How house effects work

A house effect is a recurring gap between one pollster and the pack. If YouGov has lately shown Reform two points lower than the average of other firms, that is a house effect — not proof that YouGov is 'wrong'.

ElectaMeter stores a documented starting set of offsets in the pollster_adjustments table. When you enable the toggle, those points are subtracted from each poll before the quality-weighted average. The offsets are not a claim of true long-run bias.

Leave the toggle off if you want the raw published numbers. Turn it on if you want a smoother read across firms. Either way the formula is on the Methodology page and the offsets are visible there.