I agree that there were a lot of failed projects, perhaps more than 80%. But if you learn to choose promising projects the profit even from one good project can be large.