Every score is arithmetic on what independent reviewers said about each topic. Nobody picks a product’s number; the only editorial choice is how much each topic matters, and those weights are all below. This is the whole recipe, and where 613 reviews actually land.
Reading a score
Reviewers found more to criticise than to praise.
6 reviews · 1%Praise and criticism roughly balance out, or the praise is faint.
150 reviews · 24%Clearly positive overall, with real caveats worth reading.
386 reviews · 63%Praised on most of what matters; the caveats are minor or niche.
70 reviews · 11%Near-universal praise on the topics that matter most. Rare.
1 review · <1%The recipe
Every independent source’s verdict on a topic (bass, comfort, value…) is placed on a scale from −2, broadly criticised, to +2, broadly praised. They’re averaged, weighting sources by expertise and independence; copies of other reviews are collapsed so they don’t count twice.
A topic’s weight is its importance for the category (below) times confidence: it reaches full weight once 8 independent sources cover it. Weights are rescaled over the topics a product actually has.
The weighted average sentiment becomes the score: 60 + 20 × average.
To a whole number. It’s the same number on the review, in every list and in search results.
Neutral sits at 60, not 50, so the number reads like a rating: an average of “clearly liked” is 80, four stars, and only universal praise reaches 100.
A worked example
Tonality carries 17% of the score and reviewers praise it, so it adds the most: +4.3. Bass costs the most: −0.6. Every review shows this breakdown when you click its score.
Read the Sennheiser HD 600 review60neutral+15praise−1criticism=74score
What each topic is worth
Overall tonal balance is the strongest predictor of whether people like a headphone, and treble is the most common dealbreaker. Comfort matters as much as any single frequency region: you wear these for hours. Isolation isn’t scored on open-backs: leaking sound is their design.
The same sound-first order as headphones, with fit sharing top billing: an IEM that doesn't seal loses its bass. Isolation counts more because IEMs travel; soundstage less, because every IEM images inside your head.
Electronics should disappear: no colouration, no hiss (which sensitive IEMs expose). Power only has to be enough for what you'll plug in. Past that, usability and value.
Transparency and power share the top: an amp's job is to drive the load cleanly. Noise follows closely, because hiss is the most common complaint with sensitive IEMs.
A DAC is judged on transparency and a clean noise floor, then on the inputs, outputs and filters it offers. It has no power topic: a DAC doesn't drive headphones.
A controller lives or dies by how it feels under your hands and how well it talks to your DAW. Keys and pads count only when the controller has them, so a pad controller isn't marked down for having no keybed.
Monitors are tools, so translation (mixes that hold up on other systems) sits beside tonality at the top, and self-noise matters at nearfield distances.
Hi-fi speakers are about enjoyment and the room: tonal balance first, then how they image, stage and fit the space. Amplifier demands barely register.
Where scores land
lowest to highest middle half median
Compare scores within a category. Electronics are judged on fewer, more agreed-on topics, so they run higher than headphones and speakers.
Read with care
A 76 doesn’t mean 76% approval. It’s the weighted average sentiment, placed on a 0–100 scale.
A topic that half the reviewers love and half dislike averages out near neutral. Each review lists its divided topics; read those before you buy.
Headphones, amps and controllers are scored on different topics with different weights.
A great-sounding product can score lower if reviewers think it costs too much, and only by as much as Value weighs.
Scores update when a review is rebuilt with new sources, and when the weights on this page change.