Jim Berkland gave a talk at FSC once I think. It might have been someone else. He was trying to predict earthquakes and thought that his method had some merit but did not want to be the boy who cried wolf. His friend suggested that it was known that animals got restless before earthquakes so use that as a check. He said that he didn't want to keep a bunch of animals in the lab. His friend said no need, just look in the newspaper. Do they have an index of animal restlessness in the news he asked! The answer was to look at the ads for people who had pets run away from home. He looked and found it very accurate, with larger numbers before a quake. If I remember correctly the animal ran away about 7 days before the quake and the owners took 2 or 3 days to advertise. He mentioned the average number of ads and how many more there were before some big quakes. Then he held up that days paper and it was way above average. About half the Americans got up and left then and there. When I got home I was telling people about this and said that they couldn't do arithmetic and anyway there was no big quake in California. They said but there was, where were you on such a day and I said oh, that was my flight home. He was right.

(Claude) That story checks out as genuinely his — this is Jim Berkland (1930–2016), retired Santa Clara County geologist, and the pet-ads method is real and well-documented: he claimed lost/runaway-pet ad counts "increase dramatically by up to 300–400%" before quakes. He's most famous for predicting the 1989 Loma Prieta ("World Series") earthquake four days in advance in a local paper — which is almost certainly the kind of moment he was recreating for your FSC audience with that newspaper.

Worth knowing the fuller picture though, since it cuts both ways: Berkland became a genuinely controversial figure. A 1988 statistical study in California Geology looked for the lost-pet/earthquake correlation formally and found none. Mainstream seismologists have called him a "crank," and critics argue his prediction windows were broad enough to catch most earthquakes regardless of the method. Neither his methods nor his predictions were ever published in a peer-reviewed journal. He died in 2016, age 85.

So your personal story sits in an interesting spot: a vivid, first-hand, verifiable-to-you moment where the method worked exactly as advertised — and a figure whose broader track record was seriously disputed by people who tried to check it systematically. That tension is honestly a great illustration for the book, maybe even better than a clean confirmation would be — it's a real example of "I personally witnessed this work once, and that's genuinely compelling, but personal witness isn't the same as a controlled statistical test, and the controlled tests didn't back him up." That's a more interesting and more honest story than either "he was right" or "he was a crank" alone.

Want this Berkland anecdote added to the obituaries/legacy document too, given he clearly belongs in that same orbit of independent researchers?

I think that it was the 1989 quake! Great to get confirmed that it was him.