It's certainly much worse. It's often difficult to tell when a company is dead unless they announce it.
Many other failures (likely most of YC's portfolio) will have died by being acquired for less than the most recent valuation, meaning YC lost money on the deal. These sometimes look like successful exits unless you know the details.
Companies also almost never announce their death. Startups go to die quietly. Paul Graham says the best indicator is they don't reply to emails, but this is usually something that's not very visible.
That's one signal, but it's surprising how many established companies that appear to still be going concerns also don't reply to emails even when I literally ask about buying something. And I'm not talking about my own little consumer purchases, I mean emails from my work account to vendors with the potential for significant ongoing revenue. Strange.
Also, it's not counting many of the startups from the last few years which haven't had enough time to fail. If the denominator is the total number of exits then it's probably much lower than 4k.
Zombie mode: "Others are stuck in “zombie” mode — surviving but unable to grow. They can muddle along like that for years, investors said, but will most likely struggle to raise more money."
My uncle has had his startup kickstart with investors money and then paid them back. Now he is slowly growing but not expanding the company only the assets. I wouldnt call this dead. Its just sometimes you hit a ceiling for expansion
wiki says 4K, so 821/4K = 20%. not too bad. I expected much worse.