Survivorship bias: many IPOs since then have underperformed. If you cherry-pick all the ones that did, then IPOs look like an outrageously fantastic bet.

You have to bet on the IPO knowing only the information at the time, not knowing what the future growth of the stock was.