I don't see any mistake that is obviously eliminated for Facebook apps, and problem #7 (choosing the wrong platform) is a significant risk. However, it's not

I don't see any mistake that is obviously eliminated for Facebook apps, and problem #7 (choosing the wrong platform) is a significant risk. However, it's not clear that single Facebook apps qualify as startups in Paul's sense. At least today, FB apps are more like content that plug into an existing distribution network, like movies + distributors or blogs + AdSense. If you wanted to found a business that made FB apps, many per year, that's a different story. Then you'd be more like a game studio.