Pavel, don't let success percentages mislead you. Numbers of hits means nothing.

Back in late 2005 to early 2007, over a period of 17 months, I successfully predicted 92% of the worlds quakes magnitude 4 and larger. Looking at it from the point of view of predictions, 67% of the predictions were successful.

However, the method used was not unlike flipping a two headed coin. I stacked the odds as much in my favor as possible simply by the design of the predictions.

One must compare the prediction to the probability of the quake happening anyway. I simply predicted highly probable quakes. In this context, statistically my predictions equaled chance.

http://www.skywise711.com/quakes/EQDB/index.html

Brian