Our Kansas Senate odds for Hamilton stand at 61 percent, up from 51 three days ago
Our projected margin moved from D+0.1 to D+1.5, and the one Kansas poll in our data shows a 1-point race that we treat as close.
Written by the newsroom · anthropic/claude-opus-5.5 · October 14, 2026 at 2:32 AM 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.
Our model now gives Democrat Adam Hamilton a 61 in 100 chance of defeating Republican Sen. Roger Marshall in Kansas. That is 10.4 points higher than the 51% our model gave Hamilton in the run of October 11 at 12:32 AM ET, three days earlier. Marshall's chance in our model fell from 49% to 39% over the same span.
Our projected margin moved from D+0.1 to D+1.5. The spread of plausible outcomes also shifted toward Hamilton. Our model's 5th-to-95th percentile range now runs from R+7.1 to D+10.2, compared with R+8.6 to D+8.8 in the earlier run. A range that wide means either candidate can still win comfortably.
The only Kansas poll in our data for this window is an InsiderAdvantage survey of 1,200 likely voters, fielded October 9-10. It found Hamilton at 48% and Marshall at 47%. A 1-point gap is a close race, not a lead. The survey finished fielding before the earlier run. Our data does not show when it entered the model, so it does not fully account for the movement. The rest of the change remains unexplained.
The race feeds into the national picture. Our model gives Democrats a 61% chance of winning a Senate majority, with a mean of 51.0 Democratic seats. Democrats need 51 seats for control and Republicans need 50, because the vice president is a Republican and a 50-50 tie goes to the Republicans. The election is 20 days away.
Polls cited
- InsiderAdvantage · Oct 10, 2026 · n=1,200 · Kansas Senate
Every constant behind these numbers is on the methodology page.