Maiden.bet

The beginner's guide · Chapter 3 of 9

What “68% to win” actually means

A 70% favorite loses three times in ten. That is the forecast working.

Here is the sentence that separates traders from fans, and it fits on an index card: a probability is a promise about frequency, not a prediction about Saturday.

68%the whole call

When our desk publishes “68% to win,” it is not saying the team will win. It is saying: put this exact situation on the schedule a hundred times, and this side comes out ahead in about 68 of them. The other 32 are not fine print. They are the forecast.

✓✓✗✓✓✓✗✓✓✗7 of 10, which is what “70%” means
Ten games, each called at 70%. Seven land, three don't, and the three are not the model failing. A 70% that never lost would be a mispriced 100%.

Sit with that strip for a second, because your instincts will fight it for months. When a 70% call loses, every part of your brain files it as wrong. But a 70% caller who never loses is the one you should run from: their 70% was really a 100%, which means their numbers are decoration. The three red boxes are what honesty looks like at 70%.

One result can never grade a probability. Only a hundred results can.

The Saturday multiplication

Now scale it. A college football Saturday runs thirty-plus games. Suppose the favorite in every single one is a genuine 80% shot, a strong number. Thirty games times a 20% miss rate is six upsets, every week, when nothing is wrong. The feed will call each one madness. The arithmetic called it in advance.

This is why “upsets happen” is a useless observation but how oftenupsets happen is a tradable one. If 80% favorites are losing 20% of the time, the world is in order and there is nothing to do. If they’re losing 35% of the time, someone’s 80s are fake, and chapter 4 gives you the tools to find out whose.

Your own number first

The discipline this chapter is really teaching: never look at the market before you’ve named your own probability. Watch the game, weigh the injuries, then commit to a number (I think this is 65%) and only then compare. If the market says 75%, one of you is wrong, and finding out which one, carefully, over many games, is the entire craft. Numbers like our published college football probabilities exist precisely to be a second opinion you can check against a public record, not an oracle to obey.

Frequently asked questions

What does it mean when a team is 68% to win?
It means that across many games exactly like this one, the team wins about 68 of every 100. It is not a prediction that they will win; it's a statement about frequency. A 68% call that never lost would actually be a bad forecast, because it was really a 100% dressed down.
If a 70% favorite loses, was the forecast wrong?
Not by itself. A 70% forecast promises three losses in every ten; the losses are part of the promise. A single result can never grade a probability. Only a long run of results can. What you can check is whether all of a forecaster's 70% calls, taken together, won about 70% of the time.
Why do upsets happen so often in college football?
Because there are so many games. With thirty-plus games on a Saturday, even if every favorite were a true 80% shot, you'd expect around six of them to lose every single week. Upsets aren't the system breaking. At that volume they are a mathematical certainty.

Take the whole guide with you

Sign up for the daily desk email and all nine chapters arrive as a designed PDF. One email a day, our graded numbers, unsubscribe anytime.

Already on the list? Download the PDF. And as everywhere on Maiden: these are probabilities and mechanics, graded in public. Nothing here is betting advice.