A polling outfit just admitted it faked most of its “surveys,” and big-name trackers still posted the numbers.
Story Highlights
- Wedgewood Polls said most of its published data was fabricated and called it a “social experiment.”
- The group’s fake results appeared in major aggregators and media discussions before exposure.
- Data sleuths flagged raw files with anomalies consistent with fabrication.
- A leading polling site says it had vetted some details, disputing claims of zero oversight.
What Wedgewood Admitted And Why It Matters
Wedgewood Polls posted a statement on X saying most of its polls were fake and part of a project to show weak guardrails in modern polling. The post came after inquiries from a polling outlet that was investigating the firm’s data. The admission said Wedgewood shared fraudulent raw files and sought to prove how a new “pollster” could gain credibility fast. This was not guesswork; it was their own description.
Wedgewood claimed the “experiment” worked. Despite thin credentials, it said its numbers got listed by major aggregators and discussed by media, including during high-stakes Senate coverage. One regional outlet reported those fabricated figures were used to paint a close race. That spread matters to voters, donors, and lawmakers who watch polling averages like a scoreboard. This episode shows how shiny formats can slip past filters when speed beats scrutiny.
How Investigators Flagged The Fakes
A data analysis linked to Columbia Journalism Review reviewed Wedgewood’s raw files and found strange patterns. The report said respondent and transaction identifiers did not match the behavior seen in comparable data. The fields also clashed with how the survey platform describes those values. These red flags fit what you would expect from synthetic data, not live responses from real people. That technical trail supported what Wedgewood later confessed.
The timeline also matters. A polling analysis outlet said it asked Wedgewood for comment during an active probe. It then received raw data on October 15. Wedgewood did not answer directly, but soon published the admission post. That chain shows the confession followed outside scrutiny, not just a random reveal. It also underscores how independent checks can protect voters from bad numbers.
Where The Oversight Claims Collide
Wedgewood’s post framed the episode as proof that aggregators and media lack guardrails. One outlet pushed back. It said Wedgewood had people with industry experience and had provided a methodology statement. That account suggests the group did more than slap together a website. It acted like a normal shop long enough to earn a hearing. The dispute does not undo the admission, but it shows vetting exists and can still miss fakes.
Wedgewood also said one Nebraska survey was real and used as a baseline. Outside reporting repeated that claim. But there is no public audit, invoice, or respondent-level proof that confirms it. Without a full release of files and logs, readers cannot verify what was genuine and what was not. That gap limits how much weight anyone should place on the “control poll” story. Treat that detail as unproven.
Why This Hits Home For Conservative Voters
Activists, campaigns, and newsrooms treat polls like hard facts. Lawmakers watch them. Donors respond to them. If fake numbers can slip into respected averages, it tilts the field. It can shape news cycles, set narrative “momentum,” and even sway turnout plans. Conservatives have long warned about media gatekeeping and bandwagon effects. This case backs that concern and calls for tighter verification before polls get airtime and trust.
Fixes are simple and overdue. Aggregators should require vendor contracts, platform logs, respondent-level data with privacy safeguards, and clear weighting code for any new firm. Outlets should link methods up front and flag unknown pollsters until an independent audit clears them. Wedgewood’s full archive should be released for review. Until then, readers should weigh polls as one signal, not gospel. Trust, but verify—then verify again.
Sources:
talkingpointsmemo.com, x.com, blog.fiftyplusone.news, magnoliatribune.com, storm.mg
