Comment by ryanwaggoner | Hacker News Reader

Comment by ryanwaggoner | Hacker News Reader

Parent Full thread

I'm not talking about working hard at all; I think success pretty much always requires hard work of some kind. My friend definitely worked hard to get his business up and running and to the point where it supports him.

I'm talking about the idea that you have to work so hard that it threatens or consumes everything else in your life. And no one claims that doing that is sufficient, but a lot of people claim it's necessary.

It probably is necessary if the plan is to hit a billion-dollar exit in a few years. I just think that path entails a lot of extra risk that investors overwhelmingly benefit from, and founders end up paying for in the event that things go south. And sadly, I feel that a lot of investors essentially exploit young founders who don't know enough yet to know that they're paying a very high price for a 1/10000 chance at a grand slam (making $20m in 2-5 years), when they'd be better off in almost every way by going for the 1/10 or 1/100 base hit (making $2m in 5-10 years).

And to be clear, I'm not lumping YC in with most investors here. However, YC does place a little too much emphasis on raising VC and not enough on building revenues. But that's a bubble for you :)

Your reasoning exhibits another common fallacy. While the average probability of making $20m in 5 years may be low, that doesn't imply each individual's is. For the right sort of person the probability might be 2/3 or even higher. For most people it's epsilon. The reason the average is low is that it's the average of a lot of epsilons and a handful of higher numbers.

This is easier to understand if you consider e.g. the question of being over 7 feet tall. Maybe only 1 in 10000 (or whatever) people is over 7 feet tall. But that doesn't mean your probability of being over 7 feet tall is .01%. Your probability is either 0% or 100%.

Isn't this only valid in hindsight? Isn't your probability of anything 0% or 100% once it has happened or not happened? The problem here is that we have no idea in advance which bucket you'll end up in.

So yes, I suppose some founders have a very high probability and some have almost no probability. But that's not useful when you don't know which group you're in yet. So the average actually IS useful.

You have either 0% or 100% probability of being over 7 feet tall, but what about your unborn child? Genetics aside, if 1 out of 10000 kids are born over 7 ft and you have a gun to your head, how will you calculate your kids odds of being that tall?

An individual himself can't predict whether he or she will be successful or not, but someone who's seen many such individuals in their "before" state and in their "after" state can make more accurate predictions.

For example, YC's whole investment model is based on the assumption that it's not just hindsight, and that the YC partners can use the data model that they have, that gets better with each batch, to predict which founders fall into the "success bin", and they choose to invest in those.

YC's record is a lot closer to 1% than 100% of getting huge homerun hits.

By your logic, this implies that in the best possible circumstances (YC), you're not going to get above a few percent chance a priori of a huge success if you work like crazy.

Whereas an equally intelligent person working reasonable hours on the side when appropriate has a decent chance of making a profitable lifestyle business in the millions without the possibility burnout and debt.

I am not talking about "getting huge" at all, because I don't believe there is enough space in the market for 300+ (number of YC investments so far) huge (Google/Apple/Facebook/Microsoft) companies. I am, however, talking about non-trivial exits, ones that definitely put a founder in the "success" bin. Let's say, 10+ million in 4 years. I don't think it's possible to get to that point "on the side".

Don't use generalizations, just do the math.

$10 million, minus the VC investment, split each among 2 or 3 founders. That ends up being what, $2-3 million in 4 years? I personally know several people who make more than that off of one iPhone app on the side. Apple's paid out billions to app developers.

Again, look at the numbers: what percent of YC investments have exited at $10 million+ in 4 years? Again, much closer to 10% than 100%. This is in a thread where pg implies that the chances of a big success are 2/3 or higher for some people. Empirical evidence proves that top investors are not right anywhere near the probability suggested.

Investors are incentivized to not only pick the founders they think have a very high chance of huge success, but to make as many founders as possible think they belong to that group. If they're right, big rewards. If they're wrong, the founders (and LPs) pay the price.

More: http://www.paulgraham.com/venturecapital.html

Great point. And founders who have taken the 90-hour-a-week path are emotionally incentivized to believe it's the only path to success -- because if not, they could have made different tradeoffs.

The VC firm in business for the fees was mostly a product of the Internet Bubble of the late 90s. There were still a significant number around when I wrote that 6 years ago, but in the last few years (and particularly since 2008) the bad firms have been ruthlessly pruned.

No VC whose name you recognize will invest in bad startups to get the management fees.

I don't think that's any different from most other institutions that "accept" people for challenging endeavors. If you are accepted into medical school, work insane hours, and then fail out, you still have the loans to pay back.

But it is possible to tell which group you're in. Nearly everyone correctly estimates their probability of making $20m in 5 years at a startup at near zero, and responds by not starting one. A few people (e.g. Max Levchin) can with equal ease correctly estimate that the probability is high enough to make it worth doing. There are only a few borderline cases of people who mistakenly believe they are the right sort of person to start a startup when they aren't. They're overrepresented here, but they're a tiny fraction of the population.

Even if you're one of those people, you can do way better than falling back an average. You can for example ask experienced investors. It's part of their job to judge the probability that you'll succeed.

I think that makes sense. It'd be kind of like saying broad market indexes (averages) don't matter, for only individual stocks' have effect on stockholders' capital -ignoring indexed funds.

"In a down market," for example, "the index's direction doesn't matter if you have the right stock which is going up".

Being interested in social topology, I wonder if there's a parallel here to election/predestination; anyone else see it?

I think that fallacy should be called the anthropic fallacy (I don't think the fallacy has a name yet).

The anthropic principle is that even though the number of planets hospitable to life is tiny the probability that humanity evolved on a planet that is hospitable to human life is 100%. The examples you gave are symmetrical.

Reply on news.ycombinator.com