I believe this works because the normal curve is actually also the limiting behavior for a binomial (i.e. coin-toss) experiment with n trials as n goes to infinity. Whether the ball goes left or right when it hits each peg is basically equivalent to one "coin toss". [/stats geek] |