Why I kept them separate
The combined model is only a comparison tool. The source files measure different things: individual cars, vehicle configurations, and annual fleet averages.
Data notes
I kept the main sources separate because they measure different things. This page lists the year range, unit of observation, and how I use each file.
Data coverage
The year ranges overlap, but the rows are not always comparable: some are individual vehicles, some are configurations, and some are annual fleet averages.
SourceYearsPrediction role
UCI Auto MPG1970–1982Historical vehicle predictor
EPA bridge sample1999 and 2008Vehicle-level bridge predictor
EPA Automotive Trends1975–2024Fleet and vehicle-type trend predictor
Combined-source table1970–2024Combined-source comparison
NRCan official ratings1995–2026Recent predictor after data download
How the site works
The pages are static. Predictions run in the browser; there is no server, account system, or tracking script.
Sources
UCI Auto MPG
EPA Automotive Trends
FuelEconomy.gov
NRCan ratings