Careful with the Aumann agreement theorem, it doesn’t state what EY seems to think it states and I’ve yet to see an example of the process being used correctly.
Note in particular that the final estimate need not even be between the initial two estimates. If the “Aumann process” looks like two people incrementally updating their beliefs towards each other, they’re doing it wrong.
Yes. That’s why I thought about quoting that ‘verbatim’. But it gets ugly that way. And the simplified form is admittedly oversimplified. In the same way as the sherlock quote is misleading.
Careful with the Aumann agreement theorem, it doesn’t state what EY seems to think it states and I’ve yet to see an example of the process being used correctly.
Note in particular that the final estimate need not even be between the initial two estimates. If the “Aumann process” looks like two people incrementally updating their beliefs towards each other, they’re doing it wrong.
Yes. That’s why I thought about quoting that ‘verbatim’. But it gets ugly that way. And the simplified form is admittedly oversimplified. In the same way as the sherlock quote is misleading.