Haffner (1349071) writes "Commenters over on reddit seem to believe that there exists a statistical anomaly in voting records for Romney. They've shown that for large precincts, Romney's cumulative vote distribution increases linearly over time — a statistical anomaly — and that this happens many times. Is this good evidence?" Link to Original Source
Physician: One upon whom we set our hopes when ill and our dogs when well.
-- Ambrose Bierce