Past hoax · Filed Oct 22, 2026
The Fake Polling Hoax
The numbers were a real poll, and the candidate on the wrong end of them had to get out.
The claim
The numbers were a real poll, and the candidate on the wrong end of them had to get out.
What it was
Wedgewood Polls admitted on October 20 that almost all of it was made up. Mallory McMorrow had already quit the Michigan Senate primary on July 5.
Timeline
Jun 11–13, 2026
Mitchell Research, using an open SurveyMonkey link, puts Mallory McMorrow at 6 percent in the Michigan Democratic Senate primary. This one is not Wedgewood. It is a bad method, and it is already public.
Jun 29, 2026
Wedgewood releases the survey she later calls infamous: Abdul El-Sayed 46, Haley Stevens 39, McMorrow 9. In the general-election matchups, El-Sayed leads Mike Rogers, Stevens ties him, and Rogers leads McMorrow.
Jul 5, 2026
McMorrow suspends. She says donations dried up, volunteers panicked, and reporters called her to drop out. Her internals, she says, did not have her in single digits.
Aug 5, 2026
El-Sayed wins the primary, 48.5 to Stevens’s 47.5. The Associated Press called a one-point race. The summer polls that had him up around ten did not.
Aug 2026
Median Strategies, a different firm, discloses that two of its polls were fake: the Wisconsin Democratic Senate primary and the Los Angeles mayor’s race. On this tape O’Connor also shows a Francesca Hong number from the Wisconsin governor primary and says that one was Median too. She lost to David Crowley. She did not quit.
Oct 14, 2026
Wedgewood’s Mississippi survey has Sen. Cindy Hyde-Smith at 52 and Democrat Scott Colom at 48. Trump carried the state by 24 points in 2024.
Oct 20, 2026
After FiftyPlusOne questions the raw file, Wedgewood admits the numbers were fabricated and calls itself an experiment. The same day, Senate Majority PAC confirms a first $1.2 million buy for Colom.
Oct 22, 2026
Larry O’Connor airs it on Townhall. The New York Times has pulled 15 Wedgewood polls. Decision Desk HQ has pulled 17. McMorrow has already been out of the race for three months.
What aired
Larry O’Connor on Townhall, October 22. The Facebook title is “Democrats fell for fake polls. Now the $1.2 million fallout is coming out.” He reads Wedgewood’s own post: the firm was never meant to sway an election, it was an experiment in how little anyone checks. Then he puts the Michigan numbers on the screen. In a late-June Wedgewood survey, Abdul El-Sayed beats Mike Rogers by five, Haley Stevens is tied with Rogers, and Rogers beats Mallory McMorrow by seven. He reads what McMorrow wrote after the admission. Donations dried up. Volunteers were devastated. Reporters called her to drop out. She did. He then shows Mississippi, a state he calls ruby red, where the same firm had the Republican incumbent up only four, and says Chuck Schumer’s Senate Majority PAC put $1.2 million into that race on October 20. His line is that the check was based on the fake poll.
The record
On October 20 Wedgewood said the vast majority of what it had published was fabricated. FiftyPlusOne had already found that the “raw” respondent IDs in a file Wedgewood handed over did not behave like real Pollfish data. The New York Times took down 15 Wedgewood polls released between June and October. Decision Desk HQ took down 17. The June survey that McMorrow calls the infamous one had El-Sayed at 46, Stevens at 39, and McMorrow at 9. She suspended on July 5. She wrote that the number was at odds with her own internals. In Mississippi the October 14 release was Cindy Hyde-Smith 52, Scott Colom 48. Senate Majority PAC then announced a $1.2 million television and digital buy for Colom, its first money in that race. The sequence is public. A confession from the PAC is not.
Left out
Left out of “she quit because of the one fake poll” is the other number already in the race. A mid-June Mitchell Research survey, built on an open SurveyMonkey link, had McMorrow at 6. She called the two a one-two punch. The Wedgewood result was the one later admitted to be invented. It was not the only bad survey she was staring at. Also left out of “the $1.2 million was the fake poll” is that Senate Majority PAC has not said that. O’Connor says it. The dates line up. That is an argument, not a receipt. On the same show he puts up a Francesca Hong poll from Wisconsin and says that one was a different firm, Median Strategies, which had already admitted faking numbers. She lost her primary to David Crowley. She did not drop out. The dropout on this tape is one person: Mallory McMorrow.
Why it stuck
A number with a logo looks like a measurement. Once an aggregator averages it in with surveys that were actually conducted, a made-up 46-39-9 becomes “the polls.” Reporters then call the person in single digits and ask when she is quitting. The admission arrived in October. The exit was July 5. By then the experiment had already spent a campaign, and a super PAC had already written a check in a race the same firm had drawn as a four-point contest. The correction does not unwind either one.