3 ms·

The falsehood here is in your claim that they said they would only invest in the winning companies. The NYT article you reference doesn't say anything about th

The falsehood here is in your claim that they said they would only invest in the winning companies. The NYT article you reference doesn't say anything about them wanting to invest only in winners. What they've always said is that they need to invest in all the companies that might be winners, knowing that many of them will turn out to be losers, and that in order to ensure they got all the promising companies in their portfolio, they need to be highly founder-friendly. Indeed, the strategy has a self-determining aspect to it: in order to invest in more winners they also need to invest in more losers, and by ensuring they invest in all the winners, they can afford to fund more losers, which helps to ensure they capture more winners. Andreessen explains this thoroughly in his EconTalk episode from a couple of years ago [1]. That NYT glossed over this point is their doing, not Andreessen's, but even then you've claimed something that isn't at all true and dang is right to point out that you're capable of much better. [1] http://www.econtalk.org/archives/2014/05/marc_andreessen.html http://www.econtalk.org/archives/2014/05/marc_andreessen.htm...

jacques_chester 10y ago
> What they've always said is that they need to invest in all the companies that might be winners, knowing that many of them will turn out to be losers, and that in order to ensure they got all the promising companies in their portfolio, they need to be highly founder-friendly. The key phrase here is "all the promising companies". This is the same as "I will bet on the promising horses". It's not a novel strategy. And how are the promising companies selected? Either a16z selects them, in which case they are trying to pick winners, taking me back to my caricature from earlier. Or they base it on a shotgun approach, selecting anyone who shows up. Then the selection function becomes "which jockey thinks they have the race-winning horse?", which is just another variant of picking winners. I bounced up and down the long (and very interesting, thank you) interview looking for something more. The closest I found to a thorough explanation was this: > So, basically, here's the way to think about it in the math. There's about 4000 startups a year, [?] a year, that want to raise venture capital. Of those, maybe 400 will get funded by top venture capital firms. Of those, about 15 will be responsible for over 90% of the profits for that entire year of companies. Which turned into a conversation about Google which I found myself nodding along to. Then: > And so it's one of the things, again it goes back to what's humbling about what we do, is--it's limits of knowledge. There are real limits to what you can know. Which, by the way, means that if you are going to operate in the field and if your requirement for investing in something or backing something or going to work for something is you are going to know for sure that it is received[?], you are never going to do anything. There is no return above the risk. Which is a much more reasonable position. So now I'm wondering if my whole recollection really is based on the initial publicity. > dang is right to point out that you're capable of much better. This gives me flashbacks to every school report I ever got. I'm capable of better and worse. It generally depends on the weather that follows me around.

jacques_chester 10y ago

> What they've always said is that they need to invest in all the companies that might be winners, knowing that many of them will turn out to be losers, and that in order to ensure they got all the promising companies in their portfolio, they need to be highly founder-friendly. The key phrase here is "all the promising companies". This is the same as "I will bet on the promising horses". It's not a novel strategy. And how are the promising companies selected? Either a16z selects them, in which case they are trying to pick winners, taking me back to my caricature from earlier. Or they base it on a shotgun approach, selecting anyone who shows up. Then the selection function becomes "which jockey thinks they have the race-winning horse?", which is just another variant of picking winners. I bounced up and down the long (and very interesting, thank you) interview looking for something more. The closest I found to a thorough explanation was this: > So, basically, here's the way to think about it in the math. There's about 4000 startups a year, [?] a year, that want to raise venture capital. Of those, maybe 400 will get funded by top venture capital firms. Of those, about 15 will be responsible for over 90% of the profits for that entire year of companies. Which turned into a conversation about Google which I found myself nodding along to. Then: > And so it's one of the things, again it goes back to what's humbling about what we do, is--it's limits of knowledge. There are real limits to what you can know. Which, by the way, means that if you are going to operate in the field and if your requirement for investing in something or backing something or going to work for something is you are going to know for sure that it is received[?], you are never going to do anything. There is no return above the risk. Which is a much more reasonable position. So now I'm wondering if my whole recollection really is based on the initial publicity. > dang is right to point out that you're capable of much better. This gives me flashbacks to every school report I ever got. I'm capable of better and worse. It generally depends on the weather that follows me around.