I guess that’s where belief ties in. If you believe you have no options, you act as if you have no options. But that doesn’t mean you have no options nor that you can’t change your beliefs.
I don’t like how belief and knowledge are tied together in Bayesian inference. I much prefer frequentism as it’s more solid BUT sometimes it’s impractical (perhaps most of the time) to get enough data so as a stop-gap there’s Bayesian.
But then again, maybe humans using data that’s created using a human notion of “belief of” to fill in the lack of knowledge gaps… makes sense. It’s just so human though.