Projecting the College Football Playoff with 3 Saturdays remaining

Each week, when the College Football Playoff committee releases its rankings, my reaction is basically the same: say, “That’s interesting,” to something that surprises me (in this week’s: Mississippi State falling only to fourth), then decide it’s a pretty decent set of rankings.

Then Jeff Long opens his mouth to explain the rankings, and things fall apart.

Even the best set of rankings in the world mashes together 100 criteria, and if Team A is ahead of Team B, it’s because 51 of the 100 were in its favor. “But Team B beat Team A!” Indeed. “But Team B’s strength of schedule is better, and they beat Team C by more!” Certainly. But 51 other criteria said Team A was better.

I don’t know if committee chairman Long is good or bad at explaining things. I just know that explaining rankings at all is impossible. For Long to have to go on television and give a specific reason why Alabama jumped up to No. 1 in the rankings (they “controlled the game” against Mississippi State), or why Mississippi State dropped only to fourth (“You never felt they were out of it”), or why TCU is ahead of Baylor (“The committee doesn’t feel Baylor and TCU’s body of work is comparable enough that [Baylor’s head-to-head win] would kick in”) is unfair and hard to do without some level of contradiction.

There is more than one reason for any specific ranking, but Long is asked to stand in front of a camera and say “eye test” (sorry, “complete team”) or “résumé” or “top 25 wins” or “strength of schedule” or “game control*.” And the result is that no one is satisfied, and Baylor fans are crying conspiracy because Minnesota remains at No. 25, therefore giving Ohio State and TCU another “top 25 win.”

Nobody’s happy except Alabama. I guess that means the system is working properly.

* It’s incredible the lengths to which Long and the committee are going to measure things that computer ratings measure quite well while shunning computer ratings.

Here’s a closer look at a generous list of teams still in the running:

Barring incredible chaos -- Auburn or the SEC East champion beats Alabama, Oregon State beats Oregon, the ACC Coastal champion beats Florida State, Ole Miss beats Mississippi State, Texas beats TCU, the Big Ten West champion beats Ohio State, and Kansas State beats Baylor -- we know how this will play out: if the top three win out, then the next four are jockeying for one spot.

Here are the probabilities of each of the top 12 teams finishing with a given regular season record:

Of the seven primary Playoff candidates, only Mississippi State has poor odds of winning out, because of the Egg Bowl trip to Oxford in Week 14. For everybody else, the table is set. And those who have conference championship games will be rather heavily favored in those, too.

Throughout the rest of Wednesday, I will be posting projection pieces specific to conference races at Football Study Hall. Since I have already culled the data for all of those, we can look not only at the regular season probabilities, but the conference championship probabilities as well.

Here’s a list of Playoff contenders and their odds of winning out through 15 weeks:

The Big 12 has no title game (for now!), so the odds we see above are all that matters. That said, if the Frogs perform in Austin like they performed in Lawrence, their chances of beating Texas fall close to zero percent. Teams lay eggs, and TCU survived one. Better not do it again.

My thoughts on the TCU-Baylor debate, by the way? I completely understand why Baylor fans would be mad, and Long’s explanation did nothing to tamp down outrage, but TCU is ahead of Baylor in the F/+ rankings, the Massey Composite, the AP and Coaches polls, et cetera. If nothing else, that tells you that TCU has probably been better than Baylor overall in the games that weren’t Baylor’s head-to-head win. That said, the gap was pretty large two weeks ago, and Baylor’s blowout win over Oklahoma and TCU’s near-miss at Kansas closed it almost completely.

The Bears host Oklahoma State and Kansas State and head to Arlington for a neutral-site game against Texas Tech. KSU is always dangerous, and OSU is athletic enough to throw a kink into Baylor’s plays if the Bears indeed lay an egg. But that’s a pretty safe slate overall, and Baylor’s close enough to TCU now that if the Bears simply look better than the Frogs down the stretch, they could catch them.

The Ducks aren’t home free. While the regular season finishes up with Colorado and Oregon State, they will still have to face a top-30 team in the Pac-12 Championship. UCLA has about a 52 percent chance of winning the South division at this point. Oregon has about a 79.5 percent chance of winning the title game.

The Buckeyes are almost certainly going to handle Indiana and Michigan at home to finish 11-1, that much we know. But a battle against likely West champion Wisconsin is looking tougher and tougher, and Ohio State’s chances of winning the Big Ten title game are only currently around 73 percent. That means they’ve got a greater than two-in-three chance of winning out, but Wisconsin could present a pretty big hurdle.

Here’s your reminder that odds dissipate. If a team is given a 90 percent chance to win each of 10 games, the chances are good that it will finish only 9-1, right? Florida State has three likely wins left on the schedule: Boston College (87 percent), Florida (92 percent) and the ACC Coastal champion (74 percent), which will still probably be Duke. Those are very good chances, especially considering FSU’s “make every play you need to make (and nothing more)” propensity. But from a pure odds perspective, the Seminoles barely have a 50 percent chance of winning all three.

(FSU could also lose once and remain in the title conversation, but considering the committee has bumped the ‘Noles down to third despite the undefeated record, it’s safe to say they would fall to the bottom of the one-loss pack.)

And here’s another reminder of dissipation. Auburn’s tough couple of weeks have increased Alabama’s Iron Bowl win probability to 76.1 percent. That’s good! And so are the Tide’s chances against either of two potential East champions: Georgia (the Dawgs have about a 78 percent chance of winning the East, and Alabama would have a 71 percent chance of beating them) or Missouri (Mizzou has a 22 percent chance of winning the East, and Alabama would have a 93 percent chance of winning that game).

But winning both will still be a chore for the Tide, even though they’re the No. 1 team in the country according to both the committee and the computers.

If Alabama loses to Auburn, then both Ole Miss and Mississippi State are still alive in the SEC race. That would be a good and bad thing for MSU. The Bulldogs would have a 59 percent chance of beating Georgia or an 84 percent chance of beating Missouri in Atlanta, but their odds of winning out are less than one-in-three already.

So six teams have a better than 50 percent chance of winning out. That’s a bit of a nightmare scenario for the committee, isn’t it? Yes, but here’s another reminder of how odds work: there’s only a 12 percent chance (about one in eight) that all six win out. Hell, there’s only a 49 percent chance that TCU, Baylor, and Ohio State all win out, and they’re going to be pretty solid favorites from this point forward.

With three weeks remaining, there’s still a good chance that we’re going to end up with three obvious Playoff participants and that the argument about the fourth team only revolves around a couple of teams.

That said, surprises probably aren’t going to happen in Week 13.

Boston College at No. 3 Florida State. Six of the seven primary contenders play in Week 13, but the fact that FSU is the most likely to lose tells you a bit. The ‘Noles are at 87 percent against BC, but that’s a full-on UPSET ALERT compared to the rest of the slate: Western Carolina at Alabama, Colorado at Oregon, Vanderbilt at Mississippi State, Indiana at Ohio State, and Oklahoma State at Baylor. So BC-FSU wins the “biggest game” slot by simply being least small.

No. 19 USC at No. 9 UCLA. This game and Arizona-Utah will go a long way toward determining who Oregon plays in the Pac-12 Championship. Plus, UCLA is as well-positioned as anybody to take advantage of a round of upsets and sneak into the field as a two-loss Pac-12 champion. UCLA has about a 66 percent chance of winning, but this rivalry game should be even and fun. (Plus, UCLA’s been one of those “odds don’t seem to apply to us” teams this year.)

No. 12 Kansas State at West Virginia. Kansas State is another team positioned to benefit from complete chaos, but the Wildcats will need to clear the Morgantown hurdle first.

Week 13 is just a breather before the all-out, breathless joy that is Thanksgiving Weekend. TCU-Texas hits on Thanksgiving night, and all sorts of huge matchups (Iron Bowl, Egg Bowl, and on) come up the following Saturday.