The bowl predictions page has been updated.

Once again, we’re six teams short, so I had to use both the transitioning schools (Delaware and Missouri State) and four 5-7 teams (in order: Northwestern, Auburn, Rice, and Central Florida). Once again, since I predict the games out using SP+, there is a potential that any team I have projected to finish 5-7 finds one more upset and qualifies normally. For example, Mississippi State is projected to lose out to Missouri and Mississippi, but they’re about a 10 point underdog to Mizzou, which is well within the realm of possibility. (And the Egg Bowl is, of course, the Egg Bowl.)

We’ll get more clarity when the newest CFP rankings are released tomorrow. The main question is if they will put in a three loss Alabama: I have the Tide losing to Oklahoma and Texas A&M in the SEC Championship Game. At some point, that loss to Florida State has to come home to roost, right? For now I have Utah as the last at-large bid to the playoff.

The other point of contention is going to be the status of the ACC champion. As the joke goes, when the ACC killed divisions (and therefore the notoriously volatile Coastal division), they just made everything Coastal. If everything goes to my projections, the Virginia will finish with one conference loss and Louisville would win the tiebreakers between the two-loss ACC teams (which in addition to the Cards I have as Miami, Georgia Tech, Pittsburgh, Southern Methodist, and Duke). This prevent the nightmare scenario of a 7-5 Duke making it to, and winning, the ACC Championship Game, in which case the ACC would probably miss the playoff entirely.

I’ve been doing bowl predictions since 1999. While there’s a lot more bowl games now (oh, and a playoff!), I still do my best to predict who goes where.

There’s essentially two phases to this. This phase is trying to figure out who goes where. At this point in the season, I am projecting each team’s record using SP+. Since most teams have 3 to 4 regular season games left, there’s a lot of wiggle room. In particular, doing it this way means that early on I’m usually short a few teams. As upsets happen and SP+ gets more in-season data, things usually even out. Right now, I need 76 teams, and I’m five short. This means I had to include FCS to FBS transition teams Delaware and Missouri State, and a well as 3 5-7 teams based on their APR.

Starting in 1999, and up until now, I had used Microsoft Expression Web (neé FrontPage). However, the software is several years old and it doesn’t run well anymore. I had also wanted some various modern web niceties (like displaying correctly on phones). So suffice it to say, the predictions have a different look this year. I’ll be continuing to tweak it, or if I can’t get it to work the way I want, I’ll just revert to the old system. For now though, enjoy.

Hotter and fresher than even a Little Ceaser’s pizza, you can get the latest predictions here.

Thanks to some upsets last weekend, I only needed two 5-7 teams this time (as opposed to four last time). With 77 teams having already clinched eligibility, that means I don’t need a lot of upsets to get there. Let’s take a quick look at every game involving a 5-6 team heading into this weekend:

• Oregon State (5-6) @ Boise State: The Beavers won’t be favored in this one, but Boise has gotten by with closer margins the past few weeks. For now, I don’t have them in.
• Minnesota @ Wisconsin (5-6): Wisconsin is a slight favorite here, but I don’t have them in.
• Kansas (5-6) @ Baylor: KU’s vastly improved play the past few weeks isn’t completely reflected in SP+ yet, so I still have them winding up 5-7. It absolutely will not shock me if they win.
• Louisiana @ Louisiana-Monroe (5-6): hard to see ULM pulling this off.
• North Texas (5-6) @ Temple: UNT is absolutely favored here and I have them in.
• Eastern Michigan (5-6) @ Western Michigan (5-6): The Michigan MAC Trophy and bowl eligibility are on the line here.
• Coastal Carolina (5-6) @ Georgia State: CCU is absolutely favored here and I have them in.
• Auburn (5-6) @ Alabama: this would be hilarious, but no.
• Rutgers @ Michigan State (5-6): SP+ has Rutgers as about a 6 point favorite, but the Spartans could absolutely win this one.
• North Carolina State (5-6) @ North Carolina: the Pack would have to spoil Mack Brown’s swan song, so I don’t see it.
• Texas Christian @ Cincinnati (5-6): SP+ has Cincy as about an 8 point underdog, so I don’t have them in. But they could absolutely win this game.
• Virginia (5-6) @ Virginia Tech (5-6): self-explanatory.
• New Mexico (5-6) @ Hawaii: SP+ has the Lobos as an underdog, so I don’t have them in, but they could absolute win this game.

By my count, that’s 5 very plausible upsets. So we could possibly have too many bowl eligible teams. It’s happened before, and a new bowl game was invented on the spot to accommodate them, so it figures that it will happen again.

We’ll know how everything falls after this weekend, which makes the predictions next week one of the most important I do.

They’re hot and ready here. Let’s actually talk about the process a little bit this week.

Step 1 is to project out each team’s record. To make it slightly less vibes-based, I use SP+ to project each team’s remaining matchups. I ignore any team with 8 or more losses, as there’s no chance for them to make a bowl game. Step 2 is to project the playoff. This was hard when it was the BCS, slightly harder when it was the 4-team playoff, and now with 12? All bets are off, but I’m trying my best. This is also why the new projections tend to come out after the new rankings, as I use the current rankings to try to gain some insight into the Committee’s thinking. Almost certainly the most controversial choice I made was putting Indiana in over Tennessee. If what everyone thinks will happen to the Hoosiers happens this weekend, next week’s ranking will be very useful to determine if they still have a shot.

Step 3 is to start filling in the bowls based on the non-playoff teams. I start with the Power 4 and the zombie Pac-12, as their tie-ins are slightly less to the whims of what ESPN decides to do. After that, I start filling in the G5 teams as needed. The conference matchups you see on the page are more or less… guidelines, and generally speaking for any ESPN Events-owned bowl just about anything can happen. This makes this process maddening, but I’ve been doing this since 1999 and I’m not going to stop now.

Step 4 is to then determine what, if any, backfill is needed. Right now the projected record leaves me 4 teams sort of the 6-6 mark. There’s no transitioning teams this season that will be bowl-eligible, so we go straight to APR ranking the projected 5-7 teams. That’s how Cincinnati, Central Florida, Virginia, and Boston College got in this week. It could well be different next week, as upsets tend to add more 6-6 teams.

As a reminder, I have no special insight into what anyone who is responsible for these games is actually thinking. I don’t really keep track of the accuracy of the projections because of this. It’s gotten harder over time, but it’s still a fun exercise.

Some quick sports-columnist-esque notes on the matchups:

• Boise State’s best “win” is almost certainly going to be their 3-point less to Oregon back in September.
• I don’t really love that 8 of the 12 spots are going to the Big Ten and SEC, and I love even less the change those conferences are shoving through to give themselves 4 autobids regardless. I don’t think the ACC and Big 12 are necessarily doomed to only get one team in each season (and I definitely think the Big 12 will be better next season).
• The specific order of the CFP First Round matchups is more-or-less a guess, but I would guess if Oregon gets a late game, they will get one of the later slots. The main thing to watch is how the subcontracting with TNT goes, i.e., if ESPN gets to decide what to keep (because in that case, a Notre Dame home game would almost certainly be on ESPN).

That’s about it for now. Next week will be even more robust as more teams clinch eligibility and I only have to project out one game.

Welcome back to an annual tradition mine that’s now in its 25th year. Wow, that feels weird to say. I don’t have a lot of time to explain this at the moment, so let’s cut to the chase: I try to predict who will go where in college football’s postseason. I use SP+ to project each team’s record, and then I use those records to try to predict where these teams will go.

Expect a more detailed post next week. For now, the predictions are here. I’m two teams short currently, but that means I don’t need very many upsets to get enough teams.