서사

“Statistical anomalies in the results prove the election was rigged”

Election data is lumpy for ordinary reasons — precincts differ in size, rules, demography and when they report — so almost any dataset yields something that looks anomalous to an analyst who starts from the conclusion. The recurring instances all dissolve on inspection: US overnight "spikes" were batches of mail ballots counted after election-day votes in states that forbade processing them earlier, and hand audits matched or widened the certified margin. Korea’s stable early-vote gap reflects two different voter populations. A French national vote share cannot be compared with a national seat total when seats come from 577 separate races. This is the category the site separates most carefully: an anomaly is a reason to examine records, never a finding on its own.