About rawvs

rawvs separates what has actually been measured about an ingredient from what is only repeated about it.

Not recipes. Not health claims. One ingredient at a time: I collect the specific claims English cooking sites make about it, then go looking for the primary source behind each one — a food standard, a government composition table, a peer-reviewed measurement. Often the source says something narrower than the claim. Sometimes it says the opposite. Sometimes there is no source at all, and that is the most useful thing an article can report.

Every claim here carries a link you can open and check. Not a reading list at the foot of the page — the citation sits in the sentence that depends on it. Where a figure only holds under particular conditions, the conditions are given with it: which product, which method, which edition of which document.

I have not cooked these ingredients yet, and the articles say so. A table at the top of each page states plainly what the article is built from. At the moment that line reads "not yet — this article is built from published sources" on every one of them. When I do cook something, the prices, weights, temperatures and times go into that same table and the article is updated. Published sources are the floor; cooking notes are added on top of them.

There are no photographs yet. Rather than fill the space with stock images of the wrong thing, the pages run without them until I have taken my own.

Every article ends with what I could not verify. Documents I could not open, figures I could not trace, measurements that turned out not to exist. That section is not a disclaimer — it is usually where the real finding is, because nobody who is mass-producing articles leaves blank spaces in them.

How these pages are written: I use AI tools for the research and for drafting. What the tools return is a starting point, not evidence. Every figure an article is built on, I open the original document and read myself before it goes in — and when the tool has been wrong, which has happened, the correction is in the article. The sources are listed so you do not have to take my word for any of it.

What this site got wrong: in September 2026 I used an AI tool to generate 160 articles from one fixed template and queued them to publish automatically. Google applied a site-wide spam action, correctly. Two requests for reconsideration were refused, and both times the reason was the same: I had been editing those articles rather than removing them. They are now gone — the published ones return 410, the queued ones were never published and will not be, and the scheduling tools were deleted. The site restarted from nothing, with the method described above.

What I don't cover: health benefits, disease prevention, or medical advice. If you have a health or allergy question, please talk to a professional.

rawvs is written by JaeHyun Kim, a home cook in South Korea. Questions and corrections are welcome on the contact page — if a source here does not say what I have said it says, I would like to know.