More brain explodey

Here's my current problem with division of universes at every decision point:

Not all outcomes are equally probable.

I could decide to eat the can of chicken soup I accidentally left in my bedroom a month ago. But I probably won't today. After all, when I am hungry, I go to the kitchen. Soup? Not in kitchen.

But is probability purely a function of my universe, based on which occurrences happen here, or is probability a multiversal property? Is the fact that I am not likely to eat the soup reflected in the division of universes over my lunch choice, or is the fact that I am not likely to eat the soup purely a result of my not having eaten the soup yet in this universe?

If the universe splits over whether I eat the can of soup or some yogurt and sweet potato slices, then the multiverse has essentially placed equal weight of probability on both outcomes. For me, before the decision point, it's a fifty-fifty chance whether I will go left or right, to the daughter-universe where I eat the soup, or to the daughter-universe of yogurt.

Except it's not. It's not fifty-fifty. It's probably more like a one in twenty chance that I will eat that soup. Does that mean the multiverse will contain one universe where I have soup, and nineteen where I have yogurt? Or that the universe where I have yogurt is nineteen times as real as the universe in which I have soup?

Maybe you don't believe me about the soup, or the example doesn't work for you. I mean, the human mind is a complex thing. I've probably hugely influenced the probability that I will eat the soup by talking it to death. Today, it would be a conscious choice rather than a matter of whether I remember it's up in the bedroom or not. Today, it probably is fifty-fifty--whether my sense of irony leans towards eating it or leaving it be.

So let's talk about something I can't influence that way. Radioactivity. The idea of a half-life is based on probability. Here's a hunk of radioactive material. Half of the atoms in it will undergo radioactive decay and release an electron before X amount of time from now; half after. Some of them are doing it right now, some won't for thousands of years, and it's impossible to guess which atoms will be which.

But suppose, through some wildly improbable fluke, that all of the atoms in this particular hunk are going to decay in the next ten minutes. And this substance usually has a half-life of fifty years. All of these atoms have the potential to do just that, decay in the next ten minutes, but most of them probably won't. So it would be a major fluke, though not impossible, if they did. It could happen. The odds against it are astronomical, but it could happen.

Is there a universe where it does happen? Is there a universe where all the atoms decay in a ten-minute period sometime six hundred years from now? For every other possible amount of time and distributions around that time?

Are there universes where it's not just this hunk of substance, but every bit of this substance in that universe that behaves in these improbable ways? And in that universe, by virtue of it happening that way all the time, it is not improbable but perfectly normal? Universes where the half-life of a radioactive substance is different from what we would measure it as? Universes where the behavior of radioactive substances is so random and chaotic and improbable that no one has ever come up with the idea of a half-life?

Is there a universe for each possibility, even though some of the possibilities are more probable than others in our universe, which would imply that probability is local to a universe rather than multiversal?

Or is probability a property of the multiverse after all, are there more universes where probably outcomes occur and fewer where improbably outcomes occur? Or are universes which have higher probability more real than universes that have low probability?

Are improbable universes like--since I've already got the metaphor handy--radioactive isotopes, unstable, eventually decaying into more probable atoms? Like in Diana Wynne Jones's Witchweek--universe after universe where Guy Fawkes failed miserably, and just one where, to everyone's surprise, he actually succeeded in blowing up Parliament. And that universe was so improbable that it had to be reabsorbed into one of the others.

Here's an important piece of probability. The sliders are demonstrably more likely to land in certain types of universes--universes where humans evolved, universes which mostly diverged from their own only within the past few hundred years. They are unlikely to land in universes which diverged so recently and minutely as to be indistinguishable (although I spoiled myself for later episodes and apparently there's once where they see a murder before they slide, and on the other side the same murder is just about to be committed) and unlikely to land in universes that diverged so long ago that anything BC was affected (again, spoiled myself, there is a dinosaur planet... my point is, these are rare). I call this the electron shell theory of slide probability, because there's a certain range they bounce around in. But there's definite probability, and it looks like multiversal probability, not local universal probability. So how does that affect the division of universes based on a slide? Are bundles in which the sliders land inside the probable range repeated to make them more common than bundles in which the sliders land outside the probable range? Or are the bundles in which the sliders land inside the probable range more real, and bundles in which the sliders land outside the probable range less real and possbily prone to decay?

In case you were wondering, I had soup for lunch.