Penny's and Todd's concerns about the size of Luke's forecast area are legitimate, but they contain the assumption (perhaps justified - I haven't read Luke's website or looked at his videos) that Luke's purpose is to alert people to the possibility of an earthquake so that they can, presumably, take measures to protect themselves.

If his purpose is, instead, to demonstrate that he has a predictive method that is of scientific interest, it makes no difference how large the geographical area is. It could include the entire globe, and the Moon and Mars, too! A statistical analysis to determine if his predictions are better than the null hypothesis would also take in the time and magnitude parameters. For instance, he could predict a magnitude 3.0+ globally, and a window of two minutes, and,if successful, would have chalked up one data point toward establishing a significant result.

Mike W.