How the projections work
- Each 2026 district is built from the 2024 precincts inside it (Chicago, suburban Cook and 14 other county election authorities from county and state election board files; all other counties from Redistricting Data Hub precinct results).
- Expected 2026 vote per precinct is its 2024 presidential vote scaled to midterm turnout (74% of 2024). Once whole counties finish counting, the turnout factor is re-estimated from them.
- Swing is the change in Democratic share of the two-party vote vs. Harris 2024, weighted by votes in precincts that have reported. Counties that publish only county-wide totals are compared with their in-district 2024 baseline.
- Uncounted precincts are assumed to vote at their 2024 share plus the race's swing. Until a district has meaningful returns, its swing is pulled toward the statewide swing in the Governor race.
- Win odds use the projected share and an error band that narrows as votes are counted. Seat ranges (10th–90th percentile) come from 4,000 simulations that share a statewide error, so districts move together.
- Early Illinois counts are mostly mail ballots, which lean Democratic. The error band widens to account for this, but early swings can still overstate Democrats until Election Day precincts report.
- School board races are nonpartisan: odds reflect only the leader's margin and how much of the subdistrict has reported.
- Ratings: Projected ≥99.5% with at least a quarter counted; Likely ≥95%; Lean ≥75%; otherwise Tossup. These are statistical estimates, not race calls.