You are surveying an abandoned valley in winter with a limited number of daylight hours and a shelter you must return to. There is no combat. The entire game is going out, finding something, and getting back, and it is one of the most consistently tense things we have played in years.
The reason is that the snow keeps a record.
The rule
Footprints persist in snow and are erased by wind-driven drift. The erasure rate is proportional to wind speed, which is displayed nowhere but is legible from spindrift on ridges and from the sound of it against your hood.
Since your route back is often your own trail, the wind is a timer on your safe return. In still air you can wander for hours and follow your prints home. In a rising wind your trail has perhaps twenty minutes, and the decision of when to turn back stops being about daylight and becomes about air movement.
| Wind condition | Track lifetime | Trials |
|---|---|---|
| Still | Beyond one in-game day | 6 |
| Light breeze | ~55 min | 9 |
| Fresh wind | ~21 min | 11 |
| Strong wind | ~7 min | 9 |
| Storm | Under 2 min | 5 |
Why consistency is the whole point
Those numbers held across nine storms within a few per cent, which is the finding that matters more than the numbers themselves. A system that behaves the same way every time can be learned; a system that varies for dramatic effect can only be endured.
By hour twenty we were making genuine forecasts — not from an interface, but from the way loose snow moved across a slope. Getting that wrong twice and being caught out in the open is how the game teaches, and it teaches without a single tutorial line.
The measure of a weather system is whether a player can be wrong about it in an instructive way. Randomness cannot teach; consistency can.
Symmetry
Partially verified, and this is where the score comes down from a nine. The valley has animals, and they do respond to weather: they shelter, they move less in storms, and they leave tracks that also fill in.
But the tracks left by animals are erased at a slightly slower rate than ours in identical wind, which we measured deliberately across eleven paired observations. It is a small asymmetry, it is presumably a legibility decision, and it is the only inconsistency we found in forty-one hours. Worth noting rather than worth condemning.
The eighteen-minute cycle
Well judged. Long enough that a decision to press on has consequences, short enough that a session shows you several states. Transitions build over about two minutes with visible precursors: light lifting off ridges, the sound of drift changing pitch, the temperature reading dropping before anything is visible.
We timed thirty transitions and found no case of a snap change, which after Fenwatch was a relief.
Where it is weaker
The survey mechanic itself is dull: you stand in a marked place, hold a button, and a bar fills. Given how good everything around it is, the actual objective is the least interesting thing in the game.
There is also very little variety in what you find. By hour thirty, discovery had stopped surprising us, and the tension of the journey was carrying a destination that no longer justified it. A longer game would have needed more, and this one runs about ten hours past its material.
| Element | Assessment |
|---|---|
| Track persistence and erasure | Consistent, learnable, excellent |
| Wind legibility without interface | Exemplary |
| Animal track symmetry | Slightly favours animals |
| Transitions | Gradual, well telegraphed |
| Survey objective | Dull; a held button and a bar |
Where this leaves it
A game that gets an extraordinary amount out of one rule applied without exception, and which demonstrates the argument this publication keeps making: a weather system does not need to be complicated, it needs to be consistent, symmetrical and legible without an interface. This one is three for three, and the flaws lie entirely elsewhere.