Using data from The New York Times (NYT), as provided by Edison Research.
* In the enhanced Presidential ballot CSV files, the first 9 columns are the exact values as supplied by the NYT feed (as in the raw CSV), but the subsequent columns are some additional calculated values which are useful in seeing the changes (sometimes called deltas) between batches. Unfortunately the NYT percentages are rounded to the nearest 0.1%, and when extrapolating this out to hundreds of thousands of ballots, it can sometimes introduce anomalies in the calculations such as seeing more negative ballot numbers than is to be expected (yes, there are some strange negative values which may require investigation, but possibly not as many as you're seeing in the CSV). We're working on ways to normalize these, but in the meantime, be aware.
More charts and tables may be coming soon, as our volunteer developers are able. Stay tuned, and follow us on Telegram to get notified.
Further Data Sources
In most states, the certified election results and voter registration statistics are available from the Secretary of State or State Board of Elections. Check their website for details.
The New York Times 2020 Election Site contains results for all states, with several maps and charts
USElectionAtlas.org provides detailed results for 2020 and previous years. Some data is freely accessible on their website, while some, such as detailed historic results in CSV format, is purchasable for a fee.
Jeff O’Donnell has compiled additional data at MagaRaccoon.com that may be useful:
Election night time-series data from Edison – providing more detail than is available from the New York Times, and including numerous interactive charts.
Election Data Analyzer Tool, for Canvassing Voter Registrations
USEIP is currently developing a tool that takes voter registration data, combines it with maps, analyzes it for issues such as large numbers of people at a single address, and helps focus canvassing efforts on areas with the most irregularities.
Have an idea for a chart or graph that you would find useful here? Let us know in the comments below.