Science would do well to think non-parametrically rather than prescriptively model phenomena with this theory or that formula, e.g. a percentile or range of observations may not have the hollow satisfaction of determinism, but its a more honest approach than confirmation bias. Consider Darwin's evolution: Is it random selection or a foregone conclusion? Because since time's not really a thing the sequences happen in a fairly predictable way. Even if we don't have all the variables available and in our human ignorance can't possibly know all there is to know. But it does seem that there's some idea that given a relational ontology or pattern-based reality that what's happening now is guaranteed to converge even if we don't know where it's converging.