Comment by kragen

It's hard to correct completely for survivorship bias, but it still holds very clearly for very prominent recent papers and for other papers that happen to be in the same journal issue as the old paper I went there looking for.