Polls

Bayesian aggregation

The displayed scores combine results from multiple polling institutes. A Bayesian model weights each poll based on four criteria: the institute's historical reliability, data recency, sample size, and methodology used.

Confidence intervals

Hover over a candidate to see the 95% confidence interval. A wider interval indicates higher uncertainty — often due to a low number of polls or divergent results between institutes.

Trends

Trend badges show the 7-day evolution. A green arrow indicates a rise, a red arrow a decline. These variations should be interpreted cautiously when polling volume is low.