@shoo: Pretty interesting, Stein. Thx. Taleb advocates abs. deviation, mean abs. dev. and the like. OP named reason to pick least squares: "easy to calc" IIRC,

@shoo: Pretty interesting, Stein. Thx. Taleb advocates abs. deviation, mean abs. dev. and the like. OP named reason to pick least squares: "easy to calc" IIRC, the rest of the rationale didn't exist back when it was invented and picked....