(this comment written after this one)

So you meant something like "counterfactually, it's more likely that Steve Jobs changed his mind about treatment than it is that he didn't get pancreatic cancer."

Indeed, I'm beginning to see that alternatives that are counterfactually likely have perhaps no relationship with nearby Everett branches. For example, 'if I didn't become a chemist I probably would have been a biologist' may be true but nevertheless might have nothing to say about how near the two possibilities are in terms of quantum states.

A few interesting hypotheses have occurred to me that I'd like to think about more, but right now I'm a little confused about how two nearby Everett branches would look. What would the first divergence look like?

Most of what we experience is classical, and thus an averaged process of zillions of quantum states. Thus seeing anything different would require a huge statistical anomaly. For example, it's theoretically possible that a teacup will start floating on a table but it would be very unlikely and it's kind of hard to believe in. Is this the kind of thing to expect in any nearby Everett branch that actually looks different?

So now that I know i'm looking for a locally concentrated statistical anomaly -- rather than a scattering of changed states here or there -- this directs me to consider weird/unlikely events at the molecular level.

So for example, in nearby Everett branches it is relatively near for someone to not be born (since presumably a smaller number of molecular events are required for a particular conception) but given that they're born, not near for them to make a different decision at a certain point in time because presumably many molecular interactions are averaged to form decisions.

So: easy for Steve Jobs to not be here, but difficult for him to make a different decision. However, Steve Jobs being born or Steve Jobs not being born result in very different universes.

I'm thinking that once you add 'causality' to the mix, there very little reason nearby Everett branches will look like one another, from our point of view. That is, if the differences are large enough for us to notice, the results could just as likely be huge as inconsequential.

I feel confused and wonder if someone has already dissolved this?