Ignoring whether correlation is the causation here, this post demonstrates a confirmation bias. Id like to see another list of successful companies with equal cofounders to be more objective.
I just recently saw a presentation by a guy out of Harvard at the University of British Columbia. He found the a full 1/3 of High-Tech startups divide the equity equally and that an equal division was a predictor of future trouble later on. Insert generic correlation is not causation critique here, but still the results are interesting. We were under direction not to distribute the paper otherwise I would link to it.
I figured as much. :-) Just a thought I had the other day. I imagine it's difficult, if not impossible, to prove a relation between any numerical metric on this site and start up success.
Revenue of a startup is no where close to being a reliable measure for such a small sample size. Give me any variable and I can make a similar graph to prove anything you want. Correlation != causation.
I like the idea, but how do you apply this to power law distribution outcomes and get any statistical significance? I don't know the answer.
E.g. the underlying First Round's analysis likely has no statistical significance. Assuming the power law distribution of outcomes top 5 outcomes will account for 97% of value. So we now have a study with n=5.
To make the point let's apply this to YC's own portfolio. Assuming Dropbox, AirBnb and Stripe represent 75% of its value, we'll learn that YC is incredibly biased against:
* MIT graduates
* brother founders
* founding teams that do not have female founders
* and especially males named Drew
Hard to believe these conclusions are correct or actionable
The analysis is only looking at whether there's a strong negative correlation between initial funding and startup success, using exit valuation as a measure, as this was Fred Wilson's assertion.
Of course, this is not meant to be predictive but rather meant to dispel the notion that a strong negative correlation exists between these two variables.
Not necessarily, as the distribution of startup returns is highly skewed and you don't know the correlation of his model prediction vs relative return.
Ie, if his model consistently predicts success of those companies deep in the right tail (ie, your Facebooks and Twitters), but is less correct about just 'mildly' successful startups, the algorithm will greatly outperform a coin flip.
This. Seen the same thing in 3 startups. Add me to the anecdata.
It would be awesome if we could get statistically significant numbers of people to report on this effect, although how to avoid selection bias etc. is beyond me atm.
The investors have “good data”. But even they know that only 10% of startups succeed. But HN and Reddit are full of wanna be analysts who think they can predict which startups will succeed if they have enough information.
Trying to value equity as employee or will it ever be worth anything might as well be buying books from the supermarket with winning lottery numbers.
If they just predict every start-up to fail, statistically, they would be 90%+ accurate :P
Predicting the success of a start up is probably just as hard as predicting where the market is going to go tomorrow (if not harder). There are too many variables in play. I think this is just a PR stunt; it definitely created enough buzz!
From an investor point of view, correlation can be good enough. If startups with property X are measurably more likely to succeed, it doesn't matter whether X causes success or if X is caused by some other property Y of the startup that causes success but isn't easy to determine in an interview.
In your example, X = "has multiple founders", Y = "first founder is not impossible to work with". Clearly Y causes success, but X is easier for investors to determine.
From a founder point of view, you want to focus mainly on the real causes of success rather than appearances, but give some thought to appearances. Customers are far more influenced by appearances than investors (because they have less time to dig in), so when you devote time to appearances you should think mainly about customers. Customers tend to be skeptical of one-man shows.
This is really interesting. I'm surprised you found that startups with higher centrality raised less money. Not sure how to explain that. Your investor graph has edges pointing both from investors to startups and back, right? Anyway, thanks.
It would be a start just to find an industry where some participants are VC funded and some aren't, and then look at how they each fared.
As appealing as such a question might be in abstract, I think the results would be pretty useless in practice. There's too much diversity, too many variables. Every startup is at rather incommensurable with every other, beyond the basic financial anchors. Every product's success is a highly stochastic combination of team, market conditions, direction of winds, luck, etc.
What would such a study--regardless of "scientific" rigour, which, if desired, may itself may be an unintelligible criterion given the complexity of the question--really tell us?
More than anything, it proves that you can use quotes and statistics in a manner to confirm your own biases. :) The decision tree has far too many branches to predict where the starting point will lead you but my point was, even when the recipe of successful startup (build something people want, talk to people, start small ... etc) sounds so obvious, the process of accomplishing it is anything but obvious.
How plausible is it that if you take any startup and populate it with leetcode winners, the outcome will be a high stock price?
reply