No, but it does mean that at an individual level the difference is not consequential.
a16z and Sequoia might be able to pick the winners at a higher rate than other investors, but that advantage only bears itself out with a large N of companies.
At the individual case, with a single individual looking for a single job in a single company, the difference between being to pick winners 5% better than the baseline is not consequential, since the expected outcome is still "not a winner".
Some people may be able to pick winners better than others - but the effect is not strong enough to be useful in the micro scale because the odds of success are still very, very low.
There is something not right about this argument. Do you really say to yourself: “Here are two companies A & B with the same valuation. I think A has the best chance of success based on its technology, but I don’t care which one I join since they have the same chance of success.”
For the engineer, they may be able to choose wisely between company A and B. But if company A has a 1% chance of getting rich and company B has a 5% chance, it's still only a 5% chance. So any skill in choosing a winner is diminished by the fact that winning is still unlikely.
VC's however, can place many bets and come out winning if they play enough.
Because this is all high variance, a large sample size favors those with skill. Small sample size favors luck.
Because an engineer will work at many startups over his career these seemingly small differences add up. Let assume the chance of not hitting the jackpot is 99% at companies like A and 95% at companies like B. After working at 10 companies you would have a 10% of hitting the jackpot following the company A path and 40% if you go down the company B path. This is pretty significant.
I think a 5% chance of winning is unrealistic for your math, at least without considering payout. Tie the results to the payout and the expected value should show that working at a start up with the idea of making money via equity is foolish compared to Big Co. Let's take that 1% one and say you get $500k if it works out in an IPO. And let's say you take a salary below market of $30k. It will take the start up 7 to 10 years to IPO. You gave up maybe $300k for the chance at $500k. Or the 5% case where the company exits early and everyone gets $100k. Maybe that takes 3 years. You gave up perhaps $90k for a chance at $100k. You can play start up lottery several times in your career, but equity, unless you are getting a very significant portion, should not be a large factor in choosing the start up you work for. Base salary, the environment, culture, learning opportunities, and the ability to make a difference in an organization are reasons to choose a start up. Just not equity. And that's ok :)
Sure, but the odds aren't 10% or 40%, they're more on the order of 1% to 3%. If you can pick winners at a 40% rate, you're doing orders of magnitude better than even the top VCs in the industry, and your talent is worth billions.
Realistically nobody plays at those odds, the actual odds being played are much, much lower.
Which also means that, even through a long career, you're not taking on enough jobs to get a reasonable shot at a win - 10 jobs with a 2% chance of winning each aren't good odds - you have a 18% chance of scoring a win.
All the while you're giving up a lot of money to do so - BigCo's will pay you easily $100-200k more per year. So even if you switch jobs yearly, 10 jobs = 10 years = $1-2M on the table to play for 18% odds.
Which is also why investors who are able to choose well win - YC for example invests in over 200 companies a year. Unless you're switching jobs every week you're not going to be able to raise N high enough to make a difference - at YC's scale a 2% chance of winning for each company means there's a 98% chance at least one company wins.
The entire idea here is that nobody has predictive power high enough to the degree of 10% vs. 40%. Realistically when choosing between two competing startups you're choosing success odds in the low single digits.
Not to mention the rash of articles in the past week about the low payouts you're likely to get even if the company succeeds. Engineers don't walk away from modern startups with 7-figure payouts, low 6-figures is more likely. So realistically even if you are able to pick companies at a 40% accuracy (again, astronomically higher than reality) your expected payout is lower than the $1-2M you're leaving on the table in order to play.
