I have no involvement with YC, but with a little startup and VC pitching experience, I can tell you that in my experience, lots of VCs like the idea of founders doing something incremental that builds on successful ideas. The article notably did not compare YC to any other VCs, but the truth is likely that all VCs “often” back startups that duplicate others in some way. VCs will tell you not to build something original, because there’s no demonstrated market for it. And statistically they’re right, your chances of success with an unproven idea are lower than with an incremental change to something people already buy. The article also notably did not talk about how often startups duplicate other businesses on their own, before there’s VC involvement. In that sense, despite the claim in the title that there’s data, the article is unscientific.

This article struck me as one of those “studies” that shows something everyone already knows, and the title kinda seems designed to sound dramatic and/or accusatory to appeal to readers drawn to controversy even if there isn’t really any controversy there.

YC seems to have a spray-and-pray approach, and it used to be run by Sam Altman who has repeatedly failed upward, so I think it's very reasonable to assume this is either not a conscious strategy or it's just a bad strategy.

Either way, the VC's value is to be able to predict whether Dropbox or Zumodrive deserves their bet, and they clearly couldn't.

This is an extremely wrong-headed view of what VCs do, and one thing investors do _in general_ is to have a strong idea of what they know and don't know, and in particular, what _nobody_ knows is which companies or products in particular will succeed or fail. If they knew that, they'd put all their eggs in one basket.

What VCs do is allocate capital in a way that mitigates risk for themselves.

There's actually a _really_ interesting way to think about what VCs do, which is that they _offload_ their own risk onto founders and early hires of startups. VCs invest their money across a broad basket of investments, founders invest all their time and money into _one_. VCs and early hires are taking a massive amount of personal risk. Almost all of the profits of VCs come from what is essentially a risk arbitrage -- they get more profits than they should be from the smaller risk they're taking by investing in a startup, and founders and early hires get less profits than they should be from the personal risk they're taking by starting a company.

The structure of investment deals is often setup in such a way that even events that "feel" like they should be a payday for the founders, such as a funding round or even a sale, could end up with them getting nothing because their shares get diluted, because they have lower priority ownership stakes than the VCs do.

But most VCs are paid to make bets on which companies are most likely to succeed. I never said they're supposed to know with certainty. That's a silly straw man that no one who understands basic finance would suggest.

But they spend much of their time supposedly screening deals to find good ones, and there is a preponderance of evidence that they're worse at it than a dumb index fund.

There have even been experiments to automate picking deals, and even a fairly naive algorithm is better than most VCs.

Only people on the VC side of the table think they have any added value other than lucking into being trusted to invest other people's money.

I think you meant “Founders and early hires” here?