What Vugol Me is

Polling averages get quoted as fact, and that makes them a target. Fake pollsters with plausible names have successfully placed invented numbers into widely read averages. Vugol Me was built in response: every pollster starts in quarantine, visible but excluded, until a human has verified the organization actually exists, publishes methodology, and answers for its work.

1. A poll arrives as a raw document, content-addressed and kept.
2. Staff review the transcription against the pollster’s own publication.
3. Approved results enter a normalized record; only verified pollsters’ results are eligible for aggregation.
4. The aggregate is computed by a versioned, published model and frozen as an immutable release with a checksum. Anyone can recompute the number from the release manifest.

The full method, including how partisan-sponsored polls are down-weighted and shared samples collapsed, is on the methodology page. The rules for adding, excluding and correcting sources, and the response times we hold ourselves to, are on the editorial process page.

Candidate profiles are sourced from public government records (the FEC for federal candidates), reviewed before publication, and carry the provenance of every datapoint. Candidates can claim their profile; anything they add themselves is always labeled campaign-provided and never mixes with sourced facts.

Signed-in visitors can vote in an open community poll for every race. Those are self-selected votes, they are labeled as non-scientific everywhere they appear, and they never touch the aggregates.

Vugol Me is published by VEDAVERE, LLC. The responsible editor is reachable at vugolme@gmail.com; that address also handles support, pollster verification, profile claims and press. The site uses only aggregate visit analytics, disclosed in the privacy policy.

• Subscriptions are the only revenue. Anyone may subscribe on identical, published terms, including campaigns, parties and committees. The terms of use disclose this.
• No sponsorships, no paid placement, no advertising. No rating, roster entry, poll verification or release can be bought, and none is sold.
• Subscribers never influence ratings, rosters or releases. Who pays has no channel into what is published. Editorial decisions follow the written editorial process, and every published number carries the model version and inputs that produced it.

Errors are corrected on the record: a superseding release is issued and the earlier one stays visible, marked superseded. How to report one and how quickly we answer is on the corrections page.