How we compare

1. Identical rows for both sides

Every product in a category is recorded against the same specification list, so a comparison cannot be flattered by choosing convenient rows.

2. Scored rows and explained rows

Rows with a sensible direction (more battery, lower price, longer software support) are scored. Rows that are preferences (operating system, flavour, chipset family) or where bigger is not simply better (megapixels, screen size, engine displacement) are explained in prose and, where scored at all, carry a low weight. Missing manufacturer figures are shown as "Not published" and excluded.

3. The VersusAnything Score

Scored rows are normalised against every product in the category and weighted by category-specific factors. The full weighting is on the how we score page and repeated on every comparison.

4. Editorial layer

The verdict, key differences and who-should-buy sections explain what the numbers mean in use. High-traffic comparisons receive hand-written introductions and verdicts; all pages receive a written summary of each product and a stated best-for use case.

5. Sources and verification

Each product lists its sources and the date its specifications were last verified. See the sources and data policy.

6. Curated, not generated

We publish only comparisons with genuine search or user value. The custom tool on the homepage handles any other pair without creating a page.