We graded subtitle implementation in all thirty-eight games on four counts: whether the text can be enlarged, whether it holds contrast against the game's brightest scene, whether speakers are identified, and whether sound effects are captioned at all. Nine games passed everything. Twenty-six caption no effects whatsoever.
Every game in our table has subtitles, and a store page can therefore claim the feature honestly. That claim tells a player almost nothing, because the gap between subtitles that exist and subtitles that work is enormous and entirely invisible until you need them.
We graded on four criteria chosen because each one, on its own, can make the feature useless. Text too small to read at a normal seating distance. Text that vanishes against a bright scene. Dialogue with no indication of who is speaking. And captions that cover speech while ignoring the sounds a game's design depends on.
| Criterion | Pass | Fail | Commonest failure |
|---|---|---|---|
| Size adjustable, readable default | 31 | 7 | Fixed size, too small at 1080p |
| Contrast against brightest scene | 34 | 4 | White text, no backing plate |
| Speaker identified | 26 | 12 | Off-screen voices unattributed |
| Effects captioned | 12 | 26 | Speech only, no effects at all |
Nine games pass all four. Twenty pass three. The single criterion dragging the table down is effect captioning, which is also the one that matters most in a medium where sound frequently carries information rather than atmosphere.
More than two thirds of the table transcribe speech and nothing else. A player relying on subtitles in these games receives the dialogue and no indication of a warning sound, a positional cue, or a change in the environment.
In one game we reviewed, the entire horror mechanism is a sound effect: everything else ducks to silence when the radio speaks. That game captions speech faultlessly and effects not at all, so a subtitle-reliant player receives the words of every transmission and never learns that the world goes quiet around them.
They are not experiencing a slightly diminished version of the game. The technique the whole work is built on is simply absent for them, and no amount of adjustable text size compensates for that.
The same applies less dramatically almost everywhere. A footstep behind you, a mechanism arming, a creature's approach: these are information, and a caption that says "footsteps, behind, close" delivers it. Twenty-six games in our table deliver none of it.
Four games failed on contrast, and all four look perfectly acceptable in a menu or a dark corridor. We test against each game's brightest scene specifically because that is where white text without a backing plate disappears, and bright scenes are common: snow, daylight, water, fog, fire.
The fix is trivial and well understood. A semi-transparent plate behind the text, or an outline, costs nothing and solves the problem permanently. Three of the four failing games offer three text sizes and no plate option, which suggests the feature was implemented by someone who never tried reading it against the sky.
Three size options and no backing plate is a subtitle system designed in a dark room.
Twelve games fail to identify who is speaking, and in almost every case the problem is confined to off-screen or radio voices. When a character is visible and their mouth is moving, attribution is obvious. When a voice arrives from elsewhere, a subtitle without a name is a line of dialogue attached to nobody.
Narrative games suffer most here, and it is the failure we found most puzzling: the writing team already knows who says every line, so the information exists in the project and is being discarded at the last step.
| Element | Minimum | Better |
|---|---|---|
| Size | Readable default, three options | Continuous scaling |
| Background | Optional semi-transparent plate | Adjustable opacity |
| Speaker | Name for off-screen voices | Name and colour per character |
| Effects | Captions for gameplay-relevant sounds | Direction and distance included |
Nine games in our table already do all of this, which disposes of any argument that it is impractical. It is a modest amount of work, done at the end of a project, and it decides whether a substantial number of players can follow the game at all.
