To: Michael Bakunin who wrote (85497 ) 11/15/2000 2:21:24 PM From: Don Lloyd Read Replies (1) | Respond to of 132070 mb -...Then, you arbitrarily adjust ratios "inwards", without noting the very, very odd fact that while the Bush/McCollum ratios are as expected, i.e. greater than one in accordance with the usual ballot falloff, that the Gore/Nelson ratios are anomalous at less than one. What, in all the counties more Democrats voted for senator than for president? (A couple-percent Nader vote doesn't explain a 0.9 ratio when the expected number is greater than 1.) Where you see these numbers implying missing votes for W, I see something awfully strange in the Gore ratios. What's the overall ratio in each county of votes for president to votes for senator? Do you have another explanation for the huge disparity in ratios? ... Thanks for making me look to answer your questions. Going back to full precinct data for Palm Bay, and including ALL votes, There were 435,638 total Senate votes and 432,286 total Presidential votes, 3352 less votes for President. This is a ratio of 0.9923, undoubtedly due to rejected Presidential votes outweighing ballots with no Senate votes. For the data I used, including only the Democratic and Republican candidates, the ratio of Presidential votes to Senatorial votes for each county varied between .9576 and 1.3234. Overall the ratio of Presidential votes to Senatorial votes was 1.0258. The reason that this differs from the lower ratios you noted for Gore/Nelson is a combination of third party Presidential candidates and the fact that Bush must draw more Nelson votes than Gore draws McCollum votes. This only adds to the hypothesis of missing Bush votes, subject to your comments. Regards, Don