@joe our future plans of evolution include running multiple stopped clock in parallel and picking one at random. We already achieved remarkable accuracy gains in the lab.