Marc andreesen attacks this method for good reason - the case methodology suggests you'll have a lot more data in the real world than you do and trains people to look for clues in the past about what you should do now; you simply can't always do that.