4 ms·
The trick is to ignore survivorship bias every step of the way
by Apocryphon 6y ago
The trick is to ignore survivorship bias every step of the way
- renewiltord 6y agoOh no, it's the opposite. Established companies could be established for no good reason. ULA was in space, they're the experts, but they could be beaten. The incumbent could be just a result of survivorship bias. That's what presents an opportunity. If the incumbent is just very good at the product, the only part you can disrupt is margin (or verticalize or something, but lets set aside that) and you don't want to be in a price war unless you're uniquely suited to it. So you don't fight those guys. So you count on the market leader being disruptable, which is often because they're just there as accidental survivors.
- Apocryphon 6y agoWhat you write makes a lot of sense in terms of successful startup strategy. It doesn’t detract from the vacuity of your original statement on “high variance“. When you cherry-pick two data points- from the same founder, even- while ignoring the high rate of failed startups, it makes for quite an empty fallacious statement indeed. One might as well argue for the merits of big government statism by singing the praises of the Manhattan Project and the Apollo program, while contrasting them with the DMV.
- renewiltord 6y agoWell, yes, you can and you'd be right. That's how existence proofs work. State funded projects make great moonshots, yes, and you have some great examples. It's a great argument for Big Government statists that governments are capable of large scale technological progress out of the reach of common corporations.
- Apocryphon 6y agoBut no one was disputing the existence of successful, even wildly successful startups. People are criticizing startups that behave ignorantly and arrogantly. (And quite frankly, give the entire scene a bad name.) You are attempting to play devil's advocate by pointing out that some flout conventional wisdom and established practices and succeed. Very well, we can all agree that those enterprises exist. The question is if they exist in any appreciable number, which is very debatable- again, due to the simple arithmetic of very low startup success rates. You are celebrating lottery winners. The existence of that very few does not justify bad behavior from (some of) the rest. Your proof is facile.
- renewiltord 6y agoIt isn't bad behaviour. It's just a failure. No harm in trying things that don't work. It doesn't look anyone misrepresented themselves. They came in, sucked, and got taken out by the market. The fact that they didn't listen hardly matters. It's not bad behaviour to not listen to advice from your landlord. I'm not playing Devil's Advocate. I genuinely think it is good that people are trying different things. They're exploring the state space and I enjoy that. If they were actually misrepresenting themselves or their ability to pay, or reneged on their debts without going out of business (which is a well-understood way to renege), or something like that, sure. But it looks like everything is on the level and the article is just saying that all of these guys who tried it weren't good at it. That's not a crime or even morally problematic.
- Apocryphon 6y agoPeople aren't saying they're behaving criminally. If anything, their behavior is both mockable and foolish- out-of-touch, operating in an SV bubble, detached from the needs of real-world people would their businesses ostensibly serve, and perhaps representative of the excess that dumb money empowers. They're gauche. (And yes, during times of mass unemployment and rising economic uncertainty, moralizing does start to creep in; when mom and pop stores are dying by the tens of thousands, VC-funded fail does seem like bad taste, at least. Fair's got nothing to do with the creation of this perception; but optics are optics.) Is it wrong to have higher standards for this industry? Less sloppy business practices, more empathy for the world at large? More curiosity and respect for established wisdom? What's the point of learning from "fail fast, fail often" if the failures are just going be the same mistakes over and over again, caused by foolhardy founders and negligent investors? Infinite samsara wheels, doomed by hubris.
- numpad0 6y agoI’m not so sure. When it comes to manufacturing, attempts at cutting corners seems to just disrupt startups themselves. Only in business models the break things philosophy seems to work. idk though