How the Survivor Bias Distorts Reality
scientificamerican.com
scientificamerican.com
Most are easily intelligent enough to realize how long their odds really are. Furthermore, they have the intellect to appreciate the true effort they have to make before they reach financial payback.
Any reasonable person with the math skills to visualize risk vs. reward will usually walk away --but not the quintessential entrepreneur.
This is not a flattering trait. For this reason, it is rarely, if ever, associated with successful entrepreneurs.
> By necessity, entrepreneurs are self-delusional.
Unlike VCs, entrepreneurs have no spread of risk. It is important to realize that it is in the interest of the VC that the entrepreneur remains delusional.
A successful developer, for example, starting their consultancy/agency has a high chance of building a successful venture.
When analyzing why a group of objects (usually very small, like successful companies) have a certain trait, it is not enough to find something all these objects had in common and say that it was the reason. At bare minimum, you have to look at all the rest of objects and confirm that they did not have that thing in common too. And even that would not be enough. After finding something ONLY these objects of interest have in common, you should test the prediction on new objects whose outcome is not yet known and confirm the accuracy.
It sounds really simple, but sometimes even scientific paper writers can't resist urge to release "finding" that has hidden flaws like that.
But it's enough to sell a book claiming so.
> At bare minimum, you have to look at all the rest of objects and confirm that they did not have that thing in common too.
That would rule out way too many profitable books. /s
But this assessment itself seems to suffer from a bias of another sort--the notion that entrepreneurial risk is uniformly distributed over the population of entrepreneurs. Actually, there will be some startups whose risk will be much lower than that of the "average" emerging company (due to the fact that their venture idea satisfies some market need, discovered either by genuine insight/ingenuity or by luck).
It is the belief--however appropriate--that one's own venture falls within this enlightened category of diminished risk that propels founders to pursue their ventures in the face of such an aggregate track record.
.. or by looking at what has succeeded in the past and trying to learn from it, and extrapolate those lessons to the present.
Employer: "Clearly it isn't stupid because every awesome person we have employed has passed that test. In fact, the test is awesome and is the reason we have managed to build such an awesome team."
From the article: "Instead, Smith says, Collins should have started with a list of companies at the beginning of the test period and then used 'plausible criteria to select eleven companies predicted to do better than the rest. These criteria must be applied in an objective way, without peeking at how the companies did over the next forty years. It is not fair or meaningful to predict which companies will do well after looking at which companies did well! Those are not predictions, just history.'"
In the actual research on company hiring procedures, the correct research procedure is used by giving tests BEFORE job seekers are hired, and then seeing how they do as workers over time. Industrial and organizational (I/O) psychologists have been doing this kind of research for about a century, in countries all over the world, testing hundreds of thousands of workers in numerous job categories in hundreds of workplaces in dozens of countries. Hiring job-seekers based on work-sample tests and tests of the job-seekers' cognitive ability reliably improves worker performance over the long term. There is a FAQ on this topic[1] with links to the research literature you can read for more details.
In analyzing these trends, you're biased by looking only at the companies that have done well enough to be on the current list. In this way it's particularly insidious, because you're omitting the very data that might save you from betting on a company that might be delisted.
So you are saying if I become an entrepreneur, I have a 1/100 chance of being in the top 1%? Wow, that's great! Sign me up!
checks math
Hey, wait a minute!
C = with specific characteristics
What we need is P(S | C) = P(S ^ C) / P(C), but the best sellers mentioned in the article instead address P(S ^ C)
Even very intelligent people seem to have far too many delusions about their own competence or success.