See how the evidence on a contested question splits — and whether an apparent consensus is real, or just the same few sources counted many times.
Count the sources and whoever publishes most appears to win — even when it's the same lab, funder, or dataset showing up again and again. That's how false balance creeps in.
We show an adjusted evidence-base count, not just headcount. Ten papers using one dataset count as roughly one evidence base — while quality and shared weaknesses stay separate.
Each source is fetched, labelled with a quote-backed position, and merged into a living map — positions, shared evidence, key disagreements, and what each position leaves out.