Derivative Polling

Please rotate your phone
Tracking changes in the Biden-vs-Trump vote within pollsters

If you ever find yourself asking such questions, then this is for you. Rather than combining every poll under the sun, with complex corrections for pollster reliability, bias, state effects, economic fundamentals or whatnot, this tracker instead focusses on changes from one poll to the next by the same pollster. Whatever their particular biases, tracking changes within pollsters presents a simple picture of how vote intention may be changing. Updated daily.

Fine print: All polls are reduced to two-way, normalized to 100, and aggregated by pollster and week. These data include both 2-way and 5-way polls, of both registered and likely voters. Since this isn't an election prediction but a tool for tracking changes in vote intention, within-pollster changes should remain meaningful whether a pollster polls 2-way and/or 5-way, or registered and/or likely voters, as long as they don't change their approach too often. Data generously shared by FiveThirtyEight at github.com/fivethirtyeight.

Current data

Nick Beauchamp
Comments or suggestions: n.beauchamp at northeastern.edu