On Wed, 9 Sep 2009, comex wrote: >> - If you take a Bayesian standpoint (with the process probabilities >> as your priors) you come to the conclusion that 1/Nth of each >> possible types of N cards were destroyed. Since this is >> IMPOSSIBLE (a higher-powered rule prevents destroying fractions >> of cards) the destruction simply didn't function. > > I don't follow. A Bayesian standpoint makes it clear that the outcome, > though > fixed, can be referred to in terms of probability because it is unknown, but > it > certainly doesn't require otherwise impossible events to occur.
Apologies I was leaving out a step. In blended probability assessment such as risk assessment from a Bayesian standpoint (e.g. risk assessment of a species becoming endangered) one can accept the Bayesian posteriors as "partial truths" (as a statistical construction) or if you prefer "legal truths" of prior events (probability that a species is currently or in the past has been endangered). But perhaps this is better phrased as a special case of legal method (3). You are indeed correct in the manner that Bayesian approaches modify classical statistics. -G.
