Suppose A: the evolution-over-time of the state-of-the-world is determined fully by physical laws which explain the behaviour of an objective physical reality. Then subjective experience, by definition, would not affect (in a “causal” manner) the evolution-over-time of the state-of-the-world.
By contrast, suppose B: the evolution-over-time of the state-of-the-world is partly determined by phenomena of subjective experience and actions taken in response to subjective experience (e.g. I subjectively experience pain and consequently take an action to avoid pain).
In either scenario A or B, we acknowledge that humans undergo subjective experience.
The set-of-plausible-worlds under scenario A includes both a world where stubbing my toe is subjectively painful and a world where stubbing my toe is subjectively pleasurable. Both worlds are plausible in the absence of any reason to suppose a particular relationship between phenomena in an objective reality and character of subjective experience. (i.e. In our objective reality, and supposing scenario A, we know that people don’t go round intentionally stubbing their toes, but this doesn’t tell us anything in particular about the character of their subjective experiences. Both pleasure and pain would be plausible.)
The set-of-plausible-worlds under scenario B seems to include only worlds where stubbing my toe is subjectively painful. (i.e. We know from observations of an objective reality that people don’t go round intentionally stubbing their toes, and (supposing scenario B) this suggests that people subjectively experience toe-stubbing as painful.)
i.e. In scenario B, we have reason to believe that toe-stubbing would be painful. In scenario A, we have no reason to believe toe-stubbing would be painful. The fact that toe-stubbing is subjectively painful therefore seems to lend some credibility to the supposition B, since in scenario A there seems to be a “coincidence” (that toe-stubbing is subjectively painful despite us having no reason to suspect it would be).