Press Release, November, 2026
On the Record: Baraza’s 2026 Election Forecasts
A pre-election snapshot of 506 race forecasts and 16 polling comparisons, prepared ahead of the November 3 election.
CHAMPAIGN, IL— Baraza is presenting its 2026 election forecasts ahead of the November 3 election, giving the public a view of projected outcomes across all 435 US House races, 35 US Senate races and 36 gubernatorial contests. The forecasts are available through Baraza Predictions, powered by Maya, at getbaraza.com/predictions.
The published forecast set, updated October 27, includes projected candidate margins, win probabilities and expected ballots, alongside polling averages and poll counts where available. Visitors can explore races on an interactive map or compare contests in a table ordered from closest to widest projected margin.
The national House projection is 231 Democratic seats and 204 Republican seats. For the Senate, rounded expected party totals are 50 Democrats, 48 Republicans and two independents. A separate Democratic-aligned caucus projection is 52 seats, including independents counted with Democrats in the model. The party estimates and caucus projection use different statistical summaries, which are explained on the forecast page.
A forecast that can be evaluated
Baraza’s forecasting approach is designed to test whether voter archetypes, behavioral segmentation and race-level data can add predictive value beyond polling averages. Its published outputs are numerical estimates that can be compared with election results, rather than narrative predictions generated by a general-purpose chatbot.
Polling remains an essential benchmark. Baraza displays those comparisons alongside its forecasts so readers can see where the model agrees with polling and where it differs. A disagreement does not, by itself, establish that either estimate is more accurate.
The company’s longer-term research examines whether information shared across cities, districts and states helps explain voting patterns that polling summaries alone leave unresolved. That remains a hypothesis to test, rather than a validated claim about the model’s performance.
Sixteen estimates to watch
The following selections illustrate the October 27 forecast set. All figures are two-candidate margins in percentage points: D+ indicates a Democratic advantage and R+ a Republican advantage. Polling benchmarks are those included in that forecast update; parentheses show the number of polls.
| Race | Candidates | Baraza forecast | Polling benchmark |
|---|---|---|---|
| Maine Senate | Troy Jackson vs. Susan Collins | D+2.0 | D+1.4 (11) |
| ME-2 | Matthew Dunlap vs. Paul LePage | D+0.1 | Even (8) |
| Texas Senate | James Talarico vs. Ken Paxton | D+1.2 | D+2.4 (26) |
| TX-15 | Bobby Pulido vs. Monica De La Cruz | R+2.0 | D+5.1 (6) |
| Iowa Senate | Josh Turek vs. Ashley Hinson | R+0.6 | D+0.1 (14) |
| Iowa Governor | Rob Sand vs. Zach Lahn | D+5.3 | D+8.1 (10) |
| Kansas Senate | Adam Hamilton vs. Roger Marshall | D+1.0 | D+1.7 (6) |
| Kansas Governor | Cindy Holscher vs. Ty Masterson | R+7.2 | R+4.3 (2) |
| Michigan Senate | Abdul El-Sayed vs. Mike Rogers | D+3.4 | D+2.7 (14) |
| MI-10 | Christina Hines vs. Mike Bouchard Jr. | D+1.3 | Even (1) |
| North Carolina Senate | Roy Cooper vs. Michael Whatley | D+10.7 | D+12.3 (7) |
| NC-1 | Don Davis vs. Laurie Buckhout | D+6.7 | D+10.9 (1) |
| Georgia Senate | Jon Ossoff vs. Mike Collins | D+11.1 | D+8.2 (4) |
| Georgia Governor | Keisha Lance Bottoms vs. Rick Jackson | D+0.1 | R+1.8 (5) |
| Ohio Senate | Sherrod Brown vs. Jon Husted | D+2.7 | D+3.6 (16) |
| Ohio Governor | Amy Acton vs. Vivek Ramaswamy | D+1.4 | D+3.8 (13) |
TX-15 presents a notable comparison: Baraza’s R+2.0 estimate differs by 7.1 percentage points from the D+5.1 polling benchmark. Georgia’s governor race and Iowa’s Senate race also differ in which party holds the estimated advantage. Very narrow margins, including D+0.1 in ME-2 and Georgia’s governor race, should be read in the context of forecast uncertainty.
What comes after Election Day
Baraza plans to compare its published margins with certified two-candidate results and the polling benchmarks recorded before the election. The evaluation will cover the full set of comparable races, not only the 16 highlighted here, and will report average absolute error and directional bias. Win-probability calibration will be assessed separately.
No claim of forecasting accuracy can be established before results are available. Publishing the estimates in advance creates a record against which the model’s performance can be measured.
Explore the forecasts at getbaraza.com/predictions.
About Baraza
Baraza is a campaign intelligence platform for candidates, committees and organizations. Its platform brings CRM, call time, fundraising, events and tasks into one connected workspace. Founded in 2026 by Chandra Chouhan, Pranav Volety, Rohit Singhal and Aryan Ranjan, the company is based in Champaign, Illinois.
Press contact
Pranav Volety Chief Design and Product Officer hello@getbaraza.com