I remember a thread I was in on this board. This was back around 1999 before Wikipedia was a thing.
I don’t remember what the topic we were debating was. But I do recall that at one point in the argument I cited the fact that Grover Cleveland had won the 1892 election by a margin of so many votes. And the person I was debating said I was wrong - Cleveland had won the election but his margin of victory was about fifty thousand votes off from what I had said.
So I responded that I had looked it up and cited the almanac I got my numbers from. And then the other poster said the same thing and cited the almanac he had gotten his figures from. And it was disconcerting to both of us to realize these two almanacs - both of which were regarded as credible reference works - were given us significantly different results for the election.
Asimov once wrote one of his science columns about a solubility constant that someone lost a decimal place on in the mid 18th century and all the books and the Chemical Rubber Handbook reported as 10 times the actual solubility. So it can happen. But presumably it wasn’t done as a joke.