I assert that they aren't apples and oranges becauss YC has picked every single unicorn there is to pick up until very recently. Therefore, both the 2,000 number from A16Z and 19 number from YC can be directly compared.

Do you think it is significantly easier to pick unicorns than I am describing? Because it seems very, very hard.

So hard, in fact, that YC has to pick 150 or so per year and hope that one or two hit 1000x to make up for the others.

I wish YC could have funded every unicorn! In reality there are a few hundred such companies worldwide, so it's less than 1 in 10.

https://www.cbinsights.com/research-unicorn-companies has a list.

Old tweet but here is where I was basing my "YC funded every unicorn" assertion (as of 2015)

https://mobile.twitter.com/rabois/status/634205368172814337

The tweet you referenced is talking about companies that actually applied to YC. Uber, for example, never applied to YC.