The Catch.A permanent record of checked claims
SeriesMidterm pollsOctober 20, 2026

Midterm polls · October 20, 2026 · updated 2026-10-24

Pollster admits faking midterm polls that news outlets and forecasters cited

USA Today says it removed every reference to the polls, but as of Oct. 24 its Oct. 13 Georgia debate live blog still cites one.333432

  • 33USA Today, Sept. 18 (with note)USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 34USA Today, Oct. 5 (with note)USA TODAY has deleted the mention of a poll by Wedgewood Polls in this story.USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 32USA Today live blog, Oct. 13 (fetched Oct. 24)Ossoff leads Collins by 8 to 11 points in polls of likely Georgia voters by the Public Polling Project (Sept. 21-31), Insider Advantage (Sept. 22-23) and Wedgewood Polls (Oct. 2-3).

What happened

On the evening of Oct. 20, a firm calling itself Wedgewood Polls posted a statement on X saying that most of the polling data it had published since June was made up, "generated" in its word, with no real data behind the vast majority of its releases.13830 It called the whole operation an experiment to show how easily a fake polling firm could be taken seriously, and said one poll, in Nebraska, was real.12 It posted the statement hours after FiftyPlusOne, a site that averages polls, asked it to answer an investigation into whether its data was fake.53136

  • On Oct. 20, Wedgewood said most of the data it published was generated and that it had no real data behind the vast majority of its releases.1
  • Wedgewood Polls opened its X account in June 2026 and published polls from June until October.3830
  • Wedgewood called itself an experimental project to show the weaknesses and low barrier to entry of political polling, and said the answer to how easily a fake firm is taken seriously was "far too easy."1
  • Wedgewood said one crowdfunded Nebraska poll was real.12
  • Wedgewood posted the statement on X at 3:19 p.m. Pacific time on Oct. 20, after FiftyPlusOne had asked it that morning to respond to an investigation.53136
  • 1Wedgewood's statement, Oct. 20 (archived)Most data published under the Wedgewood Polling name was generated to test whether media outlets, aggregators, and commentators would report on data without verifying the source, methodology, or underlying firm legitimacy.real data behind the vast majority of our releasesan experimental project to demonstrate the vulnerabilities, lack of oversight, and overall low barrier to entry in the modern political polling industryThe short answer: Far too easy.we did conduct one fully legitimate, crowdfunded Nebraska poll during this run
  • 38PBS News, Oct. 21The organization's X account was opened in June 2026.
  • 30New York Times, Oct. 21The pollster had been releasing surveys since June
  • 2Wedgewood's X thread, archived Oct. 21Our Nebraska poll was real
  • 53UW Center for an Informed Public, Oct. 23On Tuesday, October 20, at 3:19 p.m. PDT, Wedgewood Polls
  • 13G. Elliott Morris, Oct. 20emailed them for comment this morning, ahead of publishing the findings tomorrow
  • 6FiftyPlusOne, Oct. 20They did not respond to our request, but instead posted a public statement on social media admitting that they had been fabricating data.

The New York Times listed 15 of its polls on its poll-tracking pages, and says it did so only after Wedgewood answered the paper's questions about who ran it, how it worked and who paid for it.30 The polls fed the averages of two election forecasters: Nate Silver's Silver Bulletin, which draws its polls from the Times, and Decision Desk HQ, which counted 17 Wedgewood polls.710 News outlets including USA Today, The Hill and Newsweek reported them.3334363744

  • The New York Times removed 15 Wedgewood polls from its tracking pages on the night of Oct. 20 and published an editors' note.30
  • The New York Times listed Wedgewood's polls only after it answered detailed questions about the pollsters' identities, methods and funding.30
  • Silver Bulletin takes its polls from the Times's database, which is how Wedgewood reached its averages.7
  • Decision Desk HQ removed 17 Wedgewood polls from its race lists and averages.10
  • USA Today's editor's notes on its Sept. 18 and Oct. 5 stories each say "USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned."3334
  • The Hill removed a Wedgewood Nebraska result, Ricketts ahead by 4, from its Oct. 4 story on Oct. 21 and added an editor's note.3637
  • Newsweek retracted three articles and corrected others.44
  • 30New York Times, Oct. 21The Times removed those polls Tuesday night and published an editors’ note that said it had included the Wedgewood polls15 in allafter the pollster answered a series of detailed questions from The Times about the pollsters’ identities, methods and funding
  • 7Silver Bulletin, Oct. 22we take our polling data from The New York Times
  • 10Decision Desk HQ post, Oct. 21we have removed 17 polls from our race-level polling lists and averages
  • 33USA Today, Sept. 18 (with note)USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 34USA Today, Oct. 5 (with note)USA TODAY has deleted the mention of a poll by Wedgewood Polls in this story.USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 36The Hill, Oct. 4 (as of Oct. 4)a Wedgewood Polls survey last month put Ricketts up by a slim 4 points, within the margin of error, with 52 percent to Osborn’s 48 percent
  • 37The Hill, Oct. 4 (now)Editor’s note: This story initially included polling from Wedgewood Polls. That information was removed on Oct. 21 after the firm admitted it fabricated data in its surveys.
  • 44Newsweek, Oct. 21Newsweek has retracted three articles and corrected others.

The only evidence that the Nebraska poll was real comes from Wedgewood itself: a screenshot of a survey service's completion notice, a crowdfunding page that raised $500 of a $5,000 goal, and its own release.354 As of Oct. 24, USA Today, which says it removed every reference, still cites a Wedgewood poll in an Oct. 13 live blog, and one poll site still counts its poll of Mississippi's Senate race, whose raw data FiftyPlusOne says was fraudulent, in its average.33343264266

  • The evidence on record that the Nebraska poll was fielded all comes from Wedgewood: its word, a Pollfish completion screenshot it posted naming a "Nebraska Wedgewood Poll," a GoFundMe that raised $500 of $5,000 from 14 donations, and its release describing an online panel of 500 likely voters fielded Sept. 14 to 16; none of it independently confirms the poll.354
  • USA Today's editor's notes on its Sept. 18 and Oct. 5 stories each say "USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned."3334
  • USA Today's Oct. 13 Georgia debate live blog, as fetched Oct. 24, still says "Ossoff leads Collins by 8 to 11 points in polls of likely Georgia voters by the Public Polling Project (Sept. 21-31), Insider Advantage (Sept. 22-23) and Wedgewood Polls (Oct. 2-3)," with no editor's note.32
  • On Oct. 24, PollingSource's Mississippi Senate page still marked Wedgewood's Oct. 12-14 poll as part of its three-poll average; the site says it does not exclude pollsters.64
  • FiftyPlusOne says that in the week before its Oct. 20 post, a credible tip from a trusted source that Wedgewood might be faking data led it to request the raw data for Wedgewood's Mississippi poll; it found the data surprisingly clean and, still lacking proof of fabrication, decided to replicate the survey "to the last detail"; it also says the respondent and transaction IDs in the data did not match real Pollfish patterns.26
  • Wedgewood shared raw survey data with FiftyPlusOne on Oct. 15 that FiftyPlusOne says it has confirmed was fraudulent.6
  • 3Pollfish screenshot posted by WedgewoodYour survey "Nebraska Wedgewood Poll" is complete.
  • 5Wedgewood's Nebraska GoFundMe$500 raised of $5K14 donations
  • 4Wedgewood's Nebraska release (DDHQ copy)FIELD DATES: SEPTEMBER 14–16, 2026 • N=500 LIKELY VOTERSMODE: ONLINE PANEL
  • 33USA Today, Sept. 18 (with note)USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 34USA Today, Oct. 5 (with note)USA TODAY has deleted the mention of a poll by Wedgewood Polls in this story.USA TODAY has removed any reference to Wedgewood Polls data in all stories where it was mentioned.
  • 32USA Today live blog, Oct. 13 (fetched Oct. 24)Ossoff leads Collins by 8 to 11 points in polls of likely Georgia voters by the Public Polling Project (Sept. 21-31), Insider Advantage (Sept. 22-23) and Wedgewood Polls (Oct. 2-3).
  • 64PollingSource Mississippi page (table)| Wedgewood Polls In average | 2026-10-14 this poll | 511 LV | ±4.30 | Scott Colom 48.00% | Cindy Hyde-Smith 52.00% |We do not exclude pollsters.
  • 26Andrew Gelman, Oct. 20Last week, we received a credible tip from a trusted source that Wedgewood Polls might be faking data.the respondent and transaction IDs recorded in survey data that Wedgewood provided to FiftyPlusOne do not match these expected patternsWhen we got the raw data from Wedgewood, we found that it was surprisingly clean.But we were still left with a question: Was there anything that actually proved the data was fabricated?That’s when we decided to try to replicate their survey, to the last detail.
  • 6FiftyPlusOne, Oct. 20sharing raw survey data (which we now have confirmed was fraudulent) with FiftyPlusOne on Oct. 15, 2026

Fox News headlined Wedgewood a pollster "boosting far-left candidates." Its Michigan and Wisconsin Democratic primary polls did show progressives ahead, Abdul El-Sayed and Francesca Hong, though in Michigan most real polls at the time also had El-Sayed leading.714174755273850 The Washington Examiner said its reports "consistently" leaned left, but of the 16 general-election polls Silver Bulletin could check against its averages, 10 leaned Democratic and 6 leaned Republican.428

  • Fox News headlined its story "Democrats furious after pollster boosting far-left candidates admits fabricating surveys," and wrote that the firm projected far-left candidates as strong contenders in several primaries.7141
  • Silver sees a pattern of better-than-consensus general-election numbers for notably progressive candidates in three races: Troy Jackson in Maine, Abdul El-Sayed in Michigan and Francesca Hong in Wisconsin, though he says it is "hard to know how much to make of three races."7
  • Wedgewood's June 25-28 Michigan Democratic primary poll had El-Sayed at 46, Stevens at 39 and McMorrow at 9, El-Sayed ahead by 7; while McMorrow was still in the race, El-Sayed also led in most real polls, mostly by single digits, and several later real polls had him ahead by double digits before he won by one point.47552738
  • Wedgewood's July 21 Wisconsin Democratic primary poll had Hong at 38%, Barnes at 30% and Crowley at 9%; Marquette Law School's last poll, about two weeks before the Aug. 12 primary, had Hong at 38% and Crowley at 7%.5550
  • The Washington Examiner wrote that Wedgewood's reports consistently reflected a bias toward left-leaning candidates.42
  • Of the 16 Wedgewood general-election polls in Silver's table that had an average to compare with, 10 were more Democratic than the average and 6 more Republican.8
  • 71Fox News headline, Oct. 21Democrats furious after pollster boosting far-left candidates admits fabricating surveys
  • 41Fox News, Oct. 21Democrats are up in arms after a so-called polling firm that projected far-left candidates as strong contenders in several primaries admitted that it had fabricated its polling data.
  • 7Silver Bulletin, Oct. 22So there is something of a pattern here of Wedgewood showing better numbers for notably progressive candidates than the consensus, although it’s hard to know how much to make of three races.they had Jackson ahead by 5 pointsshowing him ahead by 5 points, better than the average3 which had AES ahead by 0.4 points insteadan early July Wedgewood “poll” did have Francesca Hong leading by 3 points in the general election against Tom Tiffany
  • 47MIRS, Oct. 20It claimed to have surveyed 353 likely Democratic primary voters from June 25–28.
  • 55PoliAgg removal listJune 28, 2026Michigan · Senate · 2026Abdul El-Sayed 46 · Haley Stevens 39 · Mallory McMorrow 9· n=353July 21, 2026Wisconsin · Governor · 2026Francesca Hong 38 · Mandela Barnes 30 · David Crowley 9· n=505
  • 27New York Times, Aug. 4 (now)Even when Ms. McMorrow was still in the race, Dr. El-Sayed led in most polls, though his leads at that point were mostly in the single digits.
  • 38PBS News, Oct. 21Several polls showed El-Sayed with a comfortable double-digit lead over moderate Haley Stevens.However, he beat her by only one percentage point.
  • 50Urban Milwaukee, Aug. 19the vaunted Marquette Law School’s last poll on the race, released about two weeks before the Aug. 12 election, had Hong at 38% and Milwaukee County Executive David Crowley at just 7%
  • 42Washington Examiner, Oct. 23 (Maine)Wedgewood’s reports consistently reflected a bias toward left-leaning candidates
  • 8Silver Bulletin table, Oct. 22| Georgia | Gov. | Oct. 2-3 | -2 | -0.4 || Georgia | Sen. | Oct. 2-3 | +10 | +8.3 || Iowa | Gov. | Aug. 27-29 | +5 | +6.3 || Iowa | Sen. | Aug. 27-29 | -2 | -1.9 || Kansas | Gov. | Sept. 22-24 | -4 | -6.8 || Kansas | Sen. | Sept. 22-24 | +2 | -1.5 || Maine | Sen. | Sept. 27 - Oct. 1 | +4 | +2.7 || Michigan | Sen. | June 25-28 | +5 | +0.4 || Michigan | Gov. | Oct. 8-9 | +10 | +7.9 || Michigan | Sen. | Oct. 8-9 | +6 | +3.5 || Nebraska | Sen. | Sept. 14-16 | -4 | +2.3 || Nevada | Gov. | July 9-12 | -3 | -4.4 || Ohio | Gov. | Aug .11-13 | +1 | +0.5 || Ohio | Sen. | Aug .11-13 | +4 | +1.4 || Texas | Gov. | July 30-31 | -6 | -3.6 || Texas | Sen. | July 30-31 | +2 | +2.7 |

Who is behind it

Wedgewood's statement said it was run by "a small group of industry veterans."1 In August it told Straight Arrow News it was a two-person shop, one progressive and one conservative; its website listed no names, no phone number and no location.51 The firm's head pollster told Zachary Donnini of VoteHub, another poll tracker, that he was a Boston University student who had worked with a polling firm called SoCal Strategies, and that his partner was in local Democratic politics in Ohio.15 FiftyPlusOne says Wedgewood gave it two names, one of them Jack Sims, and that it confirmed Sims had worked in polling.6

  • Wedgewood's statement described itself as run by a small group of industry veterans.1
  • In August, Wedgewood told Straight Arrow News it was a two-person shop, described as one progressive and one conservative, and its website listed no names, phone number or location.51
  • Wedgewood's head pollster told VoteHub's Zachary Donnini he was a Boston University student who had worked with SoCal Strategies, and that the other member was in local Democratic politics in Ohio.15
  • Wedgewood gave FiftyPlusOne two names, Jack Sims and Andrew Prater, and FiftyPlusOne confirmed that Sims had worked in polling.6
  • 1Wedgewood's statement, Oct. 20 (archived)a fictional entity run by a small group of industry veterans can be integrated into major polling averages and cited by major news organizations
  • 51Straight Arrow News, Aug. 20In response to inquiries from Straight Arrow, they said they are a two-person shopthe two-person firm includes one progressive and one conservativeWedgewood Polls’ website has no employee names, phone number or even an indication of where it is located.
  • 15Zachary Donnini (VoteHub), Oct. 20He told me he was a college student at Boston University who had previously worked with SoCal Strategies and had paid for some inexpensive polls out of pocket.the other member of the team was involved in local Democratic politics in Ohio
  • 6FiftyPlusOne, Oct. 20We were given two names, Jack Sims and Andrew Prater.We were able to independently confirm that Sims had previous experience in the polling industry and reviewed their methodology statement in detail.

SoCal Strategies says one of the people FiftyPlusOne named, once associated with the firm, has not worked there for a year, and that his actions are his own.70 Wedgewood's website said it was based in Boston, but WGVU, a public radio station in Michigan, found no such business in Massachusetts records.48

  • SoCal Strategies says the man associated with it no longer works there and has not for a year, and that his actions are his own.70
  • Wedgewood's website said it was based in Boston, but WGVU found no such business in Massachusetts records.48
  • 70Axios, Oct. 21told Axios he no longer works there and had not worked there for a yearHis actions are his own and do not reflect on SoCal Strategies,
  • 48WGVU, Oct. 22The group's website lists Wedgewood Polls as being based in Boston. However, no such business exists in Massachusetts business records.

How it was caught

FiftyPlusOne says a tip from a trusted source led it to look at the raw data behind a Wedgewood poll of Mississippi. The data was surprisingly clean, and with nothing yet proving it fake, FiftyPlusOne set out to repeat the survey itself on Pollfish, the online survey service Wedgewood said it used, "to the last detail."26

  • FiftyPlusOne says that in the week before its Oct. 20 post, a credible tip from a trusted source that Wedgewood might be faking data led it to request the raw data for Wedgewood's Mississippi poll; it found the data surprisingly clean and, still lacking proof of fabrication, decided to replicate the survey "to the last detail"; it also says the respondent and transaction IDs in the data did not match real Pollfish patterns.26
  • 26Andrew Gelman, Oct. 20Last week, we received a credible tip from a trusted source that Wedgewood Polls might be faking data.the respondent and transaction IDs recorded in survey data that Wedgewood provided to FiftyPlusOne do not match these expected patternsWhen we got the raw data from Wedgewood, we found that it was surprisingly clean.But we were still left with a question: Was there anything that actually proved the data was fabricated?That’s when we decided to try to replicate their survey, to the last detail.

The repeat gave it real data to compare. In FiftyPlusOne's own results, each respondent's ID was one of two kinds, a plain string of numbers or one starting with a hyphen, in a rough even mix, and each transaction number rose almost in step with the time the person answered; Pollfish confirmed that is how its data behaves. Wedgewood's IDs behaved neither way, and FiftyPlusOne wrote that the red flags were "definitely still waving."2630 It gave Wedgewood until midnight to answer.6

  • FiftyPlusOne says that in its own Pollfish data, respondent IDs are a rough even mix of plain numbers and numbers prefixed by a hyphen, and transaction IDs track submission time almost one to one, which Pollfish confirmed is expected; in Wedgewood's Mississippi data the IDs did not behave that way, and its repeat of the survey made it clear the red flags were "definitely still waving."2630
  • FiftyPlusOne had given Wedgewood until midnight to answer before publishing its investigation.6
  • 26Andrew Gelman, Oct. 20the ID variable for a person appears as a roughly even mix between a standard string of numbersthe transaction identifier correlates almost 1:1 with the time a person submitted their surveyWe confirmed with Pollfish that this is the expected behavior of these columnsin Wedgewood’s data, the respondent and transaction IDs do not behave the same way they do in FiftyPlusOne’s data
  • 30New York Times, Oct. 21clear that the red flags we had seen in the Wedgewood data were definitely still waving
  • 6FiftyPlusOne, Oct. 20We had given Wedgewood until midnight to respond to our questions.
The last week, 2026
DateWhat happened
The week beforeFiftyPlusOne says a trusted source tipped it that Wedgewood may be faking data.26
Oct. 15Wedgewood sends FiftyPlusOne raw survey data, which FiftyPlusOne later confirms was fraudulent.6
Oct. 20, morningFiftyPlusOne asks Wedgewood to respond by midnight, ahead of publishing its findings.65313
Oct. 20, 6:19 p.m. ETWedgewood posts its statement on X.53136
  • 26Andrew Gelman, Oct. 20Last week, we received a credible tip from a trusted source that Wedgewood Polls might be faking data.the respondent and transaction IDs recorded in survey data that Wedgewood provided to FiftyPlusOne do not match these expected patternsWhen we got the raw data from Wedgewood, we found that it was surprisingly clean.But we were still left with a question: Was there anything that actually proved the data was fabricated?That’s when we decided to try to replicate their survey, to the last detail.
  • 6FiftyPlusOne, Oct. 20sharing raw survey data (which we now have confirmed was fraudulent) with FiftyPlusOne on Oct. 15, 2026We had given Wedgewood until midnight to respond to our questions.They did not respond to our request, but instead posted a public statement on social media admitting that they had been fabricating data.
  • 53UW Center for an Informed Public, Oct. 23On Tuesday, October 20, at 3:19 p.m. PDT, Wedgewood Polls
  • 13G. Elliott Morris, Oct. 20emailed them for comment this morning, ahead of publishing the findings tomorrow

Who corrected

Two news outlets still had a Wedgewood poll in front of readers as of Oct. 24: USA Today in its live blog, and The Hill through AOL's syndicated copy of its July Maine story.3272 MS NOW also aired a Wedgewood poll of Michigan's governor race, according to Mediaite, and Politico says its Morning Score newsletter mentioned Wedgewood twice.4039

  • USA Today's Oct. 13 Georgia debate live blog, as fetched Oct. 24, still says "Ossoff leads Collins by 8 to 11 points in polls of likely Georgia voters by the Public Polling Project (Sept. 21-31), Insider Advantage (Sept. 22-23) and Wedgewood Polls (Oct. 2-3)," with no editor's note.32
  • AOL's syndicated copy of The Hill's July 7 story on Wedgewood's Maine poll, as fetched Oct. 24, still reports the poll, with no editor's note.72
  • MS NOW featured a Wedgewood poll showing Jocelyn Benson 10 points ahead in Michigan's governor race, according to Mediaite.40
  • Politico says Wedgewood was mentioned twice in its Morning Score polling roundup.39
  • 32USA Today live blog, Oct. 13 (fetched Oct. 24)Ossoff leads Collins by 8 to 11 points in polls of likely Georgia voters by the Public Polling Project (Sept. 21-31), Insider Advantage (Sept. 22-23) and Wedgewood Polls (Oct. 2-3).
  • 72AOL copy of The Hill, fetched Oct. 24Wedgewood Polls found in its latest survey on the Maine Senate race that 75 percent of Maine voters would want Platner to drop out following another scandal
  • 40Mediaite, Oct. 21Earlier this month, MS NOW featured a Wedgewood Poll showingwith a 10-point lead over Trump-backed
  • 39Politico, Oct. 22and mentioned twice in POLITICO’s Morning Score polling roundup
News outlets
Outlet, and what it ranWhat it did after
The New York Times
15 polls on its tracking pages, and references in poll stories
Removed the polls the night of Oct. 20; editors' notes on its Aug. 4 Michigan and Aug. 11 Wisconsin stories302729
USA Today
Wedgewood numbers in several stories
Notes on its Sept. 18 and Oct. 5 stories; its Oct. 13 debate live blog still cites a Georgia poll, as fetched Oct. 24333432
The Hill
A Nebraska poll in an Oct. 4 story; a July story on the Maine poll
Removed the Nebraska poll Oct. 21, with a note; AOL's syndicated copy of the July Maine story still carries the poll, as fetched Oct. 24363772
Newsweek
Several stories
Says it retracted three and corrected others44
Washington Examiner
Three mentions in July
Says it removed them43
Mediaite
The July Maine poll
Removed it, with a note58
The Daily Beast
Two stories
Updated both68
Breitbart
An August Texas poll story
Added a note69
Poll trackers and forecasters
SiteWhat it did
Decision Desk HQRemoved 17 polls from its lists and averages; its data site hosts the Nebraska release104
Silver BulletinRemoved the polls; banned Wedgewood permanently9
State NavigateRemoved the polls from its governor forecasts in nine states11
ScopesRemoved the polls; withheld its Kansas Senate figure, which no longer had enough pollsters66
RealClearPoliticsSays it never listed Wedgewood17
PollingSourceLists 23 Wedgewood polls as active; its Mississippi Senate average still includes Wedgewood's Oct. 12-14 poll, as read Oct. 246064
Prediction EdgeIts Georgia Senate page lists Wedgewood's Oct. 2-3 poll among its latest polls, as read Oct. 2459
  • 10Decision Desk HQ post, Oct. 21we have removed 17 polls from our race-level polling lists and averages
  • 4Wedgewood's Nebraska release (DDHQ copy)FUNDED BY AN INDEPENDENT ELECTION TWITTER CROWDFUND
  • 9Silver Bulletin forecast note, Oct. 23we have permanently banned Wedgewood Polls from Silver Bulletin averages and forecasts
  • 11State Navigate post, Oct. 20GA, IA, KS, MI, NE, NV, OH, TX, WI are the states where there were Wedgewood polls in our forecasts.
  • 66Scopes oversight note, Oct. 22the fair value is the Scopes Poll Average: the rules-based median of qualifying public pollsthe test compares the market against the Scopes Fair ValueEvery Wedgewood input was removed from the Scopes Poll Average and the affected races recomputed.falls below the gate once Wedgewood is removed, so its poll-based fair value is withheld
  • 17RealClearPolitics, Oct. 22RealClearPolitics, the pioneer poll aggregator, never included these bogus pollsThey were a new pollster with no track record to evaluate. They also had a ‘placeholder-style’ Google Sites website, had no staff directory on that website, and no social media history
  • 60PollingSource Wedgewood page23 active 2026 polls are stored for Wedgewood Polls.
  • 64PollingSource Mississippi page (table)| Wedgewood Polls In average | 2026-10-14 this poll | 511 LV | ±4.30 | Scott Colom 48.00% | Cindy Hyde-Smith 52.00% |We do not exclude pollsters.
  • 59Prediction Edge Georgia page| Wedgewood Polls Oct 2 - 3 | Ossoff+10.0 | 55.0% | 45.0% |

Why would anyone do this?

Wedgewood says it set out to show how little checking polls get. It says the poll trackers of the Times, FiftyPlusOne and VoteHub all added its polls, and that MS NOW cited its numbers on air.1 It says it never meant to affect a race.1

  • Wedgewood said it was added to major aggregators, including The New York Times, FiftyPlusOne and VoteHub, and that its numbers were cited on air by outlets such as MS NOW.1
  • Wedgewood said it was never meant to sway an election and was designed as a wake-up call.1
  • 1Wedgewood's statement, Oct. 20 (archived)Wedgewood Polling was added to major polling aggregators, including The New York Times, FiftyPlusOne, and VoteHubcited on air and in reporting by outlets such as MS NOWWedgewood Polling was never intended to sway an election or permanently mislead the public; it was designed as a wake-up call.
Wedgewood Polls·Oct. 20, 2026·Statement posted on X

"Wedgewood Polling was never intended to sway an election or permanently mislead the public; it was designed as a wake-up call."1

Archived copy

What the checkers say

Wedgewood's statement says the trackers did not properly vet it; FiftyPlusOne calls that false.6 The Times says it listed the polls only after its questions were answered, and its listing carried a note that Wedgewood "does not meet our criteria for selected pollsters." VoteHub gave it a D grade.304415 FiftyPlusOne says the raw data Wedgewood later sent it was fraudulent.6 RealClearPolitics, another poll tracker, says it never listed Wedgewood, citing its lack of a track record, a staff list or any social media history.17

  • FiftyPlusOne says Wedgewood's claim that aggregators did not vet it is false.6
  • The New York Times listed Wedgewood's polls only after it answered detailed questions about the pollsters' identities, methods and funding.30
  • The Times listed Wedgewood with a note that it did not meet the paper's criteria for selected pollsters.44
  • VoteHub gave Wedgewood a D grade and included its polls.15
  • Wedgewood shared raw survey data with FiftyPlusOne on Oct. 15 that FiftyPlusOne says it has confirmed was fraudulent.6
  • RealClearPolitics says it never listed Wedgewood's polls, because the firm had no track record, a placeholder website, no staff directory and no social media history.17
  • 6FiftyPlusOne, Oct. 20Their statement alleges that FiftyPlusOne, the New York Times, and other poll aggregators did not conduct appropriate vetting of their firm prior to aggregating their polls; this is false.sharing raw survey data (which we now have confirmed was fraudulent) with FiftyPlusOne on Oct. 15, 2026
  • 30New York Times, Oct. 21after the pollster answered a series of detailed questions from The Times about the pollsters’ identities, methods and funding
  • 44Newsweek, Oct. 21This pollster does not meet our criteria for selected pollsters.
  • 15Zachary Donnini (VoteHub), Oct. 20we assigned them a D grade and included their polling data on our site
  • 17RealClearPolitics, Oct. 22RealClearPolitics, the pioneer poll aggregator, never included these bogus pollsThey were a new pollster with no track record to evaluate. They also had a ‘placeholder-style’ Google Sites website, had no staff directory on that website, and no social media history

G. Elliott Morris of FiftyPlusOne says Wedgewood is "trying to cover up their fraud."13 Silver calls the explanation "a pathetic attempt to get ahead of the story," and Andrew Gelman, a Columbia statistician, finds it "kind of suspicious" that it came "on the very day that they got caught."726 In August another polling firm, Median Strategies, admitted after questions from the Los Angeles Times that its Los Angeles mayoral poll was fake, and called itself a "short term social experiment."62

  • FiftyPlusOne's G. Elliott Morris says Wedgewood is trying to cover up fraud with the social-experiment excuse.13
  • Nate Silver calls Wedgewood's explanation an attempt to get ahead of FiftyPlusOne's story, and says he does not know its motives.7
  • Columbia statistician Andrew Gelman called it suspicious that the statement came out the day Wedgewood was caught.26
  • In August, a firm called Median Strategies admitted its Los Angeles mayoral poll was fake after repeated questions from the Los Angeles Times, and called it a short-term social experiment about polls spreading without verification.62
  • 13G. Elliott Morris, Oct. 20They are trying to cover up their fraud with the "we were doing a social experiment" excuse.
  • 7Silver Bulletin, Oct. 22My opinion is that Wedgewood’s explanation is bullshit, a pathetic attempt to get ahead of the story after FiftyPlusOne was set to publish a story exposing the fraud.I don’t know what Wedgewood’s motivations were.
  • 26Andrew Gelman, Oct. 20Maybe so, but it’s kind of suspicious that they happened to release this statement on the very day that they got caught.
  • 62Los Angeles Times, Aug. 17But Monday, in response to repeated inquiries from The Times, the company said the purported “poll” was nothing of the kind.Median Strategies was created as a short term social experiment to examine how purported polling information could enter and spread through the political information ecosystem without independent verification

Motives others suggest

Silver says he does not know Wedgewood's motives. He names two possibilities: moving betting markets, where people trade on who will win, with polls in races few others poll, and discrediting polling itself. He discloses that he advises Polymarket, a betting market.7 Morris now thinks the polls were more likely meant to move betting markets than to boost progressives.14 Kalshi, a betting exchange, says a trading account people linked to Wedgewood made $194.44 on markets tied to its polls.12 In August, Urban Milwaukee read the firm the other way, as "a Republican front group that passes itself off as liberal," pointing to its "Progressive but unbiased" profile and to a conservative institute's promotion of its Wisconsin poll.50

  • Nate Silver calls Wedgewood's explanation an attempt to get ahead of FiftyPlusOne's story, and says he does not know its motives.7
  • Silver names two possible motives: moving prediction markets with out-of-consensus polls in thinly polled races, and discrediting polling overall; he advises Polymarket.7
  • On Oct. 23 Morris wrote that it seemed increasingly likely Wedgewood's polls were meant to move betting markets, not to boost progressives.14
  • Kalshi says a trading account people linked to Wedgewood netted $194.44 across markets tied to its polls, and that its review is ongoing.12
  • In August, Urban Milwaukee wrote that Wedgewood appeared to be a Republican front posing as liberal, noting its "Progressive but unbiased" X bio and that a conservative institute promoted its Hong poll.50
  • 7Silver Bulletin, Oct. 22My opinion is that Wedgewood’s explanation is bullshit, a pathetic attempt to get ahead of the story after FiftyPlusOne was set to publish a story exposing the fraud.I don’t know what Wedgewood’s motivations were.publishing out-of-consensus “polls” or fake polls in sparsely-surveyed races have more potential to move marketsAnother motive might be to discredit the polling industry overall.I’m an advisor to Polymarket.
  • 14G. Elliott Morris, Oct. 23simply part of a coordinated shadow op to move betting markets, rather than people trying to boost progressives
  • 12Kalshi, Oct. 22The user’s net profits across all Wedgewood-related markets was just $194.44Importantly, this investigation remains ongoing.
  • 50Urban Milwaukee, Aug. 19As the story noted, Wedgewood appeared to be a Republican front group that passes itself off as liberal.its Twitter account described itself as “Progressive but unbiased,”Its poll on Hong was promoted by the conservative MacIver Institute

Did it change anything?

The forecasts barely moved, though one forecaster says a race with little other polling shifted more. Removing the polls changed neither Decision Desk HQ's projected seat totals nor its odds of control in the Senate or House, Silver says his forecasts were basically unchanged, though many other new polls came in at the same time, and the governor forecasts of State Navigate, another forecaster, moved by 1.1 points at most.7101173

  • Removing Wedgewood's polls changed neither Decision Desk HQ's projected Senate and House seat totals nor its chamber control probabilities, left Silver Bulletin's forecasts basically unchanged, Silver says (a comparison he says is confounded by other new polls), and moved State Navigate's governor margins by 1.1 points at most.71011
  • Scrutinel, a forecaster, says removing Wedgewood's eleven polls moved ten of the races in its averages by less than a quarter of a point each, but in Mississippi's Senate race, which had two polls in total, its average moved from about seven points to about twelve and the trailing candidate's chance of winning that race fell from around 13% to around 4%; across its whole Senate forecast the change was nine-tenths of a point.73
  • 7Silver Bulletin, Oct. 22That’s a bit confounded because a lot of polls are coming out these daysWhen we removed them from our database yesterday, FLIPR’s various averages and forecasts were basically unchanged.
  • 10Decision Desk HQ post, Oct. 21The removal of these polls has not changed the projected seat totals or chamber control probabilities in the Senate or House.
  • 11State Navigate post, Oct. 20NE w/the largest margin change at 1.1%. Everything else 0.1%-0.3%.
  • 73Scrutinel, Oct. 22In ten of the eleven races, almost nothing. Each moved by less than a quarter of a point.Mississippi was different. Its Senate race has two polls in total, so one of them was half of everything we knew.Once it was gone, our average moved from about seven points to about twelve, and the trailing candidate's chance fell from around 13% to around 4%.Across the whole Senate forecast the change was nine-tenths of a point.

Scrutinel says removing the 11 Wedgewood polls in its averages moved 10 of those races by less than a quarter of a point each, and its forecast for the whole Senate by nine-tenths of a point. Mississippi's Senate race was the exception: it had only two polls, one of them Wedgewood's. Without it, the lead in Scrutinel's average for that race grew from about 7 points to about 12, and Scrutinel's odds for the trailing candidate in that race fell from about 13% to about 4%.73

  • Scrutinel, a forecaster, says removing Wedgewood's eleven polls moved ten of the races in its averages by less than a quarter of a point each, but in Mississippi's Senate race, which had two polls in total, its average moved from about seven points to about twelve and the trailing candidate's chance of winning that race fell from around 13% to around 4%; across its whole Senate forecast the change was nine-tenths of a point.73
  • 73Scrutinel, Oct. 22In ten of the eleven races, almost nothing. Each moved by less than a quarter of a point.Mississippi was different. Its Senate race has two polls in total, so one of them was half of everything we knew.Once it was gone, our average moved from about seven points to about twelve, and the trailing candidate's chance fell from around 13% to around 4%.Across the whole Senate forecast the change was nine-tenths of a point.

Scopes, which publishes poll-based figures to compare against prediction-market prices, stopped publishing its Kansas Senate figure, which no longer had enough pollsters behind it.66

  • Scopes, which publishes poll-based fair values that it compares against prediction-market prices, removed Wedgewood's polls and withheld its Kansas Senate figure, which no longer had enough pollsters.66
  • 66Scopes oversight note, Oct. 22the fair value is the Scopes Poll Average: the rules-based median of qualifying public pollsthe test compares the market against the Scopes Fair ValueEvery Wedgewood input was removed from the Scopes Poll Average and the affected races recomputed.falls below the gate once Wedgewood is removed, so its poll-based fair value is withheld

On Kalshi the one clear move came in Wisconsin. Kalshi says its markets mostly did not move on Wedgewood's polls, except after the July 21 poll of the Democratic primary for governor.12 That poll came three days after David Crowley, the Milwaukee County executive, rejoined the race, and before any other poll had measured it; it put him at 9%.315550 BETINT, a newsletter on betting markets, says his Kalshi price fell from 23% to 13% within 15 minutes and stayed below its earlier level until primary day; Kalshi's own analysis puts the move at 6% and calls it temporary. Real polls the next week also had Crowley behind, and he won the primary.31125029

  • Kalshi says its election markets mostly did not move on Wedgewood's polls, except after the July 21 Wisconsin primary poll.12
  • Crowley had re-entered the Wisconsin primary three days before Wedgewood's July 21 poll, no other poll had been released since, and real polls the next week also had him trailing.3131
  • Wedgewood's July 21 Wisconsin Democratic primary poll had Hong at 38%, Barnes at 30% and Crowley at 9%; Marquette Law School's last poll, about two weeks before the Aug. 12 primary, had Hong at 38% and Crowley at 7%.5550
  • After Wedgewood's July 21 poll put David Crowley at 9%, his Kalshi price fell from 23% to 13% within 15 minutes and stayed below its pre-poll level until primary day; he won the primary.31
  • After the July 21 Wisconsin poll, Crowley's market traded at about 70 times its normal hourly pace and drew about $45,000 in new money within a day, by BETINT's count; Kalshi says the poll temporarily moved Crowley's price 6%.3112
  • Crowley won Wisconsin's Democratic governor primary with nearly 40% of the vote, a few thousand votes ahead of Hong, after real polls had shown Hong far ahead.5029
  • 12Kalshi, Oct. 22Kalshi has found that election markets mostly did not move in response to Wedgewood's polls (Figure A).It appears that Wedgewood polls only impacted market prices after their Wisconsin Democratic Governor primary poll (July 21, 2026) showing Francesca Hong with 38%, Mandela Barnes with 30%, and David Crowley at 9% temporarily moved Barnes' price up 4.5% and Crowley's price 6% (Figure B).temporarily moved Barnes' price up 4.5% and Crowley's price 6% (Figure B)
  • 31BETINT, Oct. 22three days after David Crowley re-entered Wisconsin’s Democratic primary for governor, Wedgewood published a survey putting him at 9%. No other poll had been released since his return.To be fair to the market, real polls released over the next week also had Crowley trailing, and they all missed too.Wedgewood’s polls moved real money once, clearly.His Kalshi price fell from 23% to 13 within 15 minutes of the poll’s releaseCrowley stayed below his pre-poll price until primary day. Then he won, by about half a point.the market traded at roughly 70 times its normal hourly pace, and about $45,000 in new money came into the contract within a day
  • 55PoliAgg removal listJuly 21, 2026Wisconsin · Governor · 2026Francesca Hong 38 · Mandela Barnes 30 · David Crowley 9· n=505
  • 50Urban Milwaukee, Aug. 19the vaunted Marquette Law School’s last poll on the race, released about two weeks before the Aug. 12 election, had Hong at 38% and Milwaukee County Executive David Crowley at just 7%Crowley eventually won with nearly 40% of the vote, with Hong just a few thousand votes behind.
  • 29New York Times, Aug. 11 (now)The only nonpartisan survey since the field settled showed Ms. Hong in the lead by more than 20 percentage points

Mallory McMorrow, a Democratic state senator in Michigan, was running for the party's Senate nomination when Wedgewood's late-June poll put her at 9%. She says the poll dried up her donations, devastated her volunteers and brought calls from reporters to drop out.4719 She left the race on July 5.3846 She also names a Mitchell Research poll that showed her in single digits just before, and says outside spending drowned her out.20

  • Wedgewood's Michigan poll, which it claimed surveyed 353 likely primary voters from June 25 to 28, had El-Sayed at 46%, Stevens at 39% and McMorrow at 9%.47
  • Mallory McMorrow says the late-June Wedgewood poll that showed her in single digits, at odds with her internal polling, dried up donations, devastated volunteers and brought calls from reporters to drop out.19
  • McMorrow, a Democratic state senator in Michigan, left the Senate race on July 5.3846
  • McMorrow says a Mitchell poll showing her in single digits came out just before, calls the two "a one-two punch," says "Then all of the sudden it was a pattern - and people believed it had to be true" and "It killed us," and says she was also drowned out by outside spending.20
  • 47MIRS, Oct. 20Its June 29 Michigan release placed McMorrow at 9% in the Democratic primary, behind Abdul El-Sayed at 46% and U.S. Rep. Haley Stevens at 39%, with 6% undecided.It claimed to have surveyed 353 likely Democratic primary voters from June 25–28.
  • 19Mallory McMorrow, Oct. 20the infamous one in late June that showed me plummeting into single digits (which was at odds with our internals)- Donations dried up- Volunteers were devastated and pleading for guidance on what to do- Reporters hounded me with calls to drop out
  • 38PBS News, Oct. 21Mallory McMorrow, a Democratic state senator in Michigan, suspended her campaign to become a U.S. senator in July
  • 46Michigan Advance, Oct. 21McMorrow, who eventually bowed out of the primary race July 5
  • 20Mallory McMorrow reply, Oct. 20Just before this poll hit, the Mitchell poll that also showed me in single digits hit.It was a one-two punch.Then all of the sudden it was a pattern - and people believed it had to be true.It killed us.And I was drowned out by outside spending and the narrative that we had completely collapsed.
Mallory McMorrow·Oct. 20, 2026·Post on X

"It was a one-two punch. Then all of the sudden it was a pattern - and people believed it had to be true."20

  • 20Mallory McMorrow reply, Oct. 20Just before this poll hit, the Mitchell poll that also showed me in single digits hit.It was a one-two punch.Then all of the sudden it was a pattern - and people believed it had to be true.It killed us.And I was drowned out by outside spending and the narrative that we had completely collapsed.

"It killed us."20

  • 20Mallory McMorrow reply, Oct. 20Just before this poll hit, the Mitchell poll that also showed me in single digits hit.It was a one-two punch.Then all of the sudden it was a pattern - and people believed it had to be true.It killed us.And I was drowned out by outside spending and the narrative that we had completely collapsed.

@MalloryMcMorrow

Real polls had her low too, Susquehanna at 9% and Quantus Insights at 8%.4730 But Susquehanna found 49% of voters still undecided, while Wedgewood's poll left only 6% undecided, and its poll made her the only Democrat losing to the Republican, Mike Rogers.6347

  • Real polls around then also had McMorrow low: Susquehanna at 9% and Quantus Insights at 8% on July 1.4730
  • Wedgewood's poll had McMorrow at 9% with only 6% undecided; Susquehanna's real poll also had her third but with 49% undecided.63
  • Wedgewood's June poll also made McMorrow the only Democrat trailing Republican Mike Rogers, 48% to 41%.47
  • 47MIRS, Oct. 20After the Mitchell poll, Susquehanna Polling and Research had McMorrow at 9%. Quantus Insights on July 1 had McMorrow at 8%.The release also portrayed McMorrow as the weakest Democrat against Republican Mike Rogers.Its purported general-election results showed Rogers leading McMorrow 48% to 41%
  • 30New York Times, Oct. 21(Other polls at the time showed her support declining, too.)
  • 63Michigan Public Radio, Oct. 22But the Susquehanna survey placed the percentage of undecided voters at 49%. Wedgewood’s made-up numbers said only 6% of voters were still making up their mind.At the time, reputable firm Susquehanna Polling and Research Inc. also listed McMorrow in third place

In Mississippi, Democrat Scott Colom's campaign promoted Wedgewood's poll as showing a tied Senate race.4931 Senate Majority PAC, a Democratic group, then put another $1.2 million behind Colom, and says it "absolutely" did not rely on the poll.67

  • Scott Colom's Senate campaign in Mississippi promoted Wedgewood's poll as showing a tied race, and Francesca Hong's Wisconsin campaign cited Wedgewood in a press release.4931
  • Senate Majority PAC put another $1.2 million behind Colom after the Mississippi poll, and its spokesperson says it "absolutely" did not rely on that poll.67
  • 49Magnolia Tribune, Oct. 20NEW INDEPENDENT POLL: Mississippi Senate Race Tied As Race Shifts In Scott Colom’s Favor
  • 31BETINT, Oct. 22The next day its press release cited Wedgewood by name.
  • 67New York Post via AOL, Oct. 21The Democrats’ Senate Majority PAC announced this week that it is pumping another $1.2 million behind Colom.But spokesperson Lauren French told The Post that the super PAC “absolutely” did not rely on that poll for its decision.

What happens next

Brendan Carr, the chairman of the Federal Communications Commission, warned broadcasters on Oct. 21 against airing fake polls, without naming Wedgewood.22 Rep. Stephanie Bice has asked the FCC and the Justice Department to investigate whether any broadcaster knowingly used fabricated polls to boost candidates, the Washington Examiner reports.43 Commissioner Anna Gomez says the rules do not apply to a station that reports poll data in good faith, and that the FCC can act only with proof that a station intentionally distorted the news.2461

  • On Oct. 21, FCC Chairman Brendan Carr, without naming Wedgewood, said there had been a surge in fake and suppression polls and warned broadcasters against airing polls that would violate their obligations.22
  • Rep. Stephanie Bice asked the FCC, with the Justice Department, to investigate whether any broadcaster knowingly used fabricated polls to boost candidates.43
  • FCC Commissioner Anna Gomez said the news-distortion and hoax rules do not apply to stations that report polling data in good faith.24
  • Gomez wrote on Oct. 23 that the FCC can enforce its news-distortion policy only with proof a station intentionally fabricated or distorted news, and that without such evidence it has no authority to act.61
  • 22FCC Chairman Brendan Carr, Oct. 21there has been a significant surge in entirely Fake Polls and Suppression Polls designed to do nothing more than interfere with the regular functioning of the electoral processBroadcasters are advised to avoid engaging in News Distortions or running Broadcast Hoaxes by airing Polls that would violate their regulatory obligations.
  • 43Washington Examiner, Oct. 23 (Bice)investigate whether any broadcast news company knowingly used fabricated polls to boost candidates and influence the outcome of the 2026 elections
  • 24FCC Commissioner Anna Gomez, Oct. 21FCC rules on news distortion or hoaxes do not apply to a station that reports this data in good faith.
  • 61Radio World, Oct. 23enforcing the news distortion policy requires proof that a licensee intentionally fabricated or distorted news coverage of a major eventAbsent that level of clear evidence, the commission has no authority to act,

Kalshi says its review of trading around Wedgewood's polls continues.12

  • Kalshi says a trading account people linked to Wedgewood netted $194.44 across markets tied to its polls, and that its review is ongoing.12
  • 12Kalshi, Oct. 22The user’s net profits across all Wedgewood-related markets was just $194.44Importantly, this investigation remains ongoing.

Sources

The records

Everything this page rests on, numbered where the story cites it. We keep a dated copy of each one.

Primary documents

  1. 1Wedgewood's statement, Oct. 20 (archived) · posted on X; archive.today copy of Oct. 21 · what Wedgewood admitted, its stated reason, the Nebraska exception, its vetting claim
  2. 2Wedgewood's X thread, archived Oct. 21 · archive.today · the post's caption and the Nebraska claim
  3. 3Pollfish screenshot posted by Wedgewood · Oct. 20 · the only evidence Wedgewood offered for Nebraska
  4. 4Wedgewood's Nebraska release (DDHQ copy) · fetched Oct. 24, 2026 · the Nebraska poll's dates, size and funding; that DDHQ still hosts it
  5. 5Wedgewood's Nebraska GoFundMe · fetched Oct. 24, 2026 · what the crowdfund raised
  6. 6FiftyPlusOne, Oct. 20 · Radcliffe, Morris, Burton · the request for comment, the names Wedgewood gave, the vetting, the raw data
  7. 7Silver Bulletin, Oct. 22 · Nate Silver · the lean, the progressive outliers, the forecast effect, his view of the motive
  8. 8Silver Bulletin table, Oct. 22 · image in the Oct. 22 post · every row of the lean comparison
  9. 9Silver Bulletin forecast note, Oct. 23 · FLIPR landing page · the ban
  10. 10Decision Desk HQ post, Oct. 21 · X · DDHQ's removal and its effect on the forecast
  11. 11State Navigate post, Oct. 20 · X · forecast effect
  12. 12Kalshi, Oct. 22 · Kalshi News · the market effect, by the exchange itself
  13. 13G. Elliott Morris, Oct. 20 · X · the cover-story camp, in its own words
  14. 14G. Elliott Morris, Oct. 23 · X · the betting-market camp
  15. 15Zachary Donnini (VoteHub), Oct. 20 · X · what Wedgewood told VoteHub about itself
  16. 16Nate Silver, Oct. 22 · X · the vetting dispute
  17. 17RealClearPolitics, Oct. 22 · Carl M. Cannon · the aggregator that rejected it
  18. 18FiftyPlusOne on SoCal Strategies, Oct. 22 · blog.fiftyplusone.news · the related firm

Official data

  1. 19Mallory McMorrow, Oct. 20 · X · her account of the effect
  2. 20Mallory McMorrow reply, Oct. 20 · X · the other causes she names
  3. 21Neera Tanden, Oct. 23 · X · the help-progressives camp and its critic
  4. 22FCC Chairman Brendan Carr, Oct. 21 · X · the federal reaction
  5. 23Brendan Carr, Oct. 22 · X · Carr naming the Times
  6. 24FCC Commissioner Anna Gomez, Oct. 21 · X · the FCC dissent
  7. 25AAPOR, Oct. 21 · statement · the industry's reaction

Coverage checked

  1. 26Andrew Gelman, Oct. 20 · Columbia statistics blog · how FiftyPlusOne caught it; Gelman's timing comment
  2. 27New York Times, Aug. 4 (now) · fetched Oct. 24, 2026 · the Times's correction
  3. 28New York Times, Aug. 4 (as of Oct. 4) · held copy · what the Times printed before removing it
  4. 29New York Times, Aug. 11 (now) · fetched Oct. 24, 2026 · the Times's second correction; what real polls showed for Hong
  5. 30New York Times, Oct. 21 · Jess Bidgood · the 15 polls, the vetting, where the Times listed them
  6. 31BETINT, Oct. 22 · Zimmermann, Goldenberg · the Wisconsin market move; the identity accounts
  7. 32USA Today live blog, Oct. 13 (fetched Oct. 24) · fetched 2026-10-24 08:32 UTC · the citation USA Today has not removed
  8. 33USA Today, Sept. 18 (with note) · fetched Oct. 24, 2026 · USA Today's claim that every reference is gone
  9. 34USA Today, Oct. 5 (with note) · fetched Oct. 24, 2026 · a Georgia story USA Today did correct
  10. 35USA Today, Oct. 21 · Sarah D. Wire · USA Today's own account of its correction; its lean description
  11. 36The Hill, Oct. 4 (as of Oct. 4) · Julia Mueller · what The Hill printed
  12. 37The Hill, Oct. 4 (now) · read Oct. 24, 2026 · The Hill's correction
  13. 38PBS News, Oct. 21 · Maria Ramirez Uribe · who cited the polls; the Michigan primary margin
  14. 39Politico, Oct. 22 · Andrew Howard · Morning Score's two mentions; the effect framing
  15. 40Mediaite, Oct. 21 · Tom Durante · MS NOW's Benson +10; the Times tracker note
  16. 41Fox News, Oct. 21 · foxnews.com · the far-left framing; the Texas example; Holdsworth
  17. 42Washington Examiner, Oct. 23 (Maine) · David Sivak · the consistent-bias claim; Maine tracked real polls
  18. 43Washington Examiner, Oct. 23 (Bice) · exclusive · Bice's request; the 'entirely fabricated' line; the Examiner's own removals
  19. 44Newsweek, Oct. 21 · newsweek.com · Newsweek's retractions; the Louisiana poll; early warnings
  20. 45Newsweek, Oct. 22 · newsweek.com · the 'roughly 1 point' figure against Silver's 0.6
  21. 46Michigan Advance, Oct. 21 · michiganadvance.com · the June 24 date, which the poll's own dates contradict
  22. 47MIRS, Oct. 20 · Michigan capitol news service · the June poll's dates and numbers; the other polls that had her at 8 and 9
  23. 48WGVU, Oct. 22 · NPR member station · Susquehanna's undecideds; SoCal on Sims; no Massachusetts registration
  24. 49Magnolia Tribune, Oct. 20 · Mississippi · a campaign using the fake poll
  25. 50Urban Milwaukee, Aug. 19 · Bruce Murphy · the Republican-front camp; an August warning; the Wisconsin result
  26. 51Straight Arrow News, Aug. 20 · Azi Paybarah · an August warning; what Wedgewood said about itself
  27. 52Straight Arrow News, Oct. 22 · Azi Paybarah · the aggregators' lower tier
  28. 53UW Center for an Informed Public, Oct. 23 · Election Rumor Rundown · timing; how each side used it
  29. 54Campaigns & Elections, Oct. 23 · trade press · an error (McMorrow ran for Senate); Donnini's donation
  30. 55PoliAgg removal list · fetched Oct. 24, 2026 · the full poll list, including primary polls Silver's table omits
  31. 56Slate, Oct. 22 · Ian Prasad Philbrick · the good-news-for-Democrats framing; Hong's fundraising use
  32. 57The Hill via AOL, July 7 · Nexstar · the Platner drop-out poll
  33. 58Mediaite, July 6 (corrected) · fetched Oct. 24, 2026 · Mediaite's correction
  34. 59Prediction Edge Georgia page · read Oct. 24, 2026 · a site still showing the fake poll
  35. 60PollingSource Wedgewood page · fetched Oct. 24, 2026 · a site still listing the polls as active
  36. 61Radio World, Oct. 23 · radioworld.com · the legal threshold for FCC action
  37. 62Los Angeles Times, Aug. 17 · latimes.com · the earlier case with the same explanation
  38. 63Michigan Public Radio, Oct. 22 · WVPE · the undecided gap
  39. 64PollingSource Mississippi page (table) · read Oct. 24, 2026 · a live average still using a fake poll
  40. 65PollingSource Mississippi page (text) · captured Oct. 24, 2026, 09:46 UTC · the current average
  41. 66Scopes oversight note, Oct. 22 · scopes.com · a benchmark withheld
  42. 67New York Post via AOL, Oct. 21 · nypost.com · the Senate Majority PAC denial
  43. 68Daily Beast editor's note · fetched Oct. 24, 2026 · who corrected
  44. 69Breitbart, Aug. 2 (with note) · breitbart.com · who corrected
  45. 70Axios, Oct. 21 · axios.com · SoCal's response
  46. 71Fox News headline, Oct. 21 · foxnews.com · the 'boosting far-left candidates' framing
  47. 72AOL copy of The Hill, fetched Oct. 24 · fetched Oct. 24, 2026 · a copy still in circulation
  48. 73Scrutinel, Oct. 22 · published 2026-10-22 · the one race where removing the polls moved a forecast

How we check.

Source reader

Source