If you were hoping the arrival of Justin Fuente would suddenly wash away the recent ills of Virginia Tech football, you have to be feeling great right now. The Hokies' offense is scoring at rates not seen in many years, and that hasn't been to the detriment of the defense. And even injured players who could never quite get healthy like Brandon Facyson and Marshawn Williams have returned to form as important contributors.
It's times like these I like to splash a little cold water on myself by remembering that the data is emotionless and may help keep me grounded in reality. As is often the case, when things seem great they're usually not as great as the seem (Ohio State 2014), and when things seem terrible they're usually not as terrible as they seem (James Madison 2010).
Where the Season Stands
Virginia Tech's record is currently 3-1.The actual and predicted score difference and odds of winning each game are:
Football Study Hall's stat profile shows Virginia Tech as favored in every remaining game. However, the chart above is a composite of multiple computer systems that combined suggest the Hokies are still underdogs in two matchups including the Tar Heels.
The odds of each possible regular season win total are now:
If you're a berther, seeing the chance of an 11-1 regular season escalate to a realistic, if improbable, 2.7% is a huge boost of confidence. After all, as things stand now it appears a one-loss ACC champion will in fact get a playoff berth. At the very least, a continuation of the bowl streak is all but a certainty.
Realistically, however, the most likely outcome is an 8- or 9-win season, and even that is still based on a fairly small sample size of games.
Rankings and Computer Predictions
The computer rankings and predictions for each team:
As expected, the computers only give UNC a slight edge in the rankings and predicted outcome. The odds of a 2.5-point underdog winning are 42.5%.
Next is a look at any overall offensive or defensive advantages:
Oddly enough, the Virginia Tech offense is still not highly-rated despite the outburst of points over the last two weeks. It is difficult to say exactly why, but consider that Virginia Tech only scored 36 on Liberty, a team that also gave up 29 to SMU and 48 to Jacksonville State. Furthermore, hanging 54 on ECU isn't particularly impressive considering UCF scored 47 the next week.
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 mediocre offense rating is clearly due to a mostly unimpressive run game (Evans' impressive 55-yd TD notwithstanding), but the passing game is unlike anything seen in years in Blacksburg. Additionally, the Hokies are doing very well on passing downs, an indication that the offense is executing well even when the defense knows what is coming.
When North Carolina Has the Ball
Again, examine pass-run comparisons first:
North Carolina's offense has played exceptionally well in both phases of the game, and on both standard and passing downs. However, they'll face a Bud Foster defense that has been formidable as well. Tech ranks No. 2 in the country against the pass. One would expect the two to meet somewhere in the middle.
Who To Watch Out For
Marquise Williams is no longer the Tar Heels' quarterback, so that must mean a step back at quarterback, right? RIGHT?:
- Junior QB Mitch Trubisky owns the No. 6 passer rating in the country thanks to 13 TD's, 0 INT's, and a FBS-leading 76% completion percentage (the Hokie defense is the 2nd-stingiest at 41.3%).
- Senior WR Ryan Switzer is the ball-hog of the receiving core with 47 receptions (4th nationally).
- Kick returner T.J. Logan (also the most productive RB thus far) is 23rd in the country with a 27.21 yard-per-return average, but who cares since Joey Slye is first in the country with a 65.48 yard kickoff average.
Statistical Key to the Game
The statistical ratings provide concern that the recent success of Tech's offense may be just an illusion, waiting to be exposed as nothing special. The Tar Heel defense is below average so the Hokies shouldn't get completely shut down, but this game could be really telling. My key to the game is yards per play gained on offense — anything under 5 and I think Evans and company have been exposed and the Hokies lose the game.
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 |
I will not value time of possession.
Statistical Prediction
I predicted easy wins the last two weeks and the team certainly delivered, so blame me if reality comes crashing down this weekend.
Not that I'd ever predict a loss or anything...
Virginia Tech 31, North Carolina 27
As always a thanks to Football Outsiders, cfbstats.com, and Minitab Statistical Software.
Comments
As usual, great write up.
Numbers, numbers,ah, but what is impossible to calculate is the emotions, environment, lack of focus, and Bud's special plans. All things I pray that will vex UNC!,
...only if those emotions, environment, lack of focus, and Bud's plan are unique to this game!
This is like the classic "but you can't measure heart!" argument against using statistics to describe performance. Um, if heart is improving performance then you ARE measuring the effect of it - it's just all wrapped up in ability and planning and everything else.
Ah, the old debate between systematic and idosyncratic factors. This week probably is less statistically predictable than most, considering the FSU win, the ESPN attention, and Hurricane Mathew.
I still prefer the "Football Insiders FPI based ranking." /s
Question from a berther:
What is the difference between the Off Passing and Off Pass Downs? Is the first for passing on any down, and the last for performance on say 3rd & long?
Yes, but I have no idea what metrics are used to determine what is a clear passing down and what is just considered OFF Passing. Also, I always assumed that Pass Downs are included in Off Passing numbers.
http://www.footballoutsiders.com/stats/ncaaoff
Basically, think of it as "what happens when they pass?" vs. "what happens when they have to pass?"
Thanks for answering the question! This is correct. Of course they don't HAVE TO pass, but they are in a situation where a pass is much more likely. Passing Downs measures their performance on those down regardless of whether the team actually passed or not.
Thanks for the clarification. I did consider similar wording, but "downs in which conventional wisdom indicates that passing would be more beneficial" wasn't quite as nifty of a reply.
This offense has been amazing so far. Every game there has been small struggles here-and-there that they have been able to correct in game. You have to figure at some point they will really struggle, as all offenses do. How the team responds and manages to claw out a win will be paramount.
I would argue that our offense really struggled against Tenn
I'd agree that they had severe penalty and fumbling issues that could not be corrected in-game, but 18 non-penalty first downs, 394 total yards, and 3-for-3 in the red-zone isn't truly struggling to me. Throw in 4 drive-killing turnovers and a muffed punt and I think our offensive production was fairly decent.
I mean when our offense truly struggles we fumble the ball 4 times but still manage to gain nearly 400 yards and score 24 points. That's not a bad thing. I honestly don't expect this offense to struggle more against any of our remaining opponents than they did vs Tenn. I wouldn't have said that before the BC game but after seeing what our offense can do when they are focused and disciplined I don't worry as much about offensive struggles.
A very large chuck of that came in the first quarter though. The second and third quarter we struggled a lot.
The stat profile at http://www.footballstudyhall.com/pages/2016-virginia-tech-advanced-stati... estimates the offensive performance against Tennessee as the 43rd percentile among all game by all teams this season. I think the 19th-percentile game against Liberty is hurting.
touche...can't argue with stats....I'd like to think we can get back into the 80s or 90s..those first two games, hopefully, are the exception and not the rule
It seems reasonable that the game against Liberty was intentionally not high-performing. Why show much of your playbook before a bigger matchup if you don't have to?
Time will tell, however, whether the team can keep performing in the 80's or 90's, or if the typical performance will be less than that and we just happened to see back-to-back outliers.
And the Liberty game also had fumbles. Lest we forget.
Math is giving us 9 wins in the regular season, I will take it!
Should be a dandy
I'm predicting 41-31 for the turkey birds!
45-17 VT
Hey, 2 of the last 3 games we held them to 17, seems to be a theme.
Right on.
44-34 VT
38-24 Hokies. I think we score a little less than we would in dry conditions. UNC will score some points, but it'll feel like they got shut down because Trubisky will be on his back all day
27-21 Hokies. I still think Bud made the changes he did to compete against this exact style of offense. He'll be ready. The offense will score enough to win.
34 - 17 VT. Same thing I predicted for ECU and it worked that time so why change now. Would not be too surprised if it got messy on both sides though and lowered the scores a bit.
Not that I'd ever predict a loss or anything...
Virginia Tech 31, North Carolina 27
We Have a Winner ! It's science !
A smart man would hold his bet until he sees the line from Halfwits & Wagers!
Then max out the ATM and lay it on the line!
via GIPHY" target="_blank">
via GIPHY
How would you explain the difference in the spread from opening to present? I think it opened around -6.5 and is now hovering around -2 (factoring -3 for home field advantage seems like we might actually be favored come Saturday). Then on top of that, ESPN still lists UNC at roughly 65% in their matchup predictor. Something doesn't add up.
my understanding of the spread is that it's driven by bets. If too many people bet that UNC will fail to cover the spread, they reduce the spread. So it's very likely that something like the spread, which is based on gamblers' feel for the game and an "expert" prediction will not jive.
Pretty much correct...there's no new information available that would change the algorithm-driven prediction on ESPN. The spread on the other hand, is moved by the casinos to get as close to 50% of the money on either side as possible. So the spread moved heavily in our favor because too much money was being bet on the Hokies and they need to entice more bettors to take the Tarheels.
Professional casinos and gamblers tend to have some of the best experts though.
The spread is based on Vegas. The matchup predictor is based on ESPN's power rankings. Totally different things.
I get that they're totally different. So what you're saying is the matchup is just one algorithm based on one power ranking that doesn't take anything else into account? Sounds about right for ESPN.
pretty much
Could smart money be thinking a game played in a hurricane would be lower scoring and closer?
Or would favor the team that runs the ball better instead of the team that relies more on passing?
We have to assume that some of these games we are going to be in for a fight. While I would love for us to cruise to victory against UNC this has the looking of a fight. However, looking at offenses v defenses we have an advantage against their defense and at worst we are a push against their offense.
UNC has been in two dogfights the past weeks so they know the feeling and know they can come out ahead. Tenn was an anomaly for us so we really haven't seen our ability to fight back after being down.
I think we get the win here 35 - 21. It will be pretty close for the most of the game but the Hokies pull away in the late 3rd qtr and UNC isn't able to mount the comeback.
Does Matthew have any impact on the air attack for both teams this weekend?
I don't think the New Testament discussed this game specifically.
Matthew 1:1 Thou shall whoop dat tarheel ass and showeth no mercy.
must be the message "translation"
Little known fact: "tarheels" is actually transliterated from the Hebrew word "tarhelim", which means "cheaters".
Getcha popcorn ready!
Joel - who are the teams we most "look" like right now? I always enjoy those.
I'm no rocket surgeon, but based on the data I calculate the final score to be 77-6 favoring VT. I know what you're going to say...and no...I have no idea how they miss an extra point.
Sadly we don't get those stats until after week 7. So off the top of my head I'll just roll with Louisville on offense and Alabama on defense.
lolz
Also, there are two likely explanations:
1) Two field goals, or much more likely...
2) Tim Settle blocked extra point duh
duh
Both analyses: excellent comparisons here, or -to really blow your mind: maybe Louisville's offense is playing at a Justin Fuente's Virginia Tech offensive level and Alabama has finally figured out the Bud Foster Virginia Tech defense?
(explosions)
Did you see the game last week in which, in the first few minutes, three players carried the ball into the end zone, resulting in a score of 9-6?
A note on why S&P+ still doesn't like our offense that much (besides a couple of weak opponents):
It's still filtering out preseason predictions. I believe it's not until we're 7 games in that rankings are based solely on this season's play. If I remember correctly, we were projected to be something in the 80s-90s* in S&P+ offense, which should continue to drag our ranking down for the next few weeks.
* One weakness of S&P+ I've seen Bill Connelly note is that it doesn't account for coaching changes very well, so we were basically projected as if we still had Loeffler at coordinator.
All true, but he does wind down how much preseason projection matters and I believe before this past weekend it was already down to 30%. I think our offense was spectacular against BC and very very good against ECU, but only pretty good against Tennessee and not very good against Liberty. Of course that could be because we didn't need to be, and time will tell if that's the case.
well I do think that being the first ever game in that system for every single player on the field might have had something to do with it...a lot of coaches say that their teams improve the most from week 1-2 but I think ours improved the most from 2-3. 19th percentile vs Liberty, then a jump of 24 percentiles to 43 against Tenn and then a huge leap to 95th vs BC before settling at 88 against ECU. I would imagine that we can expect roughly 80th percentile for most of the remaining games we have. I chalk up the lower percentiles in the early going to growing pains.
I'll be honest, I expected this team to struggle with growing pains and inconsistencies but manage to improve over the course of the year....I have been blown away by how good our offense has looked after the Tenn game. And actually, watching it live, I thought our offense looked pretty dang dangerous against Tenn, if it weren't for some bad mistakes (penalties, fumbles). This offense is way further ahead of where I expected it to be but for some reason I just can't bring myself back down to earth...I'm expecting this offense to be absolutely dominating teams come November.
We also should consider that the first string offense has only played 3 quarters in all but one game this season. I have to assume that this has a downward pull on the stats.
Not much if any...S&P only uses plays from the competitive portion of the game. So if you're up by 30 going into the fourth quarter what you do from that point on doesn't count for anything.
This may be the single most important factor in "advanced" stats for college football. Along with strength-of-schedule, but the two go hand-in-hand.
is there any statistical advantage in coming off a bye week? And do the computers adjust for that?
I'm not a computer, but I think in the Hokies' case, the BYE week is a disadvantage.
It's a question of rest vs. rust. If the team is reeling from some bad games, and some guys are nicked up, then the BYE week serves as a needed re-up, and timeout. However after ECU, the Hokies were like a hot Malcolm Delaney raining down threes. What do you do with a shooter on a hot streak? Feed him that rock.
Don't know if it's a killer disadvantage, but I wouldn't be surprised to see the Hokies start slow out of the gate Saturday.
Personally, I preferred the Zabian Dowdell red hotness
Yes sir.
Not only could he shoot it, he was cold blooded. He shot A LOT of daggers.
While I would normally agree with you in that you never want to stop positive momentum, I feel like having a new system is an exception. Obviously most of the system gets put into place prior to the season, but the bye being this early gave another week for install for the upcoming coastal slate. I think it will be a positive and we may see more wrinkles tomorrow because of it.
Good point. Just hope they've stayed hungry.
Didn't think this needed separate topic and I wasn't sure which post to comment on. Mods feel free to move it if necessary.
Watching BC and Clemson. At the beginning of the half time show they were showing the ACC top 25 match ups. They didn't say a word about us, of course, but basically had a collective O-face over the UNC quarterback. I love it when Hokies are overlooked. I have a good feeling about this one.
Also, didn't know we were winners of three straight against ranked opponents on the road. Looks like that was Ohio State and Duke in 2014 and Miami in 2013. Here's to keeping yet another streak alive.
Yep..I learned that, and that Mack Brown is really bad in the booth.
Absolutely. Oh my Lord he's bad. His voice reminds me of a slightly more refined Donnie Baker...