SAND: Raiders of Sophie

SAND sold three hundred thousand copies in under two weeks. Its lifetime review score is 50% positive. Of the hundred most recent English reviews, all posted this month, forty-one recommend it.

A game doesn't move that fast in that direction because it looks bad. SAND looks great. It moves that way because of what happens when you press play.

The good idea

You pilot a trampler — a walking fortress, several storeys tall, that you crew, fuel, repair and rebuild. You take it out into a desert full of wrecks and hostile machines, you strip what you can carry, and you try to reach an extraction zone before something kills you. Other crews are out there doing the same thing, and they would rather take your load than earn their own.

It's a strong premise, and the execution of the fantasy is the best part of the game. The machines have genuine mass. The desert is atmospheric. Climbing around inside your own walking base while someone shoots at its legs is a specific and memorable kind of panic, and the loop of hauling something valuable home under threat delivers exactly the tension extraction games are built on.

If the rest of this review is unkind, it's because that foundation is worth being unkind about. The concept is good enough that the problems read as fixable rather than fatal.

The fight isn't fair, and not in the intended way

The single biggest complaint, and it appears in negative and positive reviews alike, is that there is no gear-based or tier-based matchmaking.

A new player takes their starter trampler into a lobby containing fully-upgraded machines with the best available loadouts. The numbers players report are stark: you die in about five shots, and you need something like eighty to threaten them. Thirty minutes of careful scavenging ends in seconds, and nothing you can do in that fight changes the outcome.

That isn't difficulty, it's a structural mismatch, and it's the mechanism by which an extraction game bleeds new players. The genre's usual counterweight — that skill and positioning can beat gear — doesn't hold when the gap is this wide. Players are asking for base-trampler-only lobbies, and the developers have said matchmaking work is underway; Update #6 on 4 August added a duo-only queue and resized crew brackets, which helps team-size fairness but not gear disparity.

Teaming and server-jumping compound it. When players in a supposedly-competitive queue coordinate across teams, the mode stops meaning anything, and there isn't yet an effective answer to it.

Cheating, and the cure that's hurting

This is the part I'd want explained before spending forty dollars.

Cheating became severe enough that on 25 July the developers ran a ban wave through BattlEye and published the results — including a list of banned Steam IDs, partially redacted — while stating their analysis put the wave at "at least 97% accurate." The remaining three per cent, by their own framing, may be false positives.

Three per cent of a large ban wave is a lot of real people, and the remedy offered is to appeal to BattlEye directly. The announcement also warns that repeated login attempts while banned make the ban harsher.

Meanwhile the anti-cheat is affecting players it didn't ban. Multiple recent reviews describe being kicked mid-expedition on ordinary disconnects or server problems, losing hours of accumulated loot, money and progress. In an extraction game, a wrongful kick isn't an inconvenience — it's the deletion of the entire session's work. There is no restitution system, and in the July Q&A the developers said plainly that they can't identify which losses were caused by cheaters, so compensation isn't happening.

To their credit, they also explained why the fix is slow, and the explanation is the most illuminating thing they've said. Properly stopping these cheats requires rewriting the character movement and interaction systems so the server can validate what the client reports. In other words, the game shipped with too much authority sitting on the player's machine — which is precisely the architecture that makes speed, no-clip and aim cheats easy and detection hard.

That's an honest answer, and it's also a warning. This isn't a tuning problem or a detection-rules problem. It's a rebuild, and until it lands the situation improves only at the margins.

Everything else

Underneath the two big issues sits a layer of ordinary Early Access roughness: ladders and doors that take several attempts, stairs you have to jump up, loot vanishing out of your hands, sound effects firing when they shouldn't, and servers that crash while you're winning. Individually these are forgivable. Stacked on top of losing runs to cheaters and anti-cheat alike, they compound into a game that feels like it's taking things away from you.

The developers are working — six updates in six weeks, real balance changes rather than cosmetic ones, a public Q&A, and a reversal on duo queues within three weeks of arguing against them. Nobody should accuse this team of coasting. Dark and Darker took a similar route through a rough opening period and came out the other side, so it can be done.

5.5 out of 10. Rated 5.5 out of 10.

That score is for the game as it exists this week, and it is not a judgement on where it's heading. The premise is worth an 8, the desert is worth looking at, and the studio is behaving well. But a PvPvE extraction game is a promise about fairness, and right now SAND can't keep it — not against gear, not against cheaters, and not against its own anti-cheat.

At nearly forty dollars, wishlist it and watch for one specific thing: the movement and interaction rewrite. When that ships, and gear-based matchmaking with it, this becomes a very different recommendation. Until then you're paying premium Early Access money for a fight you frequently can't win and occasionally aren't allowed to finish.