Tuesday, two weeks before the 2026 midterms, FiftyPlusOne, a non-partisan polling aggregation outlet, published a forensic investigation into Wedgewood Polls, a self-professed polling firm that had been conducting surveys since late June. In response to a request for comment, Wedgewood did not answer FiftyPlusOne’s questions. Instead, they posted a public statement in which they would admit their data was fabricated. Wedgewood sent FiftyPlusOne a direct message two minutes before publishing their official press statement, “Our official response to your article: Fuck you.”
FiftyPlusOne’s investigation was methodologically rigorous. The outlet paid $900 to conduct their own poll through Pollfish, the same panel provider Wedgewood had claimed to use. What they found was damning: respondent ID numbers, transaction IDs, and demographic distributions, among other things, did not match Pollfish’s known patterns in any significant way.
Wedgewood would release a public statement themselves, attempting to beat FiftyPlusOne to the punch. They are claiming the operation as a deliberate social experiment designed to expose vulnerabilites in the polling industry’s vetting processes.
“Despite having no history, formal credentials, or real data behind the vast majority of our releases,” the statement read, “Wedgewood Polling was added to major polling aggregators, including the New York Times, FiftyPlusOne, and VoteHub, and repeatedly treated as a legitimate data source.”
Many replies have pointed out that the statement was likely generated by artificial intelligence.
Join Gambit Forecaster’s Newsletter!
The social experiment framing does not hold up against any possible timeline. The statement was issued shortly after responding to FiftyPlusOne’s inquiries. Any rational person would view that as a direct response to being caught, and not as a planned conclusion to a defined research project. Further claims that they didn’t attempt to mislead participating campaigns or the public is cute, at best. This is blatant fraud. Through and through.
In a tweet, Zach Donnini, the Head of Data Science at VoteHub, described speaking on the phone for fifteen minutes with someone who identified himself as a college student at Boston University with prior experience at firm SoCal Strategies. Donnini said he extended good faith to what he thought were younger people attempting to break into the field. Adam Carlson of Zenith Research put it plainly in his own tweet: “They went to great lengths to lie to me privately.”
The long-term damage may fall on the next generation of legitimate newcomers to the polling industry. Small pollsters, aggregators, and media outlets will likely approach upstarts with a new level of suspicion (and rightfully so). FiftyPlusOne has announced more rigorous screening going forward. The New York Times has not released a statement, but has removed Wedgewood from their website.
Gambit Forecaster previously utilized Wedgewood Polls data in its models from late June through our October 1st simulation, when the firm was removed (alongside various others) as part of an update to Gambit’s own internal pollster standards. The removal preceded the FiftyPlusOne investigation, and was made for unrelated reasons. In those respective simulations, Wedgewood Polls was given the pollster rating “NR” for Not Rated. Gambit Forecaster did not independently reach out to Wedgewood Polls to confirm their methodology or legitimacy.
Check out our midterm election forecast, and our State Watch page.