12 of 49 models show statistically significant differences in performance between US and UK English, with non-overlapping confidence intervals.
Seven models perform better in US English, while five favor UK English. The average difference across all models is just 7 Elo points, suggesting no strong overall accent bias.
