I commented on your statement that new science cannot be tested with old methods, which is patent nonsense in this case, and undermines your claim to understand what you are doing.

The quakes either do or don't happen more often than chance during your predictions, which indicates their effectiveness. I'm not checking it out myself, your statements that the path of sun shadows along a wall are noticeably perturbed prior to earthquakes is so extremely unlikely that it renders the rest of your logic suspect to me.

If you've never compiled your own track record, there is little choice but to accept Roger's evaluation.