This model uses methodology similar to the 2025 Virginia Gubernatorial Forecast, but with a few changes listed below the data .

Methodology Updates:

• RECENT (3/10/2025): Due to 538 shutting down, adjusted poll scores will now be assigned based on the Silver Bulletin Pollster Ratings. A letter grade of "A" equals a value of 1 while a grade of "D" equals a value of 0 (with a slight curve in between that gets steeper as the grades drop off). Any poll with a failing grade is not included in the model.
• The Sample Score is now calculated based on the equation [0.435*(.435x-300)^(1/11)] to account for the diminishing returns when pollsters exponentially increase their sample size. A new equation existed for this with the launch of this model, but it has been updated to distinguish the varying gap in samples slightly more. The Sample Score now also takes the sample subjects into account, where a survey of general adults changes the equation to be [0.435*(.435x-900)^(1/11)]. This rewards polls that survey registered voters instead.
• This model isn't forecasting any outcome for an election and it's entirely based on public perception of the administration, therefore there aren't any fundamentals used in this model
• Pollsters are limited to a maximum of two polls in the model at any given time (with additional polls being removed based on how recent their median surveying date was). Additionally, pollsters can not have multiple polls with median surveying dates of the same day.