This article echos some thoughts I've had about why accelerators like Y Combinator and Techstars work. An investor makes money on the really big successes, but

This article echos some thoughts I've had about why accelerators like Y Combinator and Techstars work. An investor makes money on the really big successes, but these are impossible to predict. So venture capitalists are better off investing a little bit of money in every credible startup. The article draws an analogy to index funds in the public markets, but I also see similarities with how Blumhouse Productions makes movies, or even how poker players decide their bets. The lesson in all cases is to invest a limited amount of money in a ton of reasonable projects, and then ride the few winners.