It depends whether by "predict" you mean "predict" or "score hits in a prediction evaluation". It's easy to do rubbish prediction of small events. But it would be crazy to suggest that genuine prediction is easier for small events. "I can see this ant quite clearly, but you tell me there's an elephant there? I don't see it."

As I said in a previous thread, anyone who claims to have predicted some small earthquake but misses a big one is someone you should be deeply suspicious of.