How Fred works
We’re not another “Top 10” list. We’re proudly synthetic — a transparent layer over the whole review landscape. Here’s exactly what that means, so you can trust the map.
1. We ingest every substantive source
For each product we gather the reviews that actually say something: YouTube field tests and their transcripts, Reddit threads, forum debates, and written blog reviews.
2. We extract structured claims
An LLM pipeline pulls specific, checkable claims (“passed through at 40 yards,” “blade bent on bone”) and maps them to a shared attribute ontology — so a claim about penetration from one reviewer lines up with everyone else’s.
3. We score each source for rigor
Not all reviews are equal. A chronograph and a calibrated gel block outrank “felt great.” Every source gets a rigor score (0–10), and higher-rigor sources carry more weight in the consensus.
4. We publish the consensus — including the disagreements
For every attribute we show the verdict, how strongly reviewers agree, our confidence, and how many sources weighed in. When reviewers genuinely disagree, we say so and explain why. No fake certainty.
5. We link back to everyone
We extract facts and claims — we never republish transcripts or clips. Every creator whose work informed a page is credited and deep-linked. Go watch them; they earned it.
Spot something we got wrong? Fred wants the correction: fred@fredshed.tv.