11 ms·

Oh I know! It's fundamental attribution error! (http://en.wikipedia.org/wiki/Fundamental_attribution_error http://en.wikipedia.org/wiki/Fundamental_attribution_

Oh I know! It's fundamental attribution error! (http://en.wikipedia.org/wiki/Fundamental_attribution_error http://en.wikipedia.org/wiki/Fundamental_attribution_error) Oh wait - that's not right, that's actually the gross error that all "What's the X behind Y?" questions make - when in reality, it's probably about 100+ different things all coming together at the same time. For YC it's the falling cost of bandwidth, the falling cost of storage, the falling cost of computation, an increase in the number of internet users by nearly 1 billion over the last decade, a doubling in the global economy over the last 2 decades - quite frankly the list leading up to the creation of a Y Combinator is almost endless - and I haven't even dealt with the actual people who help make YC the success that it has become today. Single factor explanations of complex phenomena are stupid.

pshin45 14y ago
All factors are not created equal though, and you must admit that there is probably a combination of a few key factors that help explain most (but not all) of their success, no? (http://en.wikipedia.org/wiki/Pareto_principle http://en.wikipedia.org/wiki/Pareto_principle) I think that is what they/we are trying to get at. I agree with your general principle that causation is rarely as black and white as we make it out to be, but saying that all/most phenomena is a result of hundreds of different factors coming together at the right time and right place would make it very difficult to ever learn anything about anything. > "..the falling cost of bandwidth, the falling cost of storage, the falling cost of computation, an increase in the number of internet users of nearly 1 billion over the last decade, a doubling in the global economy over the last 2 decades." These are all situational factors common to TechStars, 500 Startups, etc. but which don't explain YC's unique success.
- confluence 14y ago
  > very difficult to ever learn anything about anything Not really - if you are talking about investigating things that have a multitude of conflating variables affecting the end result. In that case you aren't learning anything either - because then everything is just correlation. When one investigates some arbitrary phenomena, it is important to constrain one's experiment and control all but one variable so that one can be sure that the changes one observes are more likely to be due to causation rather than correlation. For example: Stating that Africa is a failure because of X thing is full of peril. However, stating that when people don't have easy access to safe drinking water, they are more likely to die of water borne illnesses, and that this has probably contributed to the failure of Africa is a much more secure statement. It's not nearly as pithy, and it includes many qualifiers - but then again, it is a much more accurate and true assessment. It is also much less likely to be subject to vague/"woo"/dumbass business like thinking. > These are all situational factors common to TechStars, 500 Startups, etc. but which don't explain YC's unique success. First mover advantage anyone? Path dependence is a bitch - just ask DVORAK.
  - pshin45 14y ago
    I don't disagree with what you're saying. If I were to sum up why YC might have succeeded like it did, I'd say that pg is great at pattern recognition and is better than most at figuring out who might become a great entrepreneur vs. who is just posturing, which increases (but doesn't at all guarantee) his chances of finding great founders capable of building great companies. Of course, things played out pretty well for pg, but with less luck perhaps it might've taken a longer time for them to find their Airbnb/Dropbox, or perhaps Drew Houston and the three co-founders of Airbnb might have failed several times before hitting their respective home runs. In which case TechStars and 500 Startups (admitted copycats of YC) might never have come into existence. pg's knack for identifying great founders doesn't guarantee that everyone who gets into YC is going to be "great," just like there are probably plenty of not-so-capable people in Harvard who somehow still managed to get in. But I would posit that the % of potential great founders is significantly higher at YC than at other accelerators (due to brand, prestige, whatever), which in the short run doesn't mean much but which over time will be greatly and increasingly to their advantage.
danielharan 14y ago
Not quite. The fundamental attribution error "describes the tendency to over-value dispositional or personality-based explanations for the observed behaviors of others while under-valuing situational explanations for those behaviors."

pshin45 14y ago

All factors are not created equal though, and you must admit that there is probably a combination of a few key factors that help explain most (but not all) of their success, no? (http://en.wikipedia.org/wiki/Pareto_principle http://en.wikipedia.org/wiki/Pareto_principle) I think that is what they/we are trying to get at. I agree with your general principle that causation is rarely as black and white as we make it out to be, but saying that all/most phenomena is a result of hundreds of different factors coming together at the right time and right place would make it very difficult to ever learn anything about anything. > "..the falling cost of bandwidth, the falling cost of storage, the falling cost of computation, an increase in the number of internet users of nearly 1 billion over the last decade, a doubling in the global economy over the last 2 decades." These are all situational factors common to TechStars, 500 Startups, etc. but which don't explain YC's unique success.

confluence 14y ago
> very difficult to ever learn anything about anything Not really - if you are talking about investigating things that have a multitude of conflating variables affecting the end result. In that case you aren't learning anything either - because then everything is just correlation. When one investigates some arbitrary phenomena, it is important to constrain one's experiment and control all but one variable so that one can be sure that the changes one observes are more likely to be due to causation rather than correlation. For example: Stating that Africa is a failure because of X thing is full of peril. However, stating that when people don't have easy access to safe drinking water, they are more likely to die of water borne illnesses, and that this has probably contributed to the failure of Africa is a much more secure statement. It's not nearly as pithy, and it includes many qualifiers - but then again, it is a much more accurate and true assessment. It is also much less likely to be subject to vague/"woo"/dumbass business like thinking. > These are all situational factors common to TechStars, 500 Startups, etc. but which don't explain YC's unique success. First mover advantage anyone? Path dependence is a bitch - just ask DVORAK.
- pshin45 14y ago
  I don't disagree with what you're saying. If I were to sum up why YC might have succeeded like it did, I'd say that pg is great at pattern recognition and is better than most at figuring out who might become a great entrepreneur vs. who is just posturing, which increases (but doesn't at all guarantee) his chances of finding great founders capable of building great companies. Of course, things played out pretty well for pg, but with less luck perhaps it might've taken a longer time for them to find their Airbnb/Dropbox, or perhaps Drew Houston and the three co-founders of Airbnb might have failed several times before hitting their respective home runs. In which case TechStars and 500 Startups (admitted copycats of YC) might never have come into existence. pg's knack for identifying great founders doesn't guarantee that everyone who gets into YC is going to be "great," just like there are probably plenty of not-so-capable people in Harvard who somehow still managed to get in. But I would posit that the % of potential great founders is significantly higher at YC than at other accelerators (due to brand, prestige, whatever), which in the short run doesn't mean much but which over time will be greatly and increasingly to their advantage.

danielharan 14y ago

Not quite. The fundamental attribution error "describes the tendency to over-value dispositional or personality-based explanations for the observed behaviors of others while under-valuing situational explanations for those behaviors."