The Bills and Notre Dame Will Be Champions. Book It. We Are Probably Wrong.

How sure should you be when you say “book it”? Here is our answer, with its number. We are picking the Bills and Notre Dame, and our own models agree they are among the best bets on the board. The Bills are the NFL’s likeliest champion at about 15%. Notre Dame is college football’s second likeliest at about 16%.

That should stop you. Most people think the favorite is likely to win. In a playoff, the favorite is the best single bet and still loses most of the time. It is one team against the whole field. By the end of this you will be able to tell, in your own head, what you believe from what you know.

The Call, in Writing

We said Notre Dame in July. Issue #109 called the 2026 national title for the Irish, dated so you could hold us to it. Six games later they are 6–0. They have beaten Wisconsin, Michigan State and North Carolina, and nobody has come closer than eleven points. We are not backing off.

Now we add the Bills. They are 4–1, with wins over the Lions and the Rams. Our ratings, built from 107 point spreads this season, rank them first in the league.

So write it down: Bills in February. Notre Dame in January. We will grade both, in public, whatever happens.

Before You Read On, Write Your Number

Stop here and try this. Pick a team you love. Write down the chance, out of 100, that it wins the title this year. Do not look anything up. Then write one line: why that number?

Hold onto it. We will come back to it.

Believe It or Know It?

Here is the idea at the heart of Bayesian statisticsi. A belief is a number. You start with one — your prior. Each game is evidence, and it moves the number up or down. Where you end up is your posterior: what the evidence lets you believe now.

Notice what “book it” does in that system. It says 100%. And a prior of 100% has a strange property: no evidence can ever move it. Bayes’ rule multiplies your prior by the weight of the evidence. Multiply certainty by anything and you still have certainty. Short of the team being knocked out, nothing moves it. A fan who is sure cannot learn from the season. That is not a figure of speech; it is the arithmetic.

So go back to your number. Was it near 100? If you love the team, it probably was. Ours would be, if we let it. The gap between that number and the honest one is your bias, and now you can measure it.

Our Biases, Named

We should say the hard part plainly. This paper covers Notre Dame first, every year. We picked the teams we write about most. That should make you more suspicious of the call, not less.

Here is how we keep ourselves honest. The pick is our belief. The 15% and 16% are what the evidence allows. We publish both, side by side, so you can see how far our hearts run ahead of our numbers. If the Bills win, the model was not wrong; 15% things happen about one year in seven. If they lose, our pick was not foolish; it was the best bet on a board where every bet is likely to lose.

A New Thread: What Do You Believe?

This starts a series. Every fan carries a set of stats they think matter. Wins in close games. A quarterback’s record. “He just knows how to win.” We want to ask three questions about each. Which stat do you trust? Why do you trust it? Do you believe it, or do you know it?

Send us yours. The best ones become issues, and we will test them in the open, the same way we will test this call.

What to Take Home

When someone says “book it,” ask them for a number. If they say 100, they have told you they cannot be persuaded. If they give you a real number, you can have a real argument.

Try it on yourself first. It works on teams, on stocks and on elections. Put a number on the thing you are sure of. Then watch what the season does to it.

How we did it. NFL: team ratings solved from 107 published point spreads, so they carry the betting market’s view as well as ours. Remaining games played out 20,000 times, seven playoff seeds per conference, higher seed at home. College: SP+ ratings turned into game odds, the rest of the season played out 20,000 times. The twelve-team field is our stand-in for the committee: fewest losses, then strength. We skip conference title games and automatic bids, so treat the college numbers as rough — give or take a few points. Records checked against ESPN and CollegeFootballData. As of October 14.