<<Dear JC: I am sorry but you dont understand how these readers for punch cards work. You ALWAYS pick up more votes when you hand read compared to punch card machine read. Thats because cards might be stuck together, chads folded over etc etc HOWEVER, it is believed that since the machine knows no party the errors are statistically balanced out when involving a large population. The catch is that when you ONLY hand read a few COUNTIES EACH OF WHICH IS SKEWED TO ONE PARTY then the results will be skewed to that party. ie if you pick up 100 additonal cards and one party has two thirds of the registration, chances are THAT party will pick up 66 votes. Fla. is roughly divided equal so to be fair would require handcounting the entire state, an impossible condition. JDN>>
It seems you're the one who doesn't fully understand, JDN. Different counties in Florida use different vote counting systems. And guess what, the ones in the big Democratic counties are 16 times more error prone than the ones in most of the Republican counties...
This from Salon:
"At a Tallahassee press conference on Saturday, Florida election officials revealed that in last Tuesday's election, there were far more problems in counties using the punch-card system than the optical scan system, or OpScan. In punch-card counties, 32 ballots per thousand cast had to be invalidated. But in counties using the optical scan system, or OpScan, only two per thousand were thrown out. Around Florida, and throughout the county, election reformers are trying to move away from the punch-card system. Massachussetts has stopped using them entirely; so has San Francisco County.
Coincidentally or not, Florida counties using the OpScan system tend to be Republican, while the punch-card counties are disproportionately Democratic. And one Republican county still using punch cards, Seminole, voluntarily hand counted its ballots when it conducted its state-mandated recount last week -- and gave another 98 votes to Bush, his largest county gain in the recount."
Steve Dietrich |
