Escape the polling paradox with Solid Purple’s approach to election modeling
Zach Heltzel's data processing says there's an important gap for uncertainty in reading the room during a rollercoaster election cycle.
Endless Twitter scrolling and daily poll updates traps people in the all too familiar stress of election news season, a feeling Zach Heltzel knows all too well. So for his sanity and the sanity of everyone wanting to stay informed, he created the political forecasting website, Solid Purple, with accuracy and peace of mind at the forefront.
“Anytime that there was a new poll that didn’t line up with conventional wisdom, I would either be ecstatic or freak out in a way that was a bit of an emotional roller coaster, even if the results were actively contradictory. So I felt that if I built something myself, I could just sort of emotionally detach from riding the roller coaster every day.”
While polls may come out with quicker data, Heltzel’s coding uses past statistics and current projections to weigh each poll’s actual impact on the races for the presidential, house, and senate races. This weighing system resolves contradicting polls and outputs a projection closer to reality.

Creator of Solid Purple, Zach Heltzel, was looking for a way to escape the bombardment of polls and have a more consistent, long-term forecasting space.
Since this is one person’s passion project, the website is updated once a week with new polling data. This constraint allows visitors to balance being informed and maintaining a life outside of politics. There is no way to sit, constantly refreshing the page.
With a deep concern for the outcome of political races, Heltzel worked on Democratic campaigns in college and often finds himself trapped in the Twitter doomscroll. As he built Solid Purple to escape this ensnarement, he became more comfortable with the unknown and the unceasing battle between red and blue.
“The more I learned, the less confident I felt about most of my underlying assumptions, but then I also felt a lot more confident in my uncertainty,” Heltzel says. “So I synthesized that idea with the likelihood of a race going one way or the other, and the name, Solid Purple, came to mind because I very solidly don’t know what’s going to happen.”
While the forecasting pages summarize carefully crafted predictions, Heltzel gives a greater explanation of the recent political event through the Solid Purple blog. His unfiltered thoughts provide a space for all the readers to connect with the insanity of our political system. It provides a way for the reader to reaffirm that they are sane and the political landscape is the crazy one.
Live versions of Solid Purple’s data sets, updated in real-time, are embedded below:
Presidential Race (Live Data Map)
Additional Probabilities
Biden wins an Electoral College (350+) landslide: 8%
Trump wins an Electoral College (350+) landslide: 20%
Biden wins the national popular vote by 10%+: 12%
Trump wins the national popular vote by 10%+: 21%
Biden wins at least one state he didn’t win in 2020: 44%
Trump wins at least one state he didn’t win in 2020: 96%
Biden loses the popular vote but wins the election: 1%
Trump loses the popular vote but wins the election: 12%
Both major candidates tie with 269 EVs each: <1%
Conditional Odds
Biden wins election while losing Arizona: 17%
Trump wins election while losing Arizona: 33%
Biden wins election while losing Georgia: 22%
Trump wins election while losing Georgia: 35%
Biden wins election while losing Michigan: 6%
Trump wins election while losing Michigan: 29%
Biden wins election while losing Nevada: 19%
Trump wins election while losing Nevada: 31%
Biden wins election while losing Pennsylvania: 8%
Trump wins election while losing Pennsylvania: 19%
Biden wins election while losing Wisconsin: 10%
Trump wins election while losing Wisconsin: 26%
Likeliest Tipping Point States
Pennsylvania – 16%
Texas – 11%
Michigan – 11%
Florida – 8%
North Carolina – 7%
Georgia – 7%
Wisconsin – 6%
Arizona – 5%
Virginia – 4%
Minnesota – 3%
Nevada – 3%