[yarou]

It's actually an interesting toy problem. Brings me back to microeconomics. Most people tend to be risk-averse, so they will never accept a 50/50 gamble (ref: http://en.wikipedia.org/wiki/Von_Neumann%E2%80%93Morgenstern...). However, those in the startup space tend to be risk seeking, thus they are more inclined for the second option. Behaviorialists will point out this is a classic example of framing effect (http://en.wikipedia.org/wiki/Framing_effect_(psychology)). While the distinction between 10 or 20 dollars is trivial, the distinction between 5,000,000 and a gamble of 20,000,000 is more substantial.