Ah. So if I understand correctly, your intuition on what will satisfice sometimes returns zero information, which certainly happens to me sometimes and I would guess most people. In that situation, I switch from optimising on the decision as presented, and optimise on <choose and implement decision procedure> + <decision as presented>.

In most cases, the variance in utility over the spread of outcomes of the decision is outweighed by the reduced cognitive effort and anxiety in the simplified decision procedure. Plus there's the chance of exposure to an unexpected benefit.

In other words, there may be a choice that is better than the current best candidate (however that was derived), and rules may exist that dominate "random", but it's not worth your time and effort to figure them out.