If you look at the math on this there are 15 quakes a year 7.0 or bigger. That's one every 24 days. So with a two day window of prediction that gives you a one in 12 chance of being right just by random selection. Additionally, there are 7.0 + quakes that he is not predicting.

If you go below 7.0 there are 134 earthquakes a year between 6 and 6.9. If you go with a two day window on those you have a 74 percent chance of being right just with random selection.

Of course the real test is the 10 that he's predicting for December.

We'll know soon enough.