How vsfind scores work

In short: the vsfind score rates a device's listed specifications against fixed scoring curves and weights that we publish below; its rank shows how it compares with the other devices in the category. We don't test devices hands-on, and the score is not a review or a user rating. Below we show exactly which specs count and how much.

Where the data comes from

Specifications are collected from public spec sheets (the source is linked at the bottom of every product page) and simplified: “5G” instead of band lists, “Wi‑Fi 7” instead of long protocol strings. The original text stays available under “Show raw spec text” in every section. We don't display other sites' own lab measurements or use them in scores; official data such as the EU energy label is used.

Field scores (the bars)

Every scored field gets a 0–100 score, and the length of its bar is that score. For numeric fields the score is read from a fixed curve listed in the tables below: straight lines between a few reference points, flat beyond the ends. For battery capacity, for example: ≤3000 mAh → 10 · 4000 mAh → 50 · 5000 mAh → 80 · 5500 mAh → 90 · 6000 mAh → 96 · ≥6500 mAh → 100. Curves flatten where more stops mattering in daily use, so capacity beyond 6500 mAh, wired charging beyond 90 W or brightness beyond 3500 nits earns nothing extra. Weight and thickness are gentle penalties: a big, heavy flagship loses a few points, not most of them. Fields like panel type, refresh rate or IP rating use hand-written tier tables; yes/no features count as yes = 100, no = 0.

Why fixed curves instead of percentiles?

The first version scored every field as a percentile against phones from the last 24 months. In a crowd of very similar phones that turned small spec differences into big score gaps: once many phones shipped with 7000–10000 mAh batteries, a 5000 mAh flagship's battery score fell into the 20s even though its real endurance was almost the same. With fixed curves a phone's score depends only on its own specs and doesn't change when other phones are added; the rank gives the relative view. The curves are calibrated to today's market and revised, with a new methodology version, when the market moves. The chipset score comes from a tier table kept relative to the newest flagship generation, so performance scores age naturally as new generations arrive.

Battery: when the official EU energy label lists an endurance figure, it outweighs raw capacity (mAh), because it is the best available measure of real battery life (≤30:00 h → 10 · 40:00 h → 45 · 50:00 h → 75 · 60:00 h → 92 · ≥70:00 h → 100). Other sites' own lab measurements are not used.

Unknown is not “no”. A value missing from the source is left out of the score and shown as “Not specified” or “?”. Taste-dependent fields such as screen size and operating system are not scored and have no bar.

Section and overall scores

A section score is the weighted average of the known fields in that section. If the known weight is below 60% of the section's weight, the section gets no score (“Not enough data”). The overall score is the weighted average of the known sections, rounded to a whole number. The best device is not stretched to 100; instead we show its rank in the category (e.g. “#7 of 458 phones”). Performance (chipset tier) and camera hardware (sensor size, optical stabilization, optical zoom) carry the most weight, because they separate classes of phones far better than RAM, storage or raw mAh.

Pros, cons and “who is it for”

Pros come from fields scoring 85 or higher and from rare features (e.g. a 3.5 mm headphone jack), at most 3. Cons come from fields scoring 20 or lower and from important missing features (e.g. no NFC), at most 3. Summaries and “who is it for” are fixed templates filled only from this data; no free-form text is generated.

“Better” in comparisons

The “better” mark is only given on scored fields, to the higher field score; ties get no mark. In the section summary, gaps smaller than 3 points count as even.