Something to think about:
The general perception is that time flows or moves "from" the past and "to" the future. It's not clear what this "motion" actually is, though. At least, I think, we can all agree that "the place memories represent" is the past (and the "from", if anything is,) and that "the place plans represent" is the future (and the "to", if anything is).
Some hold that "causality" isn't real. But it might be more accurate to say that "causality" is whatever it is that RCTs measure. Mass and photons might not "exist", either, but we still refer to "whatever causes gravity" and "whatever trips photon detectors", after all—and have a single word for each! Futher, I think that it is nonsense to say that "the future causes the past" (outside of any causal loops proper). Given that causality exists, and it's nonsensical for the future to "cause" the past, we could then establish that causality is also necessarily tied up with the "direction" of time.
The last question, then, would be: is entropy necessarily connected to causality? (I think this is likely, given the necessity of entropy in an RCT.) Perhaps a good next step to pulling on this thread would be Judea Pearl's writings.
While some disagree that entropy is the key to the directionality of time, I think that if we could clearly establish entropy as the key to causality, causality as the key to future/past, and future/past as the key to the directionality of time, we could put a solid case that entropy does explain this directionality. I think the first of these (entropy→causality) is going to be the hardest to prove, by which I mean it may require real scientific research to be sure of, while the others (causality→F/P and F/P→directionality) may be "armchair-soluble" by philosophizers.