Eric Raymond (the one responsible for the botched analysis of the stolen CRU code) responds to my post on Essex and McKitrick’s error in treating missing values as zeroes in a spreadsheet:
The error described is so stupid that I have trouble believing a statistician actually made it. Whether McKittrick understood thermodynamics or not is red herring; even somebody with my non-specialist knowledge of statistics alone would have known better, let alone a pro like him. The most plausible theory I can think of is that the spreadsheet was expressing temperatures as deviation from mean, that the “zeroes” actually pegged missing observations to that mean, and that the author misunderstood McKittrick’s response.
Raymond clung to his theory even after a commenter pointed out that I included a link to the spreadsheet and that you could easily check that they counted missing values as zeroes.
He also uncritically accepted David Bellamy’s story that the BBC cancelled his show when he came out as a global warming denier, even though Bellamy’s TV career ended in 1994 and he didn’t deny global warming until 2004. This too, was pointed out by a commenter, but Raymond did not change his position.