The weights are public
Every criterion and its weight is published before the review. They live in code, not in the post, so no single review can quietly re-weight itself.
Independent · evidence attached
Hands-on reviews of apps, websites, AI tools and hardware — tested against a published rubric. Every score shows its weights. Every review shows the version, the devices, and the hours.
0–10
Published scale with a written anchor for every band.
6–7
Weighted criteria per product type, fixed in code.
5+
Evidence items required before a review can publish.
0
Scores given to products we built ourselves.
Most review sites ask you to take the score on faith. Three things make ours checkable.
Every criterion and its weight is published before the review. They live in code, not in the post, so no single review can quietly re-weight itself.
Screenshots carry a capture date and device. Measurements carry their method and run count. A review without them cannot publish — the build refuses it.
SofNerds builds Android utilities. None of them carries an editorial score, a ranking, or review markup. They are build logs, clearly labelled.
Proof we use the whole scale
No scored reviews yet
Every product on this site so far is one SofNerds built, and we do not score our own work. This chart fills in as independent reviews publish — and it will show the bottom of the scale as well as the top, because that is the only version of it worth publishing.
Made by us · unscored on purpose
Build logs, not reviews. The decisions, the data, the things we got wrong. No scores — we do not grade our own homework.
A category opens to readers once three products have been reviewed in it. Below that it is a placeholder, and we say so. The full taxonomy runs to 10 categories.