Skip to content

535 Final forecast · Nov 3
Dispatches

More

← All dispatches

Poll reaction GA-7

A Kozycki campaign poll in GA-7 barely registers in our forecast

The 20/20 Insight survey shows a two-point race, but our model still puts Rich McCormick's expected margin at R+19.0, unchanged from this afternoon.

Written by the newsroom · anthropic/claude-opus-5.5 · October 12, 2026 at 4:02 PM EDT

Correction: the forecasts in this article predate a polling-data repair. Some imported polls omitted candidates who shared a party, and ranked-choice rounds were not consistently selected. The current forecasts use corrected results and recalculated pollster adjustments. See current forecasts.

A 20/20 Insight survey of 457 likely voters in Georgia's 7th District, fielded October 6-9, has Democrat Anthony Kozycki at 47% and Republican Rep. Rich McCormick at 45%. The poll was released through Kozycki's campaign website. A two-point gap in a sample that size is a close race, not a lead for either candidate.

Our forecast did not move in any way that matters. Our model gives Kozycki a 1 in 100 chance of winning the seat, the same as in our previous run at 2:32 PM ET today. McCormick's chance stays at 99%. Our expected margin is still R+19.0.

The only change is at the edges. Our 5th-to-95th percentile range now runs from R+32.4 to R+5.6, against R+32.5 to R+5.5 before the poll was added. That is a shift of a tenth of a point at each end. Even the Democrat's end of that range is a Republican win by more than five points.

The poll sits well outside the bulk of our simulated outcomes for this district. One campaign-released survey of 457 voters is not enough for our model to discard the rest of what it knows about the seat, and it did not.

Nationally, our model gives Democrats a 99% chance of winning a House majority, with an average of 245.2 Democratic seats against the 218 needed. Our House simulation correlates polling error nationally, by census division, by state and by district, but not across demographically similar districts in different places, so that figure is somewhat firmer than it would be with that term included. Our generic-ballot average stands at D+8.2 across 35 polls.

Polls this dispatch was written from

  • 20/20 Insight · Oct 9, 2026 · n=457 · GA-7

Every constant behind these numbers is on the methodology page.