Paul Hoffman, author of King's Gambit, sent me an e-mail yesterday in response to my review of his book. Here, in part, is what Hoffman wrote regarding my criticism that some of his anecdotes were inaccurate.
Before I wrote the footnote on Rossolimo's death, I researched it. I had remembered two conflicting stories from 1975. The one I ultimately reported, and the one that he was pushed. I tracked down people who knew him very well then, and they said there were no muggers or other people involved, so I went with that.
I am of course quite embarrassed by the Caro-Kann mistake, I've played the exchange Caro-Kann my entire chess life, so I have no idea why my neurons misfired and I wrote French Defense. Nor, obviously, did I catch it on subsequent readings. I'm not sure of the psychological phenomenon here, but I've noticed that sometimes when I make an obvious mistake, I keep on making it until someone else points it out. Which is why everyone needs a good editor. Some very strong chess players read the whole manuscript and caught much subtler stuff, but this too went by them. The good thing, though, is that when my book goes to paperback, I can fix errors like this. I've put a page on my Web site where I report corrections: http://paulhoffman.wordpress.com/kings-gambit-corrigenda. I think that you might find, if you wrote a book of 150,000 words, that some obvious-in-hindsight errors would slip in. I'm not proud of this, but it happens. Hey, I know very well what opening Noah played!
I can agree with Hoffman's statement about "obvious-in-hindsight errors" from my own experience with both books on the Philidor Counter Gambit. No matter how meticulously you proofread, you always miss something!