Petra-the test would be whether your predictions beat chance-taking into account that anyone can beat long-term chance by paying attention to which areas are currently active. Maybe it is time for Roger to come up with a list of M3+ quakes for California, and adapt his program in some way (may be too difficult?) to deal with the statistics of choosing recently active areas. My SE Loyalty Islands quake was an extreme example of chosing an active area......The reason why your predictions needs testing is that most research scientists, including me, would be skeptical that ear tones precede quakes-and even more skeptical that you can tell distance and direction.....If Roger is in to this, the next set of predictions in need of evaluation are Don's-using the "data". I don't consider Lowell's evaluations of these predictions independent. Chris