Home Statistics and Visualization Normal Deviate Wacky priors can work well?