It’s just really bad form to be predictable imo in applied game theoretic way. Generally you don’t even want to be a random process with a fixed parameter because that will be gamed easily with a mixed strat rather than them cooperating.
This is why you don’t listen to the experts all the time and completely when the issue is divisive due to core difference in terminal values.
I want to expand on this which I really, is that I just forgot to write a bunch of stuff.
Compromising by listening to the experts makes sense in a Pareto-ish utilitarian framework, but the fact you are listening to experts means that you have often have limited abilities to confirm the soundness of policy recommendations by them in contemporary public policy making. Sure one time in a high enough trust society this would work, but eventually it will select for experts who would try and sneak in stuff for their own group, or that the same experts begin to notice ways they can defect.
This is kind of the same thing with if a demographic group is loyal to a party essentially no matter what, it means that they have very little incentive to really do anything on the margins for them.