I want to touch on the issue of the Bayesian model a little more.
The Bayesian probability model conflates actual probability with perspective-specific probability, assigning values to both.
As an example, let's say that I, without looking, randomly draw a ball from a bag containing three red balls and one blue ball, and place it under a hat.
It's now sitting there under the hat. What is the probability that the ball is blue?
Well, using the Bayesian model, we state the probability as being 1/4.
I lift up the hat, and see the blue ball. Now the probability of the ball being blue is 1.
And yet, if Myriad is out of the room and hasn't seen the ball yet, he might still say the probability is 1/4.
Nothing changed about the ball to change the probability from 1/4 to 1; what changed is my knowledge. The ball was always blue; that fact just changed from uncertain to certain with an additional observation.
The important point is that the true probability and the known (Bayesian) probability are not the same.
Now, in all real-world applications we can think of, any event with a Bayesian probability of 1 also has an actual probability of 1. However, allowing for atemporal perspective or another mechanism for omniscience that is not based on mechanical determinacy, this is not necessarily the case.
Let's say I, the Oracle, flip a coin. While the coin is in the air, I make the following two statements:
1) The coin has a 50% chance of landing tails.
2) The coin will land heads.
These two statements are only incompatible if we believe that knowledge (that is, Bayesian probability) automatically usurps true probability. It doesn't, and in fact I claim the statements are compatible.
Hence, the following statements by the Oracle are logically consistent:
1) Avalon has the capacity to select any of X, Y, or Z.
2) Avalon will select X.
The Bayesian probability model conflates actual probability with perspective-specific probability, assigning values to both.
As an example, let's say that I, without looking, randomly draw a ball from a bag containing three red balls and one blue ball, and place it under a hat.
It's now sitting there under the hat. What is the probability that the ball is blue?
Well, using the Bayesian model, we state the probability as being 1/4.
I lift up the hat, and see the blue ball. Now the probability of the ball being blue is 1.
And yet, if Myriad is out of the room and hasn't seen the ball yet, he might still say the probability is 1/4.
Nothing changed about the ball to change the probability from 1/4 to 1; what changed is my knowledge. The ball was always blue; that fact just changed from uncertain to certain with an additional observation.
The important point is that the true probability and the known (Bayesian) probability are not the same.
Now, in all real-world applications we can think of, any event with a Bayesian probability of 1 also has an actual probability of 1. However, allowing for atemporal perspective or another mechanism for omniscience that is not based on mechanical determinacy, this is not necessarily the case.
Let's say I, the Oracle, flip a coin. While the coin is in the air, I make the following two statements:
1) The coin has a 50% chance of landing tails.
2) The coin will land heads.
These two statements are only incompatible if we believe that knowledge (that is, Bayesian probability) automatically usurps true probability. It doesn't, and in fact I claim the statements are compatible.
Hence, the following statements by the Oracle are logically consistent:
1) Avalon has the capacity to select any of X, Y, or Z.
2) Avalon will select X.
Last edited: