The prediction problem is one of gathering data.

How can we measure the strain level at quake depths? It's too difficult and too expensive and also we don't know the limits for every circumstance.

Roger