Last year around this time, I wrote an article trying to forecast what Clark Lea’s first year would look like from a win/loss perspective. To say the mood among fans at that time was "grim" would be an understatement. The team had just lost to ETSU – the first loss to an FCS opponent in school history. Our hets were thoroughly soaked. My motivation for the piece was mostly to try to convince myself that we were not bound for another winless season – that there was some hope in what the rest of the schedule looked like. My prediction for a 1-2 win season, which turned out to be quite accurate, but the predicted margins in the wins were quite off from the "too-close-for-comfort" real life margins. So I decided to give this another try this year and made a few small changes to the methodology to try to get this more correct.

Step 1: Instead of looking at Margin of Win/Loss vs. Opponent FPI over the past 15 years together, I broke it up and calculated separate regression parameters for each individual year. Some interesting insights from this…In most years (example from 2018; left), there was a statistically significant inverse correlation between the two measures (i.e., the harder the opponent, the worse the outcome). One notable exception to this was 2016 (right), when the line was essentially flat, which confirms why that season felt like an anomaly. The blue point on each graph is where the 2022 Hawaii game would be.

Step 2: This then allowed me to then take Hawaii’s FPI (updated after Week 2) and predict how each Vanderbilt team since 2005 would have performed. There were 3 years when the predicted margin of victory was fairly close to the actual 2022 margin: 2011, 2012, and 2018. Interestingly, these were also statistically VU’s best seasons over the past 2 decades.

Step 3: Based on the slope and intercept of the regression lines, I picked three seasons and characterized them as the Best (2011), Worst (2021) and Average (2006) cases based on the regression intercept, with higher being better, lower being worse, and zero being average. Again, the blue point on each graph is where the 2022 Hawaii game would slot in.

Step 4: Using the regression line from each of these seasons and the 2022 FPI numbers (after week 1), I calculated predicted margins of win/loss for each remaining opponent in the 2022 season. From these predicted margins, I characterized each game as Win (margin greater +7 points), Loss (less than -7 points), or Toss-Up (between -7 and +7 points). The total number of wins is then the two we already have + the number of games in the "Win" category + ½ the number of games in the "Toss-Up" category. Here’s what that table looks like:

First thing to note is that this is a very hard schedule. Even a really good VU team has only one guaranteed win left. If we take Hawaii’s current FPI at face value, then we are headed for the 2011 trajectory, and Clark Lea’s bowl game dream might manifest. However, this team would have to play on par with the best VU football teams in recent history. This seems unlikely. If the team is as bad as it was last year, we will need a bit of luck to beat Northern Illinois and improve to 3 wins, which would not be great.

If Hawaii is worse than they’ve shown in the first 2 weeks, then this team may be closer to ‘average’, and the expected win total is between 4-5 with expected wins against Northern Illinois + 2/3 of Wake, the Game Penises, and the Big 12 team. This seems like the most reasonable expectation (and is my personal prediction), but the margin of win/loss in the next game should tell us a lot more about which trajectory we are on for this season, because an n=1 isn’t really ideal for this. But hey, optimism!