I've been having trouble with this the entire time I've been working on my current game (tentatively named 'Tyche' pronounced like 'psyche'). Since I wrote the sound positioning engine, my strategy has been to apply reverb to sounds that come from behind the player. Unfortunately though I have no good way to tell whether this is effective.
What have you guys done to make this work? Is there some kind of math I can apply to this? I've heard of HRTFs but I'm not sure how to implement them.
My github.