I think that, at this point in time, after what we've seen at both Jupiter and Saturn, we can make the following generic statements:

Rocky bodies with no atmosphere: Predominated by craters.

Solid icy bodies with no atmosphere: Predominated by craters and then cracks (i.e., tectonic formations).

Partially solid (i.e., with liquid water underneath a solid ice crust) icy bodies with no atmosphere: Predominated by cracks and then craters.

The less solid an icy body is, the fewer craters we see and the more cracks we see. But at no point do we seem to lose the tectonic formations -- if it's icy, it's cracked.

So, we seem to have a rough rule of thumb in terms of the potential for subsurface oceans -- the higher the ratio of cracks to craters, the greater the chance of an ocean.

-the other Doug

Interesting, but can you really reduce all the idiosyncrasies to variations in one or two parameters? How would, for instance, Triton or Miranda fit in here? Then there's the question of histories. The Galileans especially may have undergone huge fluctuations in the amount of tidal heating over various timescales. So do lots of cracks necessarily indicate the continuing presence of a liquid water layer, or merely preserve the record of intermittent internal melting? Of course it's easy to find awkward cases. I agree it's a useful enterprise to try to identify trends, if only to make the exceptions stand out more clearly.

Followed by lava flows, and then cracks (depending on size). The moon and vesta both have extensive lava flows, with Mercury behind.