9 out of 10 times five? Isn't that like having everyone in the room have the same birthday in five places?
No, your problem is you are looking for facts to support your hypothesis rather than the other way around. Since you are simply looking for a corrolation of events, the birthday paradox explains why you are able to do this. You are not trying to find a corroloation between specific events, simply any event.
This is quite clear in your post. In one case it is when a hurricane changed direction. In another it is when it was formed. In another it is when it hit. There is no method at all to your logic besides picking dates and trying to find a connection, no matter how loose it is.
This is no different than people who look to Nostradamus to try to explain past events. If he, or in your case the weather, is so predictable, why can't you predict what will happen ahead of time instead of simply finding a 'connection' after the fact?