Wedgewood Polls Says Its Surveys Were Fabricated in Social Experiment
A new pollster that won media and database listings says the operation tested how easily a made-up firm could gain recognition
By AXY World Desk ·
Key points
- Wedgewood Polls said on October 20 that its surveys were fabricated and described the deception as a social experiment.
- FiftyPlusOne researchers confronted the firm with evidence that its underlying data was fake.
- Newsweek retracted three articles and corrected others; Decision Desk HQ removed 17 Wedgewood polls from its lists and averages.
- Warning signs included a website with no staff names, phone number or location, and an X account created in June.
Wedgewood Polls has said that the surveys it published were fabricated, and that the whole operation was a social experiment designed to expose weaknesses in the polling industry. The firm made the admission on October 20, after researchers at FiftyPlusOne, an election analysis website, confronted it with evidence that its underlying data was fake. Before that, prominent media outlets had reproduced its results, and several of those results had been placed in major polling databases.
In a statement posted on X, Wedgewood said the operation was meant to establish how easily a fictitious polling company could obtain recognition from polling aggregators and media outlets. It answered its own question directly. “The short answer: Far too easy,” the firm said.
Newsweek, which reproduced Wedgewood's surveys, has retracted three articles and corrected others. Other publications that reproduced Wedgewood surveys included The Hill, Breitbart, The Daily Beast and the Washington Examiner. Newsweek has contacted Wedgewood Polls for comment, and the firm's response had not been reported at the time of the statement.
How a Firm With No Named Staff Reached the Averages
A polling aggregator is a website that collects many polls and combines them into an average, which many readers use as a quick guide to where a race stands. Journalists, forecasters and databases lean on those averages because they are faster than reading every survey one by one. A new pollster that publishes regularly and looks legitimate can slip into them, and once it does, its numbers start to carry the same weight as those from long-established firms.
That is the gap Wedgewood appears to have exploited. Its results were included in prominent polling databases, including those maintained by The New York Times, Nate Silver, VoteHub and FiftyPlusOne, before FiftyPlusOne exposed the firm. Decision Desk HQ said it had removed 17 Wedgewood polls from its race-level polling lists and averages. Each removal pulls one number out of a calculation that readers may have already seen, which means every average built on those polls has to be rebuilt.
The New York Times had already listed Wedgewood's results when the firm was exposed, but with a warning attached. The disclaimer said: “This pollster does not meet our criteria for selected pollsters.” That line shows that at least one major gatekeeper had reservations, even though the numbers still reached readers through other channels.
Louisiana's Early Senate Poll Pointed the Wrong Way
Wedgewood emerged in June 2026, presenting itself as a new independent polling organization. One of its first surveys covered Louisiana's Republican Senate primary and showed John Fleming leading Julia Letlow by 4 percentage points. Letlow went on to defeat Fleming by almost 14 points on June 27.
A single poll that misses is not proof of fraud. Surveys miss results for ordinary reasons, including small samples, late shifts in voter opinion and imperfect methods. But this one came from a firm with no track record, and its result pointed in the opposite direction of the final vote. That made it an early reason to look harder at Wedgewood, not a conclusion in itself.
The poll also shows why a new pollster's first release matters. Early numbers are often the ones that draw the most attention, because they are the first data point anyone has. If those numbers are wrong, the damage can spread before anyone checks the methods behind them.
The Warning Signs Were Public by July
Questions about the firm's credibility surfaced within weeks of its appearance. On July 10, Washington Monthly columnist Bill Scher said that Wedgewood had only recently entered the polling industry and that its first survey had been badly wrong. The criticism was out in the open, yet Wedgewood's numbers continued to circulate in media coverage and in databases.
That gap between public doubt and continued use is part of the story. A critical column does not automatically remove a poll from an average or pull a headline from a website. Newsrooms and database managers have to decide for themselves whether a warning is serious enough to act on, and many of them rely on a pollster's own disclosures to make that call.
How Straight Arrow News Traced the Firm's Background
A more detailed look came on August 20, when Straight Arrow News examined Wedgewood's background. Its investigation found that the firm's website, hosted on Google Sites, listed no staff names, no telephone number and no location. Those are basic details that most established polling organizations publish, because they let reporters call, ask questions and check who stands behind the numbers.
The investigation also found that Wedgewood's X account had been created in June, and that the website was registered on June 26. A firm that appears in June and is already being quoted in national coverage has very little time to build a reputation, which is why its lack of history mattered.
Straight Arrow News also reported that Adam Carlson, a Democratic pollster at a different firm, said he had communicated with someone at Wedgewood. Carlson said that person told him the two-person firm included one progressive and one conservative. According to Straight Arrow News, Wedgewood declined to identify its operators publicly or take part in a video interview, saying one member had previously experienced online harassment.
Taken together, those details would have given a careful editor reason to pause. A pollster that publishes results but has no named staff, no listed phone number and no stated location does not meet the basic tests of transparency that readers expect from a survey. Requests for identity can be met with privacy concerns, but a firm that wants its numbers taken seriously has to accept some level of scrutiny.
Who Exposed the Firm and How
The investigation that finally exposed Wedgewood was carried out by FiftyPlusOne researchers Mary Radcliffe, G. Elliott Morris and Cooper Burton. Their work moved the question from suspicion to evidence. It was the confrontation with that evidence, according to Wedgewood's own account, that prompted the firm to say its underlying data was fake and its results fabricated.
Wedgewood's explanation is that the operation was designed to test the system from the outside. That claim is now on the record. It does not, however, change what happened to the readers, outlets and databases that relied on the numbers while they were in circulation. A test that succeeds at fooling careful institutions also puts false information in front of the public, even if the intent was to prove a point.
Why Fake Polls Reach Beyond One Firm
Polls shape how campaigns plan their spending, how donors decide where to give and how voters read a race. A fake survey can therefore affect people well beyond the newsroom that published it. A misleading average can change a headline, a forecast or a judgment about who is ahead, and those judgments can feed into decisions that have real consequences, such as where a campaign sends its money in the final weeks.
The Louisiana example shows the stakes in a single race. A survey that showed one candidate ahead when the other went on to win by almost 14 points could shape how voters, donors and reporters understood the contest, at least until the result was in. That is why the early signals from a new firm deserve more attention than they usually get.
What Newsrooms and Databases Now Have to Fix
This episode could push outlets to ask tougher questions before they publish a new pollster. The useful questions are simple. Who runs the firm? Who funds it? Where is the data collected, and can someone outside the firm check the methods? Those checks are not new, but this case suggests they are most often skipped for pollsters that look new and convenient.
The next test is whether newsrooms and databases apply those checks consistently, not only to well-known names. Decision Desk HQ's removal of 17 polls, and The New York Times' disclaimer, show that some gatekeepers can act quickly once evidence is in hand. The harder question is whether they can spot a problem before the numbers have already spread.
For readers, the lesson is practical. A poll from an unfamiliar firm deserves more scrutiny than one from a long-established organization, especially when it is the only survey showing a result. Retractions and removals can correct the record, but they cannot take back the time a false number spent being read as fact. Wedgewood's own statement made that point in three words: far too easy.
Questions and answers
What did Wedgewood Polls admit?
Wedgewood Polls said on October 20 that its results were fabricated. The firm described the deception as a social experiment intended to show how easily a fictitious polling company could gain recognition from polling aggregators and media outlets.
Who exposed Wedgewood Polls?
Researchers at FiftyPlusOne, an election analysis website, exposed the firm. Mary Radcliffe, G. Elliott Morris and Cooper Burton conducted the investigation that confronted Wedgewood with evidence that its underlying data was fake.
Which outlets were affected by the fake polls?
Newsweek retracted three articles and corrected others. Other publications that reproduced Wedgewood surveys included The Hill, Breitbart, The Daily Beast and the Washington Examiner, and the results also appeared in several polling databases.
Were Wedgewood's polls removed from averages?
Decision Desk HQ said it had removed 17 Wedgewood polls from its race-level polling lists and averages. The New York Times had already listed the results with a disclaimer saying the pollster did not meet its criteria.
Topics: Wedgewood Polls, FiftyPlusOne, Polling aggregators, Newsweek, Decision Desk HQ, Louisiana Senate primary, Straight Arrow News
Sources: Newsweek