By the (Advanced) Numbers: Dominating Duke

The numbers see Virginia Tech taking advantage of a down year for Duke.

The Hokies win at Pitt felt so good, Whit Babcock got a contract extension just for hiring the guy who coached the game.

Now comes the part of Virginia Tech's football season that will define how fans feel from January through August. Virginia Tech is currently, and likely will remain, favored in every game left on its schedule. But the opponents are not cupcakes, and even if Tech wins the next three, Virginia will be determined to spoil the party.

First, Duke must be conquered. While the Blue Devils have taken a step back this season and fallen short of expectations, they remain well-coached and certainly a threat to call into question the Hokies' success thus far. So what do the numbers suggest is in store?

Where the Season Stands

Virginia Tech is currently 6-2 and will go bowling for the 24th straight season. All irrational optimism aside, before the season few realistically thought bowl eligibility would be secured with four games remaining, and Tech would be favored down the stretch. The actual and predicted score difference and odds of winning each game are:

Important note for the graph above and win totals below: the predictions used before were a combination of multiple systems put together by a site called Superlist. The predictions the site was providing, especially for Virginia, have been getting farther from what most others predict and no longer seemed realistic. So this week I switched to using the estimated score difference provided on the Football Study Hall Advanced Stat Profile of Virginia Tech.

The odds of each possible regular season win total are now:

So...almost a 99% chance of going 8-4 or better in the regular season. Oh, you want to know what Virginia's odds are of making a bowl? This website covers Virginia Tech, not Virginia, and there is a level of professionalism to be maintained that does not include poking fun at others' misfortune.

Seriously though it's 0.1% according to Football Study Hall. But hey, that's the same chances of Tech going 6-6 also!

Rankings and Computer Predictions

The computer rankings and predictions for each team:

The worst computer ranking given to the Hokies is better than the best ranking given to the Blue Devils, and every computer but one predicts a win. The odds of an 11.5-point favorite winning are 80.0%.

Next is a look at any overall offensive or defensive advantages:

In a scenario not often seen in Blacksburg over the past few seasons, Virginia Tech brings a better defense AND a better offense to the table.

When Virginia Tech Has the Ball

Here is an explanation of S&P+ ratings, and FEI ratings. All statistics are now opponent-adjusted.

Who has the advantage in the passing and rushing game when the Tech offense has the ball?

The Hokies continue to demonstrate much better passing than rushing, as well as significantly better performance on passing downs than standard. The Blue Devils, fortunately match up well as passing and passing downs are the weaknesses in their defense. In other words, looks for another game on Jerod Evans' shoulders.

Offensive and defensive line performance are compared using Football Outsiders' metrics:

The biggest weakness on the Tech line remains power success and stuff rate, indicating that the team is getting stopped far too often without gaining even short yardage. The Duke defensive line does not appear poised to take advantage, but isn't particularly weak either.

Now let's take a look at the FEI personality traits of the Hokie offense versus the Blue Devil defense:

Virginia Tech remains balanced in personality characteristics, as does the Duke line.

The Virginia Tech offense is closest in personality to:

  1. Utah
  2. Temple
  3. Nebraska

The Duke defense is closest in personality to:

  1. Oklahoma State
  2. Miami-OH
  3. Mississippi State

When Duke Has the Ball

Again, examine pass-run comparisons first:

The Blue Devils don't stand out on any categories, rating a little below average across the board. The Hokies, meanwhile, are very good and especially difficult to pass against.

Offensive and defensive line performance are again compared using Football Outsiders' metrics:

This again appears to be a good matchup for Virginia Tech, with the one above average category for Duke — stuff rate — being matched by an elite unit in that regards.

As for personality traits:

The mediocre Blue Devil offense exhibits low scores across the board in all personality metrics, and the one category in which the Hokies haven't stood out — drive-ending turnovers — is exceptionally poor for them. Again a good matchup for Tech.

The Virginia Tech defense is closest in personality to:

  1. San Diego State
  2. Clemson
  3. Boston College

The Duke offense is closest in personality to:

  1. Charlotte
  2. Georgia Southern
  3. San Jose State

Special Teams

First a look at the Hokies' kicking units:

As indicated last week, Tech's units seem good on one side of each kick (punt returns and kickoffs) and mediocre at the other side (punting and kickoff returns). But net field position has been elite.

When the Blue Devils kick:

So there you have one of the worst field goal kicking units in the country, along with the best kickoff unit.

Who To Watch Out For

There aren't many Blue Devils who stand out for anything statistically, but dig long enough...:

  1. SR S Corbin McCarthy leads the nation among DB's with 10.5 tackles for loss.
  2. SR Breon Borders is 11th in the country with 11 passes defended.
  3. This week's Mr. Irrelevant is JR Shaun Wilson with a kickoff return average of 26.89 yards (19th nationally), because Joey Slye remains tops in the country with a 64.98-yard kickoff average.

Statistical Key to the Game

The matchups in this game seem to heavily favor Virginia Tech, so let's talk about where it's the Hokies game to lose. One way would be to let the opponent sustain drives by converting 3rd- and 4th- far too often (i.e. Syracuse), and the other would be turnovers. I don't envision Duke's offense having the potential to convert in the way that Syracuse did, so turnovers it is. If Virginia Tech is neutral or wins the turnover battle, it seems highly likely they will walk out with a win.

The Stats that Define Virginia Tech's Season

Revisiting my column from this summer on the stats that will define the season, here is how the Hokies performed thus far (each listed as VT - Opponent):

Opponent PPP Pace (sec/play) ToP
Liberty .40 - .21 22.7 - 25.5 33:42 - 26:18
Tennessee .33 - .71 26.1 - 26.1 31:44 - 28:16
Boston College .64 - 0 27.5 - 26.1 35:15 - 24:45
East Carolina .79 - .25 26.5 - 26.0 30:03 - 29:57
UNC .41 - .05 28.9 - 19.4 39:59 - 20:01
Syracuse .24 - .31 22.2 - 20.1 26:35 - 33:25
Miami .55 - .21 26.4 - 23.5 29:28 - 30:32
Pitt .48 - .65 24.4 - 29.1 33:17 - 26:43

I will not value time of possession.

Statistical Prediction

Duke is a well-coached team, and a 3-5 (0-4 ACC) slide has been a huge surprise, and probably ensures David Cutcliffe another year at the helm as bigger schools are less likely to pursue him. It is natural to wonder whether this is a temporary blip in a successful run, and Duke may start to show potential again on Saturday, but I'm not buying it, especially with leading rusher Jela Duncan out. Hokies win easily.

Virginia Tech 37, Duke 17

As always a thanks to Football Outsiders, cfbstats.com, and Minitab Statistical Software.

Comments

Seriously though it's 0.1% according to Football Study Hall. But hey, that's the same chances of Tech going 6-6 also!

0.1%? Now that's poopy

UVA fan in my office Monday morning, after a long, play-by-play summary of the Louisville game......

"I like this team. We're only a 3 pt underdog to Wake, if we get that win, we always play Miami well. And Bronco has shut down all the triple option offenses he has faced in the past, we will beat GT. And we're due against Tech. I definitely think 6-6 is a possibility."

This is the same guy who wrote "9-3, beat Tech, win bowl" on his whiteboard before the season started to document his prescience and bravado.

The absurdity of any argument that begins "We're only a 3 pt underdog to Wake..."

...and I forgot the part of "...and that's what Wake gets for homefield, so Vegas sees us as the better team."

Again, Wake Forest. And you probably aren't the better team, by like....any...data.

This delusion of his makes me hope that UVA does pull out a win against WF, Miami, and GT...so we get to crush their dreams AGAIN!!!!

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

Luckily, if they lose all three we still get to crush their dreams further.

Nah, I would love for LOLUVA to beat Miami. I want Richt & Miami to not even bowl qualify. We already know that LOLUVA isn't going to make it, so share the hate...

Hard to talk shit about wake after 6-3...

I can imagine no more rewarding a career. And any man who may be asked in this century what he did to make his life worthwhile, I think can respond with a good deal of pride and satisfaction:
“I served in the United States Navy"

KCCO

You mean 3-6?

Virginian by Birth, Hokie by Choice

More LolUVA fans in your office than at their spring festival...

man those 6 field goals really murdered our ppp stat against Pitt..

If a tree falls in Scott Stadium does it make a sound?

Yeah. TOP was important in that game because it appears Pitt was getting more PPP.

Stick it in! Stick it in!

Now let's take a look at the FEI personality traits of the Hokie offense versus the Panther defense

Think you missed a Find/Replace here, Joel.

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

Maybe I just REALLY want to face the Panther defense again.

And maybe I never wanted to have to go to Durham again, but here we are.

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

Unfortunately I won't be in Durham this weekend because I'll be in Greenville with my wife's family but I'm hoping to see Wallace Wade stadium full of maroon & orange. Anybody hearing anything about our expected turnout?

I want to see a LOT of this.....

"Everyone has a plan 'till they get punched in the mouth."

Based on the number of ticket I have in my name, I can safely say there will be at least two.

so there will be at least 2 people still in the stadium after halftime...

"Everyone has a plan 'till they get punched in the mouth."

+4 for me

Jerod 4 President

+2 for me!

Get Angry, Bud!

+ about 160 for my group....

Don't let this comment take away from the fact that Arkansas blew a 24 point 2nd half lead in the Belk Bowl.
Don't let the Belk Bowl take away from the fact that Virginia blew a 16 point 2nd half lead in the Elite 8.

I remain unable to attend.

And this tailgate remains awesome-sounding.

You're the only ones that I can talk to about this, you guys.

I mean... we're going to have multiple kegs, Whit Babcock, and probably the Hokie Bird and cheerleaders...

So yeah, I mean, sounds like a typical Saturday to be honest

Don't let this comment take away from the fact that Arkansas blew a 24 point 2nd half lead in the Belk Bowl.
Don't let the Belk Bowl take away from the fact that Virginia blew a 16 point 2nd half lead in the Elite 8.

ppp was less than Pitts, the only time in our season we have won with a lower ppp. how often does that happen?

Logically I would only expect that in a close game with an abnormally high numbers of field goals for the good guys (this game) or turnovers after long drives to a team that couldn't convert (should have been UT game)

Warning: this post occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors).

Pitt's turnovers had a lot to do with it.

We won with a lower PPP because we played at a faster pace and held the ball longer. So I would think that happens with some regularity, especially with closely matched opponents.

and we came away with 3 points most drives while they came away with 7 at the end of their longer drives.

Warning: this post occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors).

The worst computer ranking given to the Hokies is better than the best ranking given to the Blue Devils, and every computer but one predicts a win

Can someone shine a little light on this for me? I would think that home field advantage may have the potential to sway a matchup that the computers thinks is close, but it looks like more than 5 points on the chart? Not that it matters... just curious...

As always, great stuff.

Save a collar, pop a Wahoo

I can't dig very deep at the moment, but I will lob out that the two graphs have different sources so there's not 100% overlap in the computers that are shown on rankings (from masseyratings.com) and those shown on predictions (from predictiontracker.com).

Another aspect is that, especially away form the very good and very bad teams, a wildly different ranking may not be worth many points and each system rates their own home field advantage.

Otherwise, I'll have to see what the specific computers are to figure it out...

Thanks for the response, Joel. Seems like the different sources and lack of consistent home field advantage can explain it away well enough. Since I was likely over-analyzing anyway I would have also accepted this response...

Save a collar, pop a Wahoo

The computer prediction predicting Duke by 6 is simply labeled "Least Squares w/HFA" on Prediction Tracker, with no link to a source. Safe to say it is not likely included in Massey's comparison of computer ratings, and may be a basic system run by Prediction Tracker.

Joel, love your columns. Quick question, on your "Statistical Key to the Game" would you mind providing a bit of analysis/interpretation? In your link to the pre-season article, you explain the importance of each and it makes a lot of sense, but i'm having a hard time understanding if we are currently doing good/bad/as expected.

I feel like i'm missing out on something profound!

Good question! PPP is a good measure of how effectively each team created points on offense and prevented them on defense. When that doesn't align with the outcome you know something different must explained what happened, like turnovers, unusually high or low frequency of stalled drives, or dominant field position. Pitt is a good example - they were much more effective at creating points, but thanks to turnovers and field position the Hokies walked out with the win.

Pace was really just to highlight that Fuente likes to run at a faster pace. What's shown in the table, however, does include periods of games when he is trying to kill the clock and running at a slow pace. Going in I didn't realize just how often the team would be in a position to kill clock. By S&P's adjusted pace number, which removes "garbage time", VT is running the 19th fastest pace in the country.

And time of possession is just to illustrate that it is meaningless. The games where we dominated time of possession were only due to clock killing.

One computer really does love us, huh?

I like that one, we should adopt it

Don't let this comment take away from the fact that Arkansas blew a 24 point 2nd half lead in the Belk Bowl.
Don't let the Belk Bowl take away from the fact that Virginia blew a 16 point 2nd half lead in the Elite 8.

The poll is Nutshell. They have us #2

So someone keeps a poll exclusively to set up "In a Nutshell" jokes.
.
.
.
I like these people.

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

Warning: this post occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors).

I think he did this joke in Wayne's World as well. Very funny.

In a nutshell, yes

If it was a physics program, it'd be named "In a Vacuum"

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

We swarm

"Jerod Evans had more moves on that run than Bayer has aspirin!"

-Mike Burnop, 9/24/2016

This is how I feel everytime is start looking at the graphs on a "By The Numbers" post.

That's me when I start to read French's analysis.