Because I come from the exact or hard sciences (math, physics, engineering), where there is extremely high predictive power, carried out in repeatable experiments, with precise measurements and agreement.
Empirical evidence often precedes the science. In sports science studies, it is pretty common to see interventions with some non-responders and athletes that decline rather than improve. However some studies do provide a compass, but “your mileage may vary” as they say.
Of course there are a good collection of first principles, however this is a branch of science with less predictive power and fuzziness which gets more into discussing the probability of an intervention. In other words, a fancy way of saying “it depends.” Which is perfectly fine.