After a dominating 51-24 win at Purdue, it was easy for Hokie Nation to feel good about the remainder of the season. The offense was rolling. Beamerball was back. The less-than-elite performances by the defense were an anomaly; certainly correctable given the talent and Bud Foster's acumen. Just a week later, the sky is falling. Virginia Tech's defense was inept in a 28-35 loss at East Carolina. Moreover, miscues and penalties plagued both the offense and special teams. And now Foster's unit will have to get by without star corner Kendall Fuller.
In either case, the counter to irrational emotion is analytical thinking. Looking at hard numbers and film reviews restores a sense of control and clear thinking about what to expect from the Hokies moving forward. Even in areas where the team is clearly mediocre, seeing how Virginia Tech stacks up nationally is a good reminder that while the team may not be where fans want, there are many, many teams that are clearly worse. Thanks to how teams cluster towards mediocrity, the numbers also offer reason for genuine hope—even a moderate improvement moving forward can move the national ranking by leaps and bounds.
Moving on from East Carolina, a team the Hokies inexplicably seem to struggle with, this week's By the (Advanced) Numbers takes a look at Pittsburgh, a team the Hokies inexplicably seem to struggle with.
Where the Season Stands
Virginia Tech is currently 2-2. The actual and predicted score difference and odds of winning each game are:
But hey, upsets feel really good and look at how many opportunities Tech has!
The odds of each possible regular season win total are now:
It is currently more likely that Virginia Tech does not make a bowl than does.
Rankings and Computer Predictions
The computer rankings and predictions for each team:
Most computers are seeing Pitt as a slightly better team, but given homefield advantage Virginia Tech tends to be favored. The odds of a 4.5-point favorite winning are 63.2%.
Next is a look at any overall offensive or defensive advantages:
S&P+ is one of the computers that has Virginia Tech rated higher than Pittsburgh, but the margin is very thin. In what is a clear trend at this point, the Hokies' offense is rated higher than the defense.
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?
Thanks to the inept rushing committee formed in Blacksburg, Tech is not rated particularly well at running the ball; fortunately, the Panther's rush defense appears to be even worse, so there is some room for success. According to Football Study Hall's advanced statistical profile of Virginia Tech, Travon McMillan needs to be the Hokies' feature back. The profile credits the offensive line for the first five yards of a carry, and terms those yards beyond that "highlight yards". Travon is averaging 7 highlight yards per opportunity, compared to the next best running back — Trey Edmunds — at 6.0, and no one else is close. But the bigger reason is the ability to hit the right hole and gain opportunities for highlight yards — McMillan has an opportunity rate of 58% compared to 29% for Edmunds. The eyes of every fan watching Tech's games agrees, McMillan needs the lion's share of touches.
When Pittsburgh Has the Ball
Again, examine pass-run comparisons first:
This could turn into a long day on the ground, as the Panthers field a top 15 rushing unit against a Hokies' run defense that is ranked in the top 102.
Who To Watch Out For
- Oh my. 6-2, 230 redshirt freshman running back Qadree Ollison might make me wish James Conner was playing. Recalling the highlight yards per opportunity stat, Ollison is sitting on an incredible 12.7 yards, making him a huge home run threat.
- Elite WR Tyler Boyd is targeted on 36% of throws (next closest: 13%) and hauls in 84% of them for 9 yards per target.
- Diminutive CB Avonte Maddox has 6 pass break-ups on the season, but may struggle if he is put into one-on-one situations with Bucky hodges. Hodges has 10 inches on the 5-9 Maddox.
Statistical Key to the Game
TECH CANNOT GET GASHED ON THE GROUND. It's been discussed plenty thus far — the statistical key to the game is highlight yards gained by Pitt per opportunity. Bud must find a way to stop the home run threat the Pitt rushing game brings and force them into long drives.
Statistical Prediction
Statistically, Virginia Tech is projected to gain 241 yards passing and 171 yards rushing while Pitt is projected at 161 yards passing and 208 yards rushing, putting the Hokies at an overall yardage advantage. There is no evidence one team is more likely to turn the ball over more than the other, so I'll take the Hokies to narrowly escape and back us all (slightly) off the edge.
Virginia Tech 27, Pittsburgh 24
As always a thanks to Football Outsiders, cfbstats.com, and Minitab Statistical Software.
Comments
What a difference a week makes.
At this point, I'd be happy with a 3 point victory! Any W is a W.
Quite the nosedive after that loss. I knew we'd take a hit, but wow.
it's amazing how a couple bad calls can turn a team's season from most likely going to a bowl game to most likely missing a bowl game. That's something else.....
But preseason games shouldn't have any affect on our stats, right?! /s
We didn't lose because of officiating.
there were two drives by ECU kept alive by really bad calls from the refs. Both drives we had ECU stopped dead to rights and they got free first downs and ultimately scored without giving the ball back to our offense to try and do something. You are correct that we didn't lose as a direct result of the officiating, but it certainly didn't freakin' help.
You can say, "but for bad officiating, ECU scored two touchdowns."
This gif is SO appropriate this week...
I just gotta point this out.
You said:
And then Joel said:
The running backs are averaging 5.0 YPC.
You, sir, are a goddam genius.
I believe you meant to say French and Mason are goddam geniuses
I wonder how the team would have scored in these rankings in 2010 after losing to Boise and JMU.... Just saying, this is still a young season and the data is still relatively thin. Let's get halfway through the ACC schedule before we pronounce the end of the bowl streak.
That's how statistics work, though. The odds of us winning so many games that year were low, but not impossible. More games means more stats which refines the odds.
Funny you should ask - you can find out the answer to that very question at http://www.masseyratings.com/cf/arch/compare2010-2.htm, where it shows us as being ranked on average #36 in the country after the JMU loss. We finished the season with an average ranking of 14th.
To be fair, I haven't pronounced the end of the bowl streak...I just said it is currently more likely that we don't make a bowl than we do. Obviously every game presents new information and a more refined estimate...blow out Pitt and suddenly we're back in business. Take another ugly loss and this could be a really painful season.
First, I love this feature, Joel, even when the results are depressing, like today. A purely objective look at our prospects is a good thing.
It's too bad that you didn't do that histogram for odds of winning N games last year. I'd be really curious to know if this is the first time since 1993 that a losing season was objectively the most probable outcome. Things looked pretty dicey for awhile last year, so maybe not. It's a sobering thought, though.
edit: Grammar fail corrected
It is near the top...right below the schedule odds and "Not great, Bob".
I think we're misunderstanding each other, Joel. I was curious about previous seasons, e.g. if during our losing skid last year the expected value of the number of wins ever dropped below 6. I looked back at some of last year's BT(A)N posts and didn't see the same kinds of projections for the rest of the season. Just curious to know if projecting a losing season as the most likely outcome is new territory for us.
The more I think about it, though, I've got to think that after the BC game or after Wake last year, the projection would have been for a losing season. So, probably not new. And the good news is that with a good showing this week, the projection will probably pop right back up since the models are so fluid at this point in the season.
I see now...I don't have the numbers to run it exactly, but based on rankings and game locations I'm guessing that after the BC game we were under a 6 game expectation...maybe somewhere around 5.5-5.75 wins. After Duke would have been positive since we would have been projected to beat Wake Forest and Virginia.
Unfortunately the Wake Forest game never happened, but going into Virginia we'd have to be under 6 because we only had 5 wins and there was (very slightly) less than a 100% chance of beating UVA.
ohh my lord.. will my liver be able to handle this week?
If you take a shot every time we score, it just might survive.
If you take a shot every time Pitt scores, you may want to watch the game in a hospital so they can tend to you immediately
"Here's your room, Ms. Jugs. How was the game?"
No way we don't make a bowl game. ACC schedule is weak and its quite unlikely we will have as many things go wrong in all of our remaining games as went wrong against ECU. We couldn't catch a break.
you mean after ECU spotted us 14 points?
i haz a sad
I seriously dont think I've ever been this low on the hokies, not even during the past 3 seasons...
yeah I know...it's rough
worst part is that before the game I was talking to a friend who went to VT for grad school, not undergrad, and doesn't really keep up with CFB in general. He asked me if ECU was any good and what my expectations were for beating them. I told him that ECU always plays us tough and I don't expect it to be an easy win at all. 5 minutes later we're up 14-0 with momentum and he says to me something along the lines of "Man, I thought you said they were tough..this is a cake walk!" and I warned him that there was plenty of time left. Even up 14 early, I had very little confidence that we would win. I just knew deep down that it wasn't going to be that easy. It's really sobering when your team goes up 2 scores and you can't be confident that they'll hang on to win. 5 years ago if we went up 2 scores I would, without a doubt, confidently believe that this team would win handily. Not anymore.
fair enough. After that, we couldn't catch a break.
We seriously need said everlasting bottle of scotch...
We've managed to survive as bowl eligible thanks to a good defense keeping us in games with a poor offense.
Now we have a top 50 offense but a bad defense. We are ranked 102nd in Defense against the run.
There aren't many games that match up well for us, in our current state.
I need an antidote after digesting all of this data....Please send FOE-Rensics ASAP!!
statistical key to the game (hokie07me style): score a TD on every possession. Whenever we're down on defense, offense has to step up and force the hand of the other team. I like hodges against a 5'-9" corner all day. If they gash us on the ground, let's gash them in the air. Also, get Travon enough carries to break 100 yards and see what that does for us.
I like that matchup, though I think that match up works on a dry day. Bucky doesn't seem to have good hands for a wet football.
Hard to say about his hands. That ball he dropped on 4th down hit him right in the shoulder pads.
He has dropped a couple this year in good weather and bad now.
Beat pitt and flip this lousy script. Otherwise just light next weeks by the (advanced) numbers on fire and punt it off a bridge.
A lbit off topic, but has anyone else noticed Michigan is 22?
Weren't they unranked for longer than us?
Michigan will continue to be good this year and become elite very soon. Harbaugh is legit. Just have to wonder how long he will stay.
Merp. Popping my head up again.
Stats clearly indicate that this is a challenging match up for both teams, so lift your heads a bit. Pitt has lost to Youngstown State AND Akron in the past couple years. ECU is far an away a better program and any team that plays them can fall into a trap.
VT's biggest worry (Albeit I'm not convinced this will happen) should be if Pitt has the run game comes to life, which, by god, any team should fear. That being said, the Pitt O-Line has not shown the ability to open holes large enough for a solid positive gain (like you saw last year from Conner). Pitt has a TE-Turned Inner Lineman who is still working on run blocking and injured center. If Conner was still in the backfield VT loses, Conner can produce yards out of nothing. Ollison, the second coming of James Conner he may be, but he's freshman James Conner... not Sophomore ACC POY.
Pitt has a weakness against running QB's too... if Motley can get enough O-Line support to get beyond the first level with spread receivers he can get consistent 7-10 yard gains... Honestly I think Motley is a better QB match up against Pitt than Brewer. Pitt has a secondary that can be exploited across the middle, where slants have been particularly effective. LB's are solid players, but nothing extraordinary.
VT's biggest worry beyond Pitt's running game coming to life should be Tyler Boyd, no questions asked. With Kendall Fuller out - he shouldn't have been playing anyways from what I've seen - I don't see anyone who will be able to compare with Tyler's talent. If Pitt's QBs do an adequate job of getting the ball out and somewhat on target he'll get the ball, get yards and then some. Keep Pitt's QB under duress, taking longer passing away and you'll get opportunities for interceptions and remove Tyler's 'yards after catch' ability when he gets the ball out in space.
This is always one of my favorite games of the year, I'm excited for Saturday, rain or no rain!
Thanks for checking in and reading TKP! More in-depth film review will be coming from the site in the next couple of days. Your explanation of the soft spots on the Pitt D make me think Motley can exploit the running QB opportunities but Brewer is probably better for the quick slants. Either way it is irrelevant as Brewer will not be playing on Saturday.
Pitt beat VT the years they lost to Youngstown and Akron so there's that. Thanks.
I don't believe the data. I like this team.
I'm completely incapable of predicting a win now. I knew both teams were weak to the run, but I didn't realize the gap was so large.
This is one of those games that is bizarro. Pitt has played noone and rushing yards are skewed by the first game where they gained over 200 yds rushing but were lack luster the other two games. They have a week rushing game which means we will struggle to run against them. Pitt has had an atrocious run defense for I don't know how long and we always seem to be inept when it comes to running against them. Perhaps in the past due to the OL but this year will probably be due to rotation. And because stats show that McMillian should get the bulk of carries and trey second, JCC will start and Shane will attempt to get him the bulk of carries to make up for ECU. McMillian will get carries on jet sweeps and look impressive running the ball but will somehow still get only a few carries.
I think defense rebounds from the passing game and Voigt is not the runner that Summers is. Dont' know about the other QB's running abilities but assume from what I've read not as good as Voigt's. Also, because the run numbers are skewed for pitt and they had a less than stellar running performance against Akron they will somehow look like the #1 rushing team in the country....other than ECU of course.
All of this because you know....Pitt. I think we either win big or lose a close one.
Thanks for pointing this out...I hadn't looked at the game logs and then did have fairly weak rushing totals against Akron and Iowa. A few of the Akron rushes probably got tossed because they were garbage time but not many. Currently Iowa is #12 against the run and Akron #17 so that explains it a little bit, and Youngstown State still counts...although a weak team Pitt still put up 325 rushing yards and a 7.56 YPC which is impressive. I chalk this up to small sample size somewhat and think they'll come down soon...hopefully after tomorrow but I don't have a warm and fuzzy about that.
For what it's worth Akron held Oklahoma to 3 YPC (some of that is surely garbage time) and Iowa has held opponents to 2.65 YPC.
Neither do I. Pitt always seems to look like an elite rushing team when they play us regardless of how they have done in previous games. I've watched enough Hokie games to know that. I'm expecting at least 150 on the ground for them. Hopefully we can stop the carousel that is the rb rotation at VT and get at least one of them going.