Bottom Line: Antihero compresses a full night of board-game scheming into the length of a coffee break, and almost nothing in the genre matches its economy of design. Buy it for the tactics, not the campaign — and know you're buying an orphan.
The Gameplay Loop
The loop is three verbs deep and bottomless. Move your thief. Spend your gold. Deny your opponent. That's it — and the reason it works is that all three verbs are competing for the same scarce resource, which is turns.
Your thief is your only real explorer, your only burglar, and the unit that unlocks everything. Every turn she spends walking somewhere is a turn she isn't burgling. Every lantern she steals is a lantern you could spend now on a cheap unit or hoard toward a tier that pays off in six turns you may not get. Antihero never asks you whether you want a thing. It asks what you're willing to be behind on to have it.
The urchin economy is the sharpest piece of design here. Infiltrating a business is cheap, slow-yielding, and permanent-ish — a bet that the game runs long. The buildings are contested, so an infiltration is simultaneously income for you and income denied to them, which means the same action reads differently depending on whether you're ahead. A tempo play and an economy play are frequently the same click, and knowing which one you're actually making is the skill ceiling. New players see a board of buildings and grab the nearest. Good players grab the one their opponent's thief was three tiles from.
Then the violence. Thugs, gangs, saboteurs, assassins — each is a scalpel, not a hammer. This is not a game of stacks and doomstacks; a single assassin removing a single urchin can swing a match, because that urchin was two gold a turn and two gold a turn was the difference between reaching the next tier before the point threshold hit. Combat here is arithmetic with a mask on. It's also, satisfyingly, reversible in cost but not in time. You can rebuy the urchin. You cannot rebuy the four turns it was working.
Where the Tension Comes From
Fog of war does enormous lifting. You know roughly what your opponent is doing — the board tells you — but not when. Every match settles into a bluffing rhythm where you're reading spend patterns rather than positions. Did they just buy an assassin, or did they bank it? Are they racing the point total or setting up a blackmail chain? The information gap is small enough to reason about and large enough to be wrong about. That's the correct amount.
The win condition deserves credit too. Points arrive from too many directions to defend all of them, which prevents the turtling death-spiral that ruins so many two-player strategy games. You cannot sit on a lead. You can only outrun it.
Onboarding and Friction
The campaign teaches well and ends early. It's story-driven, charming, competently written, and over roughly when you've stopped needing it — which is either elegant restraint or thin content depending on how much you paid attention to the "story-driven" part of the marketing. Treat the campaign as an extended tutorial and you'll be delighted; treat it as the main course and you'll feel short-changed. Steam reviewers land on that criticism constantly, and they're right.
Skirmish AI covers several difficulty tiers and is a fine sparring partner at the top end, but it plays the board rather than the opponent. It won't bluff you. Once you've internalized its priorities — and you will, in maybe fifteen matches — the AI becomes practice, not competition. The real game is PvP.
Which is where the structural problem lives. The online population is small. Ranked matchmaking can take a while, and "a while" is fatal for a game whose entire value proposition is I have twenty-five minutes. Async play is the workaround and it's a genuinely good one, but it depends on you having friends who own this game. The single best strategy for owning Antihero is buying a copy for someone else.