Statistics may save us
Would that be the smoking gun?
Lots of work going on in the background. Statistics may be the one tool that bring us the proof, if that election has been rigged. In Florida, people have uncovered a disturbing bias between the use of optical ballot scanners and results favoring Bush.
Look at this:
The orginal figure and full discussion is here. What is plotted here is the % registration difference, i.e. relative to the political affiliation of the voters (Dem vs Rep) vs the % voting difference, relative to the actual voting numbers. Explicitely,
%Votes Difference = (Republican Votes – Democratic Votes) / Total Votes * 100%
%Party Registration Difference = (Republican Registration – Democratic Registration) / Total Registration * 100%
The blue points plot E-touch screens results, and they follow the expected trend: a good correlation with a slope of one. The red points are for Opt-scan precincts, and there the behavior is radically different! An inexplicable bias. Again, this page (other data here) present the analysis in greater details and split the sample to look at tendencies with other variables (like county size).
