got from raf at recurse-center.
generalization of weighted average.
so:
in a case for a coin flip where if you get heads you win 10, and if you get tails you lose 0,
the expected value is 5.
10 * 0.5 + 0 * 0.5 = 5;
so:
E(X) = x1 * p(x1) + x2 * (px2) ..