Silence And Space
On the parts of games where nothing happens
The procedure in full, including the four conditions, the per-game figures and the places where a stopwatch is the wrong instrument.
Silence And Space
On the parts of games where nothing happens
The procedure in full, including the four conditions, the per-game figures and the places where a stopwatch is the wrong instrument.
Every figure in this journal was produced with a mechanical stopwatch and a sheet of squared paper. No capture software, no telemetry, no automated audio analysis. The instrument is crude on purpose, because the thing being measured is an experience rather than a signal.
A stretch of play is recorded as a silence only if all four of these hold for its entire duration. Fail any one and the clock stops.
Footsteps and breathing count as ambience, not address, even when they are clearly performed. A distant unexplained sound counts as ambience unless it repeats on a schedule, in which case it is a cue. Weather starting is ambience; weather forcing an action ends the silence.
Inventory screens stop the clock, because the player has left the space. Photo modes also stop it, for the same reason and because they usually pause the world.
Each of us plays the whole game to its end at our own pace before any timing begins, so that no measurement is taken by somebody who does not yet know what the game is doing. Timing happens on a second pass.
On that second pass, the stopwatch starts when all four conditions come true and stops the moment any one fails. The figure goes on the sheet with a two-word note about the location and a mark for which condition broke. Where a stretch is repeatable — a crossing, a corridor — we time it at least three times and record all three, because the spread tells you whether the length is a decision or an accident.
Nothing is timed once. A single measurement of a silence tells you about the reader's mood that afternoon.
| Essay | Genre | Longest | Median | Timed |
|---|---|---|---|---|
| Snowfield | Survival | 22m 07s | 4m 10s | 19 |
| Plateau | Open world | 19m 30s | 5m 02s | 28 |
| Lake | Exploration | 14m 12s | 6m 44s | 23 |
| Waiting hall | Adventure | 8m 40s | 2m 18s | 11 |
| Corridor | Horror | 6m 55s | 3m 30s | 16 |
| Garden | Narrative | 5m 02s | 0m 11s | 14 |
| Unwritten A | Tactics | 3m 41s | 1m 04s | 12 |
| Unwritten B | Racing | 2m 12s | 0m 38s | 9 |
| Unwritten C | Action | 1m 55s | 0m 22s | 9 |
Three games were timed and not written about, and their figures are the useful part of the table. The action game's longest uninterrupted stretch in a fourteen-hour campaign is one minute fifty-five seconds. Its median is twenty-two seconds. That is not a criticism — some games are meant to be relentless — but it establishes the scale the other figures sit on.
Note the garden's median of eleven seconds against its longest of five minutes. That single outlier in an otherwise wall-to-wall scored game is the subject of essay six, and the gap between those two numbers is the whole argument.
The bottom row is the one we watch. Six per cent of our recorded silences ended because the reader chose to move on rather than because the game intervened, and those are the only stretches where the ceiling is set by the audience rather than the designer. All six essays in this issue are about stretches in that category or close to it.
A stopwatch cannot measure attention. Two stretches of identical length can feel four minutes apart, and nothing in our procedure captures that. Where it matters, we say so in the essay and give each reader's subjective account separately.
Three readers is not a sample. Every figure here is the experience of three people, one of whom is consistently more patient than the others. When we report that all three of us walked to the unmarked arch, that is three data points and should be read as an anecdote with a number attached.
Knowing the game distorts the second pass. Timing on a replay means the reader knows nothing is coming, which almost certainly makes the silences feel calmer than they did the first time. We accept this because timing on a first pass would require pausing to write, which breaks the condition being measured.
Ambience is a judgement call. The line between a mixed-in ambient layer and a very quiet piece of music is not always clear, and in four cases across this issue we disagreed among ourselves about which side of it a sound fell on. Those four stretches are excluded rather than argued about.
We do not name games, do not score them, do not publish anything before finishing them, and do not accept review copies. Nothing in this journal is a purchasing recommendation, and the figures are not benchmarks — they describe six specific stretches of six specific games as three people experienced them.