Hi all;

At Boyko's request I have evaluated his 2012 predictions and am posting the results here.

He had 371 predictions that year using a 300km radius about a given lat/lon point.
I'm not sure about his date and mag range and my request for clarification brought no response so I gave him the most generous ranges that seemed reasonable. I allowed mag to go one unit below his value and accepted anything larger as well. I allowed a 3 day window centered on his given day. No location range was needed as a 300km radius is large enough.

Probability was a problem because he had so many predictions and each one required the program to examine every quake in the file. I chose to use 2012 for probability just to keep things reasonable timewise.

The results show that he had 50 hits out of 371 predictions and his significance level was over 1.7 standard deviations below chance.

I'd say that's a conclusive failure since any tightening of parameters would make things even worse.

Roger