There are two ways to take a risk: 1) Bet on something unlikely and get lucky 2) Have an advantage over others so that your chance of success is higher than t

There are two ways to take a risk: 1) Bet on something unlikely and get lucky 2) Have an advantage over others so that your chance of success is higher than the average chance of success. #2 is pg's approach to startups - starting a company is "risky", but starting a company if you're determined, smart, and connected to a system that turns entrepreneurs into companies isn't risky at all, with a ~50%+ success rate. On a sinister note, insider trading falls into this category too - the market isn't a random walk if you know information that will move it, and your information gives you an advantage over the average investor.