UNC Update
The weather forecast has changed, drastically. Here are a few extra nuggets to consider.
What a difference 12 hours makes! Now that the forecast has cleared up, I thought it might be helpful to put this matchup in some historical context.
ELO is a rating system originally created to rank chess players. I have included an ELO trend graph on my data visualizations all season because I think it is a really helpful way of quantifying how good a team has been over time.
ELO reflects past performance, and past is not always prolog, but it usually gives a pretty good indication of a game will play out. Virginia Tech’s current ELO rating of 1388 is 171 points lower than North Carolina’s. Here is a look at Virginia Tech’s point differential compared to ELO differential since 2002 for regular season games.
When regressing VT Opp ELO Diff on VT Score Diff, we get an R^2 of 0.2183, meaning the ELO difference explains 21.83% of the score differential. There is a clear positive association between the two factors, but the strength of the association is only moderate.
The following table lists all games since 2002 in which Virginia Tech entered with an ELO disadvantage of between -140 and -200:
The game most similar to today’s was Virginia Tech’s October 5, 2019 matchup with Miami. The ELO difference was -174, just three points more than today’s game. Virginia Tech’s ELO entering that game, 1418, was 30 points higher than its current mark, so the 2019 team was on a very similar rating level to today’s team. In 2019, the Hokies were coming off the humiliating home defeat to Duke that signaled the beginning of the end of the Justin Fuente era. Miami had a hot young quarterback in Jarren Williams. The weather was wet and muggy. The Hokies, shockingly, raced out to a huge lead in that Miami game, then hung on at the very end to win by a touchdown.
Two other comparable games played away from Lane Stadium were the 2004 matchup with USC (-177) and the 2007 game at LSU (-191). Both of those teams went on to win the National Championship. UNC obviously is not a title contender in 2022, and the fact that Virginia Tech enters this game at a similar ELO disadvantage as those high profile matchups from more than 15 years ago shows just how far the Hokies have fallen since the peak Beamer years.
A few other notes. Virginia Tech was at a -190 ELO disadvantage entering last year’s game against UNC, but still won by a touchdown. Two other home games provide good examples of why the ELO difference has only a moderate correlation to VT point differential.
2017 vs. Clemson (-192), Tigers won 31-17
2019 vs. Wake Forest (-195), Hokies won 36-17
Remember, ELO is a rating system that looks only at past performance. It is moderately correlated to future score differentials because it does not take into account many factors that can influence the outcome of a game, e.g., weather, key injuries, motivation, suspensions, home vs. away, etc. Put simply, ELO is great metric for quantifying how “good” a team is at a given point in time.
Entering today’s game, North Carolina is better than Tech by a margin similar to 2004 USC, 2007 LSU, and 2017 Clemson, but also other teams like 2019 Miami, 2019 Wake Forest, and 2021 North Carolina, which all faded after losing to the Hokies.
I wrote in yesterday’s article that under clear skies, UNC was likely to win big. I put the margin at 38-17. I’ll stick by that, but I’ll throw in the caveat that this UNC team appears to be teetering. A VT win could be the start of another Tar Heel fade to 6-6 and a low-tier bowl.