ODU Week: New Beginnings
Changes abound, and positive momentum is palpable, if tenuous
Finally, the longer-than-anticipated offseason has come to a close. Like Opening Day in baseball, the first full weekend of the college football season arrives with so much promise.
While there is a lot to analyze on the field, I want begin this column with some news off it.
Elevate or relegate
Deion Sanders, when he took the Colorado job, talked about how, in coaching as in every other job, one eventually reaches an inflection point: elevate or terminate.
I have meditated on that idea in the months since that televised profile. I came back to it again as Hokie Analytics marked its first anniversary.
Also floating in the back of my mind was Will Stewart’s experience in the early days of Techsideline (then Hokie Central). Here I was, 26 years later, facing a similar situation.
Cranking out real, insightful, and imaginative analyses takes time. One year in, I know the ropes of the Substack platform and have developed a digital library of college football data, code, and visualizations.
The time had come to elevate the publication, or terminate it.
I chose elevation.
You have probably noticed the updated branding. Logos, watermarks, a few tweaks in the look and feel of the site. It was long overdue, and I am pretty satisfied with how it has turned out.
One change you may not have noticed if, like the vast majority of readers, you consume Hokie Analytics articles via email or the Substack app, is the new domain: hokieanalytics.com.
Plenty of Substack writers manage with the sub-domain. Some even have massive subscriber lists (measured in the hundreds of thousands).
It irked me, though. Felt unprofessional. It also enabled Elon Musk to throttle down anything shared on X, so long-term, it was a barrier to growth.
Other changes will be rolled out in the coming days and weeks. I am also planning to publish more frequently, especially during the season.
Concurrently, I do have plans to roll out a paid subscription tier. The free written content will continue similar to before, with the paid option offering increased, and more diverse, content.
Importantly, I want to stress that I do not view Hokie Analytics as a direct competitor to any other current Tech-focused website or podcast. Rather, I see it as complimentary.
My aim with Hokie Analytics, from an in-season coverage perspective, is to bring the stories hidden in numbers to life, shedding light on strategic decisions and offering data-driven predictions of future games.
In the offseason, I want to continue to shed light on the history of Virginia Tech football. The recent article on the Southern Conference split remains a personal favorite, and I want to do more like it.
I have built a career out of developing analytics products to support C-suite level decision makers. Hokie Analytics is no different. You, the reader, are the executive, and I want to help you understand, whether in the past, present, or future tense the likely outcomes as well as indicators that may signal a diversion from the expected.
It’s been a fun ride so far, and now it is onward and upward.
The matchup with Old Dominion
Last year’s game sticks in the craw of every Hokie fan, but that result, and indeed the 2022 season results, tell us very little about what is likely to transpire inside Lane Stadium on Saturday night.
Both teams have seen significant roster turnover, as well as key additions to their offensive coaching staffs. That is always the issue with trying to analyze week 1 games - there are no data available!
The temptation is to extrapolate from 2022 data, but those data are generally not applicable, with the exception of prediction models that look at trends over hundreds or thousands of games stretching back years.
So, yes, I included lots of 2022 data on the preview viz, but more so as a reference point than to actually inform where things might go this week.
The most valid numbers on the viz are the predicted point totals: 34 for Virginia Tech and 15 for Old Dominion.
When forecasting a home team victory, the model is correct nearly 80% of the time. Is there reason to believe the ODU game will be the one in five games in which the model gets it wrong? Nick Gallo injury aside, no.
Still, the last time the Hokies scored at least 30 points in a game (November 14, 2021 vs. Duke), Justin Fuente was the head coach. So, let’s examine the factors that indicate the model is correct and then delve into hidden variables that may signal a different outcome.
Key performance indicators
There was a segment during the Hokie Hangover Podcast preview of this week’s game in which the guys discussed an X (Twitter) poll Ricky LaBlue ran this week.
On the podcast, the discussion centered around a couple of comments about their “Other” votes, disclosing that they would have voted for Da’Quan Felton and/or Jaylin Lane. Full disclosure, I was one of the two.
Why?
Focus on how the question was worded. Which transfer will make the “biggest impact”? The best player does not always make the biggest impact.
I agree that Ali Jennings is, at this point, likely the best all-around receiver of the three transfers. However, let’s imagine a world in which Felton and Lane did not transfer in to Tech.
If the receiver room comprised Jennings, holdovers from last year, and true freshmen, would we be justified in expecting a significant increase in production from WR1 over what Kaleb Smith produced in 2022?
Opposing defenses would double team Jennings all day long and force Tech to either throw into coverage or hope someone else steps up. The net impact on the team would likely be positive (Jennings is an NFL prospect, Smith was not), but limited.
The Hokie offense struggled mightily in two areas last year:
The short-to-medium passing game
The offensive line
Everything else stemmed from those two issues.
Felton and Lane were strategic acquisitions, meant to turn #1 above into a team strength. For full details, I encourage you to revisit this article from the offseason:
In short, their impact is likely to stretch beyond their own individual performances. A better short-to-medium range passing game means less help over the top against Jennings. It will also force defenses to play in nickel packages more often, which could open up the run game. It’s amazing the possibilities when an offense can complete 10-yard passes with regularity.
Likewise, Ron Crook appears to be much more aligned with Vance Vice’s offensive line philosophies than was Joe Rudolph. As we saw in Dublin last Saturday, Rudolph did not come to Blacksburg and magically forget how to coach. It was just a poor fit between players and scheme (and perhaps line scheme within the broader offensive scheme).
Crook figures to be a major improvement from a fit perspective. If you missed this article from shortly after he was hired, I would definitely recommend reading it in order to gain a sense Crook’s strengths and what he is likely to bring to the offense:
On Saturday night, these are the two KPIs to watch most closely. If, in the first half, the Hokies are getting a good push up front, RPO bubble screens are popping for solid gains, and Tech is able to convert a few 3rd and 7s, we can feel good about where the game, and the season, is headed.
Hidden variables
Ask any coach about keys to a game and they will undoubtedly answer penalties and turnovers. Variables yes, hidden no.
For this segment, I want to focus on hard to quantify variables that nonetheless impact the game; the sorts of things that figure into the proverbial “eye test”, but are not tracked in datasets.
Player Health
Coaches don’t talk openly about injuries anymore. The public is lucky if a team discloses whether or not the injury is to the upper or lower body.
Outside of Nick Gallo, things have been quiet on the injury front for the Hokies. Or so we think. Nagging minor injuries are normal coming out of camp and can impact snap counts. Keep this in mind if you see a nominal third stringer getting more playing time than expected.
Jitters
Anecdotally, I cannot tell you how many times I have seen the Hokies come out way too amped up for a big game. ODU is not a major opponent, but an opening week night game at Lane Stadium creates the potential for early mental lapses (e.g., blown coverages) and over-running misdirection plays.
Make a mental note of how many instances of the above you see in the first quarter. Less is more in terms of potential to win the game, and win comfortably.
Strength and Conditioning
Heat won’t be a problem, so the focus here is mostly on the strength part. Heights and weights tell only so much of the story. Is the weight good or bad? Ten extra pounds of muscle on a linebacker can be great, whereas it can slow a receiver and hurt his ability to generate separation.
Most players in the two-deep have now had two years in the strength and conditioning program led by Dwight Galt. Andy Bitter of Techsideline had a story just yesterday talking about how one of the inbound transfers, Derrick Canteen, has had a positive impact on how Mansoor Delane approaches his diet and film study.
The average television viewer should notice a real difference in the strength, quickness, and speed of the Virginia Tech players in comparison to Old Dominion’s. Another year of slow players with “skinny legs” would signal another year of five to six wins, at best.
Final thoughts
I am, as a matter of practice, always careful to discern between what the data can actually tell us, and what is otherwise unknown.
It’s week 1 and there is very little in the way of applicable data on either team. So, in closing, I will again point to the 34-15 score prediction, from the new, validated Hokie Analytics score prediction model. That, alongside the key takeaway from this recent article - an average Scot Loeffler-era level of offensive proficiency is likely sufficient to ensure bowl eligibility this year - portends good things for the Hokies as they kickoff the season against Old Dominion.
