We’ve added bowl projections to Movelor’s outputs, and you can find those here. Forecasting bowls is an inexact science, but here’s how we’re approaching it:
High-Level
At the highest level, we’re looking at each team’s average season in Movelor’s 10,000 weekly simulations and comparing them against one another. Teams are stacked in order of average season. This creates some problems. Entering Week 1, only one team is more than 50% likely to make the playoff. Entering Week 1, only one team is more than 50% likely to win its conference. To address those, we designate the four likeliest playoff teams as the four playoff teams in our projections (even if some of those teams’ playoff probabilities have especially inverse relationships, meaning they play one another head-to-head and/or are in the same conference), and we designate the likeliest champion in each conference as the conference champion.
(Update: As of Week 6, we’re using best average CFP ranking rather than highest playoff probability as our determiner of the four playoff teams. Update to the update: As of Week 7, we’re back to using the highest probabilities.)
Determining Bowl Eligibility
Anyone who’s projected to win 5.5 games or more in their average season is bowl-eligible in our approach’s eyes. We round to the nearest whole win, because 5.6-win teams are, generally, going to be 6–6 more often than they’re 5–7. On the 5.5-win issue, we round up.
One caveat here is that if we end up with too many bowl-eligible teams through this approach, we cull the list by eliminating those with the fewest average wins. This isn’t how it will really work, and as the season goes on, we will adjust to circumstance, but it seems like the option most likely to give us an accurate projection. Should we have too few bowl-eligible teams, we will consult the respective APR numbers for teams projected to finish with five wins.
(Update: As of Week 3, we are now going in order of bowl eligibility probability rather than average wins. This seems like a more accurate way to approach the issue. We are not yet considering APR.)
(Update: As of Week 9, we are aware of the NCAA rule which says FBS transitioning teams with six or more wins can fill bowl slots before 5-win teams with the highest APR.)
(Update: As of Week 13, we are now using each team’s most probable final record, and not their average final record, because for most teams there is only one game left. We are then filling open slots with Navy (projected to finish 5–6 but favored to beat Army) followed by the projected 5-win teams with the highest APR.)
New Year’s Six Bowls
Something we can do objectively is line teams up based on their average final College Football Playoff ranking, so we do that to determine the New Year’s Six matchups once the Orange Bowl and the CFP Semifinals are set and we have all Power Five champions accounted for (the Orange Bowl is the only NY6 bowl this year with conference ties).
Bowl Ties
Again, this whole thing is an inexact science. There are lists of which conferences have ties to which bowls, but very few conferences have a direct, linear lineup of which team goes to which bowl, and even there, there are ties. We do what we can to follow any relevant precedent, but ultimately, this process involves a lot of guessing. We’re mostly doing this manually. It’s something we’d like to study further, but right now, we’re mostly throwing teams into possible bowls and avoiding rematches of regular season games.
Help!
To this last point: If you know things about bowl projections, feel free to send them our way. This is an area where we’re going to learn on the fly. Our ultimate guess is that the first thing fans care about is the likelihood of their team’s bowl eligibility and that the second thing fans care about is which bowls their team could go to. To the first end: Bowl eligibility probability is included in Movelor’s outputs. To the second end: We do think these are bowls your team *can* go to. If this ends up being incorrect, our apologies, and please don’t hesitate to correct us.
