Moonlight Logo
Sign In About Team Contact
Hero background

Moonlight Metrics.

Win votes. Skip the spending arms race. Go on offense during policy debates. Do it all with a decision engine that answers questions your opponent won’t even think to ask.

The average candidate leaves 7% of votes on the table

Our candidates don't. By creating a direct map between change in policy platform and change in vote share, we help you find and tap into a gold mine of overlooked, undervalued issues.

Below, a comparison of prediction accuracies for voting outcomes given a policy bundle [inputs = single-issue polling, y-axis starts at 50% accuracy]

Linear
Regression

Polynomial
Regression

>90%
Accuracy

Moonlight
NeuralNet

Request a demo