Ice
A small figure, an endless frozen slope, and a surface so slick you can never quite stop. Ice is the first game in months that made me feel the gap between the line that keeps you alive and the line that scores — and it holds those two lines apart on purpose.
Verdict. A beautifully judged study of momentum that plays clean in short bursts. The only marks against it happen in the seams between runs, not on the ice itself.
Scorecard
Holding a side to carve is intuitive and the drift eases off exactly as steeply as you expect. A touch of dead zone in the first frame of a hold keeps it out of the higher band.
You understand the whole thing in two runs without a word of tutorial. Steepening slope, one input, a visible streak counter — everything to learn is already on the screen.
Runs land at the right length and end cleanly, and you can walk away mid-descent with nothing lost. Points off because the advert cadence stretches every third restart by an unpredictable half-minute.
The streak counter is a genuinely good thing to chase and the skill ceiling is real, but the hazards stop introducing new shapes after about a week. I still opened it on day fourteen.
Adverts sit between descents rather than inside them, which is right. Parking the revive prompt under the thumb's resting spot is not, and it caught me often enough to read as a choice.
- Genre
- Frozen-slope glider, single-player, distance and streak attack
- Session length
- About 30 seconds to 3 minutes per descent
- Controls
- Hold the left or right side of the screen to carve that way; hold longer for a harder cut; release to drift straight; a quick opposite tap bleeds off speed
- Played on
- 3 GB budget handset and a mid-range 2025 phone
- Cost
- Free to play; supported by short video adverts between descents
What you actually do
You steer a small figure sliding down a frozen slope that scrolls forever. The surface is slippery by design, so you are never standing still: you are always carrying momentum in some direction, and the job is to point it well. Hold the left side of the screen to carve left, hold longer for a sharper cut, let go to drift straight while the slope keeps pulling you down. Every stretch of clean descent adds to your distance, and that is the whole of the counting.
Over the top of it sits a glide streak in the corner. Keep a smooth line and it climbs, adding more to each stretch; clip anything or scrub off a lot of speed and it drops to nothing. That number is the game's whole argument, because the line that keeps you safe and the line that keeps the streak alive almost never coincide — the game lives in the gap between them. The hazards themselves read instantly: cracks fork across the ice and you weave around them, darker patches are thin ice that shatters if you dwell, snowdrifts are safe but scrub your momentum to nothing, and late gaps of open water can only be crossed by arriving fast enough to skim the top.
How it feels in the hand
Very good, and the main reason this entry exists. The carve responds on the frame you press, and the drift when you release decays at a rate that quickly becomes muscle memory, so after twenty minutes you are placing the figure rather than reacting to it. I played it one-handed, thumb only, on a moving train — a high bar most games in this shape trip over. My one complaint is a sliver of dead zone at the start of a hold: on the older handset a fast flick-and-hold sometimes lost its first fraction of travel, which on a tight line is the difference between clearing a crack and clipping it. It never once affected the newer phone, but on a long streak run it ended me more than any hazard did.
The best descents are the ones where you stop steering around each hazard and start steering the momentum, setting up the line two obstacles before you reach it.
Session shape
A poor descent is over in thirty seconds; a good one runs two to three minutes; the best I logged lasted a little over four and ended, as they all do, in a gap of open water I saw coming and glided into a fraction too slow. That is exactly the right length for the moments a phone gets picked up in — a queue, a kettle, the last few minutes before something starts.
The friction is not in the descents but in the joins between them. After roughly every third failure there is a video advert of fifteen to thirty seconds. The adverts are fine and skippable at the end, but the game also offers a revive for watching one and puts that offer where the restart button appears half a second later. Twice in my first week I tapped the revive by accident because my thumb was already on the way — a layout a studio chooses deliberately, and the single thing holding this score down.
There is also a cosmetic collection — different figures and trails, none of which change how the game plays — unlocked with an in-game currency you earn only by playing. I unlocked five over the fortnight without trying, and nothing sits behind a wall you cannot walk over with a bit of time.
Where the difficulty actually comes from
Not from raw speed. The slope steepens and the figure quickens, but smoothly enough that your hands keep up. What tightens is density: cracks arrive closer together, and late descents stack a crack field directly before a water gap — so the moment you most want to scrub speed to thread the cracks is the moment you most need to keep it for the jump. Being a hair too slow at the lip of the water is the commonest way a strong run ends.
The thin ice is the fairest hazard and the one that catches most people, because it punishes hesitation rather than speed. The darker shade warns you clearly, and as long as you keep moving across it you are fine. My distances went up the day I stopped watching my own figure and started watching the ice two lengths ahead of it.
Five pointers
- Learn to feather on day one. A quick opposite tap sheds a little speed without breaking your streak; a hard carve sheds a lot and usually breaks it. Most of the intermediate skill is choosing the small correction over the big one.
- Read which way a crack forks before you commit. Cracks spread away from their wide end, so the narrow side is where the ice stays whole. Aim for the point, not the gap you can see right now.
- Protect your speed going into open water. The instinct is to slow at the edge, and the water punishes exactly that. Set your line early and arrive fast — the skim across is forgiving, the slow approach is not.
- Treat a snowdrift as the brake, not the enemy. When a crack field and a water gap stack up with no clean line through, a drift costs the streak but saves the descent. Losing the counter is cheap; losing the run is not.
- Wait a full second after a wipeout before touching the screen. It is the cheapest fix for the accidental revive tap, and it stops you restarting while still annoyed — which is how the next bad descent begins.
Who it is for
Anyone who wants a game about flow and momentum with no meta-progression to manage, no timers to wait out, and no obligation to return on day two. If you resent a game that pings you at six o'clock, this is the opposite of that. If you want depth that unfolds over months, look at Garuda in entry 43 instead — Ice is a short, sharp, honest thing, and content to be exactly that.
This page is an informational write-up of a free casual video game. Ice as played above has no paid element at all — a distance counter is a distance counter, nothing is put in and nothing comes out — and the entry exists to record what playing it was like, not to promote it. Thumbstick Diary operates independently of every studio and app store, and the title is a trade mark of whoever published the game, named here purely to identify the subject.