I’m a bit confused about what your methodology is, in the appendix. You start by saying:
We should demand that our decision theory gives the right answer “for the right reason”.
I agree! I think this is really important. I argue for something similar here. But then, later you say:
Here are the cases which make me skeptical of physically causal CDT: [lists bottom-line answers given by CDT in the cases]
This seems in tension with the first quote. If what we care about is the reasons for the decision recommended by the decision theory, shouldn’t we be focusing on questions like “What does it mean for my decision to have good consequences in the relevant sense? What do I care about when I say I want to ‘maximize EV’? What does this say about whether the relevant notion of EV is causal or evidential or [something else]?” Rather than “Does this decision theory recommend bottom-line answers that I intuitively agree with?”
(Maybe you’d say “those bottom-line answers are evidence about the right reasons”? See this part of the linked post for my response to that.)
I’m a bit confused about what your methodology is, in the appendix. You start by saying:
I agree! I think this is really important. I argue for something similar here. But then, later you say:
This seems in tension with the first quote. If what we care about is the reasons for the decision recommended by the decision theory, shouldn’t we be focusing on questions like “What does it mean for my decision to have good consequences in the relevant sense? What do I care about when I say I want to ‘maximize EV’? What does this say about whether the relevant notion of EV is causal or evidential or [something else]?” Rather than “Does this decision theory recommend bottom-line answers that I intuitively agree with?”
(Maybe you’d say “those bottom-line answers are evidence about the right reasons”? See this part of the linked post for my response to that.)