Amit;

You have the wrong idea here. I'm not asking you to be 99% correct; I'm asking you to be 99% better than chance.

I've been running tests on random predictions, sets of 100 dates chosen at random for quakes of 6.0 to 6.6 mag. As it turns out, it's very hard to beat chance that way.

The odds on a hit are 0.65 so it takes 65 hits in 100 tries to equal chance and more than that to beat it. But after over a thousand tries the program got 65 hits and the test said that was at chance level. Thus the program works correctly.

When I run your predictions thru the program it says you are over 2 standard deviations BELOW chance.

Case closed.

Roger