now. Wont let you do a 7 Vick though..
https://www.fanatics.com/college/virginia-tech-hokies/virginia-tech-hoki...
Edit: Wrong link, this is the custom one
What's a man gotta do to get a jersey with stitched numbers and letters these days
DIY baby!