Trending news
Automation pulls headlines from neutral wire services every 4 hours; AI proposes questions, editors review.
How we work
Every survey is a question + the context to answer it.
Each published survey ships alongside the glossary terms, civics guides, and issue briefs that explain the question. No survey runs in isolation.
Survey sources
Where survey questions come from.
Evergreen civics
Durable questions about American government — Electoral College, filibuster, judicial review — seeded from an editorial topic list.
Reader requests
Survey ideas submitted via the contact form. Editors evaluate every suggestion against the neutrality bar.
Editorial calendar
Anniversaries, court terms, debate windows — moments where a question becomes timely.
Voting + integrity
How we keep votes honest.
One vote per person
Anonymous browser fingerprints dedupe repeat votes per survey. No login required. Raw IPs are never stored.
Live tallies
Results recompute continuously. State + political-lean breakdowns appear once a survey crosses minimum-sample thresholds (10 votes per state, 5 per lean).
Seed baselines
New surveys ship with a small randomized vote baseline (50–200 votes) so the page renders coherently before the first organic voter arrives. Seeded counts are documented and dilute as real votes accumulate.
Balanced briefs
Every issue brief must include the strongest version of both the supporting and opposing argument. Editorial review rejects briefs that read one-sided.
What these results don’t mean
Survey results are sentiment snapshots — not scientific polls.
Site visitors are self-selected. Demographic breakdowns reflect who chose to vote, not the U.S. population. Where we apply scientific weighting (rare), the survey is labeled accordingly and the methodology is published alongside the result.
Want to go deeper?
Read how the platform is built.
Editorial standards, neutrality mandate, and the data architecture behind surveys all live in the public spec.