To: Skeeter Bug who wrote (85697 ) 11/20/2000 12:36:18 AM From: George Acton Read Replies (1) | Respond to of 132070 Somewhere in my exposure to election news, there was a report from the National Bureau of Standards comparing different methods of counting. Paper ballots with scanning, with hand counting and lever machines were comparable, with about a 0.3% error rate. The rate with punched cards was 3%, mainly because of chad. This makes a good case for hand counting here. Hand counting in the real world might not be as accurate as in the controlled conditions of the study. OTOH, it might be better because people would be paying more careful attention. Bush's margin is 0.00015% of the votes cast, or to express it another way, 1.5% of 1%, a ratio of 1:20 to the best error rate in the study. This is so far below the noise level of the counting process that we will never know who won. It's like trying to weigh an object to an accuracy of 50 milligrams with a scale whose finest gradation is a gram. The best way to legitimize the election appears to be hand-counting the whole state, or perhaps only counties that used punch card ballots. It would have been better if a week ago the candidates had agreed to this and promised to abide by the outcome. I hope the Florida Supremes come down that way tomorrow, although it would have been better if the candidates had negotiated it. In passing, I note that the "Bureau of Standards" has been blanded down to the "National Institute of Standards and Technology". The current rot set in when they replaced the "War Department" with the "Department of Defense". Since then, we haven't won any WW II-type wars that I can recall offhand, and we lost a medium-sized one that wasn't, by many standards, even defensive.