[sidman]

So it seems for investors, the big winners are simply based on luck, there is no logical factor that you can easily pick out that will give you a better chance of winning accept for possibly experience over a number of years that allows you to pick the features of a successful startup that initially doesn't look like they are going to be successful (maybe a glimmer in the founders eyes, something that isn't completely obvious to the general investment population).

But whats the thought process and how does it change given this insight ? It should still be the case that as an investor you still should focus on the best opportunities and try to maximize a decent return from the majority of your investments then focus on picking that facebook, dropbox or AirBnB right, just as you would if you were a trader in the stock market ... otherwise your just gambling ?! Focus on the technicals and data that give you the best odds and just take those trades cause taking trades where you are waiting on information from the fed chairman or draghi without insider info for example (even though they have big pay offs if you swing the right way) will earn you pretty much the same amount of returns as losses in the long term.

If the potential to win big time is counter intuitive what do you trust from data in front of you to make the decision to invest other than the "gut feeling" ?

In any case, i would think with dropbox it never really sounded like a bad idea and the founder/founders (Drew and Arash) sounded very promising even from the start. If this is the case even good bets do pay of big time to :) maybe these are easier to pick then the facebooks where at the time it does look like a bad idea.

Basically, Im curious as how this insight in PG's essay would change how an investor should invest or is it just an FYI and something to keep in the back of your mind as an investor ?

[pmarca]

> So it seems for investors, the big winners are simply based on luck...

That's not quite right. It's more a question of taking enough of the right kind of chances in funding the companies that you do fund, as opposed to all the companies you don't fund.

I think of it as the concept of "search space" -- you need to define your search space properly, so that you have a good chance of having some big winners in the search space. You have to define your search space because you can't invest in everything. The investors that do the best job of defining their search space (having the most accurate criteria, best judgment, best mental models, etc.) have the best chance of funding some of the winners -- almost as a side effect.

When you talk to old successful VCs, what they tell you is that they had almost no ability to judge which of the companies they funded would go on to be the big winners in their portfolios at the point of funding them. They only learned that later. The challenge was to get enough of the right kinds of risk "above the line" and then give the portfolio time to develop.

[rsheridan6]

I think there is a way to improve your chance of winning big - don't invest in startups that serve niche markets. Look at this list of YC companies from last summer: http://www.quora.com/Who-are-the-Summer-2011-Y-Combinator-co...

Don't you think you can you pick out some that had zero chance of becoming massive-scale homeruns, because they served sort of a niche market? There are none that were obviously (to me, or apparently to pg and co) going to be unprofitable, but if you want the next Facebook, you're not going to get it by investing in, say, Codecademy, which, as laudable as its goals are, only appeals to newbie coders, who are a tiny slice of the population.

[sidman]

> but if you want the next Facebook, you're not going to get it by investing in, say, Codecademy, which, as laudable as its goals are, only appeals to newbie coders, who are a tiny slice of the population

Well you would just never know. Computer science and software engineering could become part of the basic school curriculum in the next few years. In the news you see that "the 14 year old boy developed an X for the iPhone and is now a millionaire". Enough of these stories could prompt the education system to deem coding as a basic skill like math, english and science. Code academy could end up being the platform that schools use to teach it.

But i guess the chances are slim because as it stands now the population is small though the possibility is there.