Review the closest contests, open a matchup, then change incumbency or fundraising to see how the fitted model responds.
Scenarios are model-sensitivity illustrations. They are not official forecast updates or estimates of what would causally happen.
All races
Published margins determine the order. Select a matchup to inspect its full distribution or try a model-sensitivity scenario. An asterisk (*) marks an incumbent candidate.
| Race | Matchup | Likely winner | Win chance | Likely Margin |
|---|
Model sensitivity
Hypothetical result
Predictive uncertainty
Read this correctly
The model associates race outcomes with district partisanship, incumbency, and—where complete—candidate receipts measured at the forecast horizon. Changing fundraising here does not show the causal effect of raising that money; fundraising can itself reflect how competitive a race appears.
The published forecast stays fixed. A scenario applies only your edited inputs to the same paired posterior draws, so differences are not browser resampling noise.