Prediction Methodology
Truth In Polling uses a transparent, multi-factor Predicted Floor Position (PFP) model to estimate how elected officials will vote on pending legislation.
Federal Factors
- Party voting patterns (40%)
- Campaign donor interests (35%)
- Historical voting record (15%)
- Cosponsorship status (10%)
State Factors
- Party voting patterns (45%)
- Party loyalty (30%)
- Historical votes (25%)
All data is sourced from Congress.gov, FEC, LegiScan, and direct state APIs. No AI or machine learning is used in the prediction model itself.