Thanks for this. I understand your point now. I was misreading this:

In my scenario, the information given is not about which propositions are true about the outcome, but rather about which algorithms are controlling the outcome.