The bowls are a different beast.

With the advent of the College Football Playoff, transfer portal and opt outs becoming more and more frequent, games have really become a combination of glorified exhibitions or a chance for coaches to play younger players and get them experience for the next season.

The chart below shows how bowl games impact our results. Now we’ve been inconsistent in terms of the spreads we use to measure the spread results (this year and last year we are using midweek Vegas line numbers) and they have noticeably dropped our overall number against the line. However, I would argue the overall accuracy has improved.

Even the actual win probability for bowl games based on point spreads are different for bowl games vs regular season games. It’s not overly significant, but close to half a percentage point difference in the probability.

During the regular season, we benefit in picking winners straight up with all of the Power five vs Group of Five games and in bowl season, the games are generally more even matched, but still, almost everywhere across the board, our numbers are poorer for bowl games vs regular season games.

It’s a major reason that although I will post final ratings based on the outcome of bowl games, outside of the three CFP games, the results of those games will not figure into my 2023 pre-season ratings.

After the bowl games, I will post the CFP title game pick, but also what I believe are really insightful looks into the 2022 Boston College season, focused on reasons why BC was always more likely to be a sub .500 team this season.

Very pleased with our overall results vs other computer systems as captured in the Prediction Tracker. If you scroll to the very bottom, you will see those. The Book and New Book are dead even, tied for 10th, which gives me great hope that I can make even marginal changes into how the New Book is calculated and get a bit higher.

Enjoy the games and Go Eagles!