Here is a past one for you.

Find the sci.geo.earthquakes archives and do a search for E.D.G. and May 27, 1998. What you will find is one of my rare public predictions which was for an earthquake to occur in Iran or Afghanistan by May 30, 1998. Actually, I pretty well knew exactly where one might occur in either of those countries if that happened. But I did not want to be that specific in a public forecast.

On May 30, 1998 there was a devastating earthquake in Afghanistan, right when and where I expected one could occur. And something like 5000 people perished.

To evaluate such a prediction you need to compare it with others that I might have made to see if it was just random chance or a significant forecast. And if you did that you would discover that it was about the only public forecast I have ever made for that area, or for any area for that matter. And that means that it is highly significant, nor random chance.

By itself that one prediction is in my opinion proof positive that my forecasting procedure works. But, people don't want to believe that earthquakes can be forecast. And so they generate all manner of excuses aimed at convincing themselves and others that ones like that are just coincidences.

Present forecasting efforts are focused on improving data circulated through my Data.html Web page etc. As I said, my notes posted to this bulletin board are largely intended to be progress reports, not specific earthquake warnings.

Regarding that sequence of events I proposed, Yes. It does sound frightening. But even though you yourself would probably never encounter such a situation, many people around the world have and will. And as a scientist who can have some impact on what is taking place I have to deal with the reality of the situation, not with what we might wish were the true.

These are personal opinions.