How we run polls, protect data integrity, make editorial decisions, and handle participant identity.
AmericanElectionHQ publishes opt-in public sentiment polls. These polls are not scientific probability surveys. They are a tool for measuring the preferences of engaged visitors — people who sought out a specific race page — and giving them a way to participate in the national conversation about the races that matter to them.
Our polls are voluntary, opt-in sentiment surveys. Any visitor to a race page can participate. There is no invitation-based recruitment, no random selection, and no attempt to build a statistically representative sample of any population. Results reflect who chooses to engage with a race on this platform — not the broader electorate, registered voters in a jurisdiction, or any other defined population.
Respondents self-select. Anyone who finds a race page through search, a shared link, social media, or direct navigation can vote. We do not recruit participants, weight responses by demographics, or apply any sampling correction. Results can be influenced by which communities share a link, where traffic originates, and when during a news cycle a race receives attention.
We enforce one response per session per poll using a session identifier stored in your browser's local storage. When you submit a vote, your session ID is linked to your response. If you attempt to vote again in the same poll from the same browser, the system returns your original response rather than recording a new one. A user who clears browser data, uses private browsing, or switches devices can technically participate again — this is a known limitation of session-based deduplication without account creation.
Submission requests are rate-limited per IP address (five requests per ten-second window) using a sliding window algorithm. Cloudflare Turnstile is active on Step 1 of the poll flow — all votes require a passing security check before being written to our database. These measures significantly reduce automated and scripted abuse.
Our systems automatically flag responses that match known abuse patterns — including anomalous submission timing, repeated IP clusters, and response sequences inconsistent with organic behavior. Flagged responses are held for review by our moderation team before they affect displayed vote counts. We make no guarantee that displayed results are free from organized participation efforts, but we actively work to identify and remove manipulation.
AmericanElectionHQ publishes opt-in public sentiment polls. These polls are not scientific probability surveys. They are a tool for measuring the preferences of engaged visitors — people who sought out a specific race page — and giving them a way to participate in the national conversation about the races that matter to them.
Our polls are voluntary, opt-in sentiment surveys. Any visitor to a race page can participate. There is no invitation-based recruitment, no random selection, and no attempt to build a statistically representative sample of any population. Results reflect who chooses to engage with a race on this platform — not the broader electorate, registered voters in a jurisdiction, or any other defined population.
Respondents self-select. Anyone who finds a race page through search, a shared link, social media, or direct navigation can vote. We do not recruit participants, weight responses by demographics, or apply any sampling correction. Results can be influenced by which communities share a link, where traffic originates, and when during a news cycle a race receives attention.
We enforce one response per session per poll using a session identifier stored in your browser's local storage. When you submit a vote, your session ID is linked to your response. If you attempt to vote again in the same poll from the same browser, the system returns your original response rather than recording a new one. A user who clears browser data, uses private browsing, or switches devices can technically participate again — this is a known limitation of session-based deduplication without account creation.
Submission requests are rate-limited per IP address (five requests per ten-second window) using a sliding window algorithm. Cloudflare Turnstile is active on Step 1 of the poll flow — all votes require a passing security check before being written to our database. These measures significantly reduce automated and scripted abuse.
Our systems automatically flag responses that match known abuse patterns — including anomalous submission timing, repeated IP clusters, and response sequences inconsistent with organic behavior. Flagged responses are held for review by our moderation team before they affect displayed vote counts. We make no guarantee that displayed results are free from organized participation efforts, but we actively work to identify and remove manipulation.
Our goal is a platform where results reflect genuine public sentiment — not manipulation, gaming, or coordinated campaigns. This page describes our commitment and the specific controls we have in place.
AmericanElectionHQ is an opinion platform, not an official voting system. Our polls do not affect any election outcome. But the integrity of our results matters: bad data misleads voters, distorts public discourse, and undermines the purpose of the platform. We take that seriously.
We operate three layers of defense against result manipulation:
All responses are written to a secured database. Vote counts are derived from raw response records — there is no separate tally that can be manipulated independently. Result percentages are calculated at query time from the live record set, so removing a manipulated response immediately affects the displayed result.
Campaigns, supporters, and third parties are prohibited from organizing coordinated voting efforts designed to artificially inflate results. In 2025, we identified and removed approximately 50 responses in a gubernatorial race that matched the pattern of a coordinated campaign — submissions clustered by timing, IP range, and identical enrichment data. Those responses were removed and the published results updated before broader public attention reached the race. We will continue to enforce this policy without advance notice to the parties involved.
If you believe a result on this platform has been manipulated, contact us at integrity@americanelectionhq.com with the race name, the specific concern, and any supporting information. We investigate all credible reports.
Our controls reduce manipulation but do not eliminate it. A sufficiently distributed and patient coordinated campaign — using residential proxies, varied devices, and staggered timing — could affect our results in ways our current systems might not catch. We are honest about this limitation. Our results should always be interpreted as sentiment indicators from a self-selected sample, not authoritative measurements of voter intent.
This page describes how we decide which races to cover, how poll questions are written, how we handle corrections, and what editorial independence means for this platform.
Coverage decisions — which races we add, when we add them, how we present candidates — are made solely by our editorial team based on the significance of the race and the interest of voters. No candidate, campaign, political party, PAC, donor, advertiser, or government entity has any influence over these decisions. We do not accept payment to add a race, feature a candidate, or weight results.
We prioritize races based on their likely impact on voters and the availability of meaningful candidate information. Federal races (US Senate, US House) are added first because they attract the broadest national interest. Statewide races (Governor, Attorney General, Secretary of State) are added next. Local and municipal races are added when there is clear public interest or when we have the candidate data to cover them responsibly. We do not cover every race in every jurisdiction — our goal is depth over breadth until the platform matures.
Candidate polls use a standardized question: “Who do you support in this race?” — applied consistently across all races. We do not tailor question wording to favor any candidate. Opinion polls (those not tied to a specific candidate race) are reviewed before publication to ensure the question and answer options are balanced and do not presuppose a preferred response. We do not publish push-poll style questions designed to lead respondents toward a particular answer.
If we publish inaccurate information about a candidate, race, or result — including errors in candidate names, party affiliations, election dates, or district information — we correct it as quickly as possible. Corrections are made in place; we do not maintain a separate corrections log at this time. If you spot an error, contact us at editorial@americanelectionhq.com.
For editorial inquiries, race suggestions, or corrections, contact editorial@americanelectionhq.com. For press inquiries, see our Press page.
This page explains how we handle participant identity, why we do not require voter registration verification, and what that means for how you should interpret our results.
Participating in an AmericanElectionHQ poll does not require proof of voter registration, citizenship, or residency. There is no age verification requirement. This is intentional. AmericanElectionHQ is an opinion platform, not an official voting system. Our polls do not affect any actual election. They are a tool for measuring public sentiment and giving people a voice in the national conversation about the races that matter to them.
During the poll flow, we collect your name, email address, and ZIP code. We collect this information for two purposes:
We do not verify that the name, email, or ZIP you provide is accurate or associated with a real registered voter.
We enforce one response per session per poll using a session identifier stored in your browser's local storage. When you submit a vote, your session ID is linked to your response. If you attempt to vote again in the same poll from the same browser, the system returns your original response rather than creating a new one. A user who clears their browser data, uses a private browsing window, switches to a different browser, or uses a different device can technically participate again. This is a known and inherent limitation of session-based deduplication systems that do not require account creation. We address this through complementary controls — rate limiting, security checks, and behavioral flagging — but we cannot guarantee that every duplicate submission is caught.
Requiring verified voter registration before participation would create significant problems:
The trade-off is that results should always be interpreted with the understanding that participants are self-selected, not verified voters.
We are evaluating the possibility of an optional verified-voter badge for users who wish to voluntarily verify their voter registration status. This would allow participants to signal that their response comes from a confirmed registered voter, and would allow us to display results segmented by verified vs. unverified participants. This feature is in early consideration only. There is no current timeline for implementation.