[etizzle]

not saying i agree with sklansky, but i think the crux of his argument is this.

Your FA analysis determines the price at which you think the purchase of the stock is neutral EV (vs avg market returns).

The market price thinks its 'value' for the stock will also be EV neutral. The market price is almost def better than a random price, regardless of how far it is from the true value.

For example, if your FA says it should be 140, and the market says its 100, would you say the true value is more likely to be 100 or 180? Independent of the market price it should be neutral (maybe not exactly because the diff between 140 and 180 is less then 140 and 100 % wise), but with the market price known it is almost certainly more likely to be closer to $100.

Thus the market price is a better predictor then a random selected price, even if that price is likely to be off given your FA. Your coinflip ESP analysis would then apply, regardless of how badly mispriced the stock is relative to your FA value.