Yeah, I like Brian's way of explaining a fault better than a couple rocks banging into each other. There are patches on faults, including at Parkfield, that break over and over and make identical earthquakes. A couple complications: Faults are not 2 dimensional: they are not smooth planes; they are 3 dimensional with lost of bends...so, as the blocks on either side of each other slide by, the part of the stress that is holding the fault together is at a different angle relative to the 3D fault and it might clamp the fault a little tighter. Plus...fluid in the fault zone tends to work against the stress that is holding it together...and high fluid pressure helps weaken a fault. It may be the fluid in the fault that would give some clues about impending quakes.
Chris