Paste the strains you are choosing between.
Most strain sites make you open one page at a time and read a paragraph. Paste a whole dispensary menu here and get one table — every cannabinoid and terpene side by side, each with the number of laboratory results behind it and the range those results actually covered.
What this does differently
It shows a range where everyone else shows one number
Every strain page on the internet prints something like "Blue Dream: THC 19%", as though that were a property of the plant. Across 1,793 real laboratory results, Blue Dream ran from 14.4% to 22.2%. The single number is not wrong so much as it is a fiction about how cannabis works: the same name, from different growers and different batches, is genuinely different material. This site shows the spread, and the count behind it, in the cell where the number is.
It does not let three samples beat five hundred
Sort any strain database by an ingredient and the top fills with obscure names that were tested a handful of times, because a small sample can average anything. Here every figure is pulled toward the typical strain in proportion to how little evidence stands behind it, by an amount worked out from the data rather than picked. The arithmetic is on the method page, and the comparison table prints its own working line by line.
Different ingredients are compared on equal terms
THC runs around 18% and limonene around 0.2%. Ranked on raw numbers, every list would secretly be a THC list. Each ingredient is measured in how far a strain sits from a typical strain, so asking for more limonene and less THC gives both requests the weight you actually assigned them.
It never ranks anything for you
The site has no opinion about which strain is good. You set the weights, the weights live in the URL, and the ranking shows the arithmetic that produced it. Somebody shopping for a workday morning and somebody shopping for 11pm are asking opposite questions, and one ranking cannot answer both honestly.
What is missing, said plainly
There is no effect data here. Not because it was overlooked: no source publishes per-strain effect rates under a licence that permits reuse, and the one site that has them forbids both scraping and commercial use. Rather than borrow numbers we cannot stand behind, or invent them, the column is empty and this paragraph explains why. If that changes, the same table will fill in.