Under ideal theoretical conditions, yes. In practice, friction and material deformation will probably give the top of the crest some degree of stability, possi

Under ideal theoretical conditions, yes. In practice, friction and material deformation will probably give the top of the crest some degree of stability, possibly enough for a demonstration. At the very least it will probably be much slower to roll from such a metastable orientation. Edit: also, based on the following gif someone else posted, I don't think the top of the crest is the metastable point. You can see the object linger at that point before seemingly spontaneously finding true equilibrium - illustrating my point rather well. https://en.m.wikipedia.org/wiki/File:Gömböc.gif https://en.m.wikipedia.org/wiki/File:Gömböc.gif