Prior to posterior analysis

If there are no single correct priors (u8, m347) what about sufficient statistics?

The shape of the prior represents the beliefs about the $\theta$ before the data are observed

low wide priors represent weak beliefs, high narrow priors strong

the most likely value of $\theta$ should be at the mode of the prior distribution to reflect the "most likely" density of the $f(\x)$

is the recipricol of variance