The "SGU," the Skeptics Guide to the Universe is probably the Fool's favorite podcast. Along with its companion blog and web site, it is a rich source for intelligent and entertaining discussion of the very broadest range of skeptical topics. The Fool has listened to every single episode, and these people are terrific—they really hit on the right formula. If you are of a skeptical turn of mind, even if you subscribe to no other podcast, you should listen to SGU regularly (available on iTunes or via RSS feed from their site).



In a recent blog entry, Steven Novella, MD (

As I said recently on the SGU - prophesy is easy.  It is easy to make stuff up. Typically, stuff made up as prophesy is poetic and vague in form. The advantage of this is twofold - it sounds more profound, and it places the burden on the receiver of the prophesy to do all the hard work. The one receiving the prophesy has to scour world events to find a fit for one possible interpretation of the poetic vagaries.

Nostradamus, of course, perfected this style. His quatrains are literal poetry. He was deliberately vague, mostly so that he could trash talk his contemporaries and have plausible deniability. It is important to recognize that many of his prophesies were meant for his patrons, and so they dealt with subjects his patrons cared about - namely their own time and place. They did not care about distant lands and times.

However, proponents of Nostradamus as a legitimate prophet increase the probability of finding a “hit” by removing his prediction from all context. They allow themselves all of human history to scour for pattern recognition.

It is important to think of it this way - if Nostradamus were just making up random crap, would we be able to find apparent matches between that random crap and the subsequent centuries of world history? The answer is clearly yes - and therefore the fact that apparent matches exist does not mean the prophesies have any reality to them.

The broader concept is that the use of vague or open-ended criteria allow for finding correlations even where they do not exist. Humans are great at pattern recognition. Vague patterns therefore have no predictive value, because you can find them anyway. Only specific hits (for example when an actual date and place are unambiguously mentioned) are of any value - and it turns out that every time Nostradamus gave us a date his prophesy was wrong.