If you throw a dice and read the results it’s reasonable to assume that it’s a trial independent of the rest of universes happenings. Thus conditioning on the dice result should not shift any probabilities concerning the rest of the world. If I throw an additional dice it doesn’t help determine what already thrown dice are. Your expectation doesn’t shift so no probability can shift.
One of the things I take from the post is that you don’t have any new information about what they rolled, but you DO now have an indication that they had some reason to roll a die. If you know what kind of decisions they make based on die rolls, you know they’re making such a decision.
For some things, that is a lot of information about the universe from the act of divination, not from the results of the act.
This is sort of the inverse of what the post is saying (that preparing for the act ensures that you consider the question with sufficient weight).
You can’t. But for example if they say what they rolled and you assume there is a correlation what they actually get and what they would say then you have a chance to narrow it down. If you know it’s not corrrelated to anything (ie is pure dice) you know it can’ t be evidence.
You can guess. You can roll the die yourself (and guess that it came up the same way). You can also examine the die, and then guess.
If I throw an additional dice it doesn’t help determine what already thrown dice are. Your expectation doesn’t shift so no probability can shift.
Also, this contains some assumptions that aren’t always correct. I can throw a die a bunch of times, and notice that it comes up “6” or “1″ an awful lot an conclude it’s weighted. (A shift in expectation.)
A guess would be equally good without dice throwing. Indeed if you have access to the dice that generated the result we want to know about you can infer distribution information. But if you have a different die and determine that it’s weighted it doesn’t tell whether the orignal die is weighted. If you knew the dice came from the same factory you could infer something. But you manufacturing a fresh dice is justified to assume to not be distributionally connected. If you have information that you know to correlate your manunfacturing process to be similar then that contains your information and the actual rolling of the die doesn’t tell you anything.
If you throw a dice and read the results it’s reasonable to assume that it’s a trial independent of the rest of universes happenings. Thus conditioning on the dice result should not shift any probabilities concerning the rest of the world. If I throw an additional dice it doesn’t help determine what already thrown dice are. Your expectation doesn’t shift so no probability can shift.
Someone rolls a die, and writes down the result. How do you guess what they rolled?
One of the things I take from the post is that you don’t have any new information about what they rolled, but you DO now have an indication that they had some reason to roll a die. If you know what kind of decisions they make based on die rolls, you know they’re making such a decision.
For some things, that is a lot of information about the universe from the act of divination, not from the results of the act.
This is sort of the inverse of what the post is saying (that preparing for the act ensures that you consider the question with sufficient weight).
You can’t. But for example if they say what they rolled and you assume there is a correlation what they actually get and what they would say then you have a chance to narrow it down. If you know it’s not corrrelated to anything (ie is pure dice) you know it can’ t be evidence.
You can guess. You can roll the die yourself (and guess that it came up the same way). You can also examine the die, and then guess.
Also, this contains some assumptions that aren’t always correct. I can throw a die a bunch of times, and notice that it comes up “6” or “1″ an awful lot an conclude it’s weighted. (A shift in expectation.)
A guess would be equally good without dice throwing. Indeed if you have access to the dice that generated the result we want to know about you can infer distribution information. But if you have a different die and determine that it’s weighted it doesn’t tell whether the orignal die is weighted. If you knew the dice came from the same factory you could infer something. But you manufacturing a fresh dice is justified to assume to not be distributionally connected. If you have information that you know to correlate your manunfacturing process to be similar then that contains your information and the actual rolling of the die doesn’t tell you anything.