You’re scrolling through your feed, and there it is: a poll showing your favorite candidate surging. It feels good. It confirms what you want to believe. But before you hit that share button, pause. That data might be completely fake.
We just saw a textbook example this month. An unknown outfit called Median Strategies dumped a series of "polls" into the political ecosystem. They claimed to show shifts in races from Wisconsin to Los Angeles. Campaigns—even incumbents who should know better—amplified these numbers to brag about their momentum. Days later, the company folded, admitting the whole thing was a "social experiment."
It was a wake-up call. It’s never been easier to manufacture legitimacy. If you’re not vetting the source, you’re just a tool for someone else’s misinformation.
The Mirage of Modern Polling
The barrier to entry for releasing a "poll" is basically zero. You don't need a degree in statistics or a decade of experience in survey research. You need a website, a Twitter account, and the ability to cook up a spreadsheet of numbers that look plausible.
When Median Strategies released its bogus data, major polling aggregators didn't touch it. Why? Because the firm provided zero transparency. They couldn't name the vendor who fielded the sample. They couldn't explain their voter file sources. They refused to answer basic questions about their methodology.
Reputable pollsters follow the transparency standards set by organizations like the American Association for Public Opinion Research (AAPOR). They disclose:
- How they reached respondents (phone, email, text).
- The exact wording of the questions.
- Their weighting process to account for demographics.
- Who paid for the survey.
If a poll lacks this, it’s not data. It’s a press release. Treat it like one.
Why Campaigns Fall for the Bait
You might wonder how a professional campaign team could fall for such an obvious trap. It happens because of confirmation bias. Campaigns are desperate for good news to energize donors and volunteers. If a "study" shows you’re winning, you don't look for flaws—you look for a megaphone.
Once a candidate shares a fake poll, it takes on a life of its own. Media outlets, sometimes hungry for content, report on the campaign’s "internal momentum." The cycle reinforces itself. By the time someone digs into the methodology, the narrative is already set. Damage done.
How to Protect Yourself from Bad Data
Don't be a vector for misinformation. Use these steps to sanity-check any numbers you see online.
Check the Aggregators
If you don't see the numbers on sites like RealClearPolitics, The New York Times, or established academic polling centers, be skeptical. These aggregators have teams whose entire job is to vet methodologies. If they aren't citing it, there's usually a reason.
Look for the Transparency Statement
If a polling firm hides behind a generic "contact us" form and doesn't explicitly state their sampling techniques, walk away. Legitimate firms are proud of their process. They want you to know how they arrived at their conclusions because that’s their product: credibility.
Consider the Sample Size and Source
A "poll" of 200 people on a social media platform isn't science. It’s a focus group of the loudest voices in a room. Look for randomized sampling that reflects the actual population of the district or state.
Watch for "Push Polling"
Sometimes the poll isn't "fake" in the sense that the numbers were pulled from thin air, but it's manipulative. Push polls use leading questions designed to plant an idea in the respondent’s head. If a question asks, "Would you still support Candidate X if you knew they were a [scandalous description]?" you aren't being polled. You’re being sold a smear.
The Bottom Line
Polling is an industry built on trust. When that trust is eroded by bad actors running "experiments" or partisan groups pushing propaganda, the whole system suffers. Don't let your desire for a specific election outcome blind you to the reality of the data.
Demand transparency from the firms you see cited. If the information isn't there, the poll doesn't exist. Stop sharing unvetted numbers. Start looking for the fine print.
Your vote is your power. Don't let a manufactured spreadsheet tell you how to use it.
