Knights in Tight Spaces

The original Fights in Tight Spaces had one of the cleanest hooks in the deckbuilder boom: a lone secret agent in a corridor, a hallway, an elevator, and every card played was a punch that moved a body. It borrowed Slay the Spire‘s economy, borrowed John Wick’s choreography, and welded them together with a grid. It worked because the fantasy was singular. One man. One deck. One problem at a time.

Knights in Tight Spaces takes that machine apart and rebuilds it as a party tactics roguelike. Same universe, same studio — Ground Shatter, once again published by Raw Fury — but now the trench coat is chainmail and the elevator is a medieval antechamber. You recruit up to three adventurers from a roster of archetypes: knights, archers, sorcerers, rogues. Each contributes cards to a shared hand, which is the central design gamble of the entire project. You no longer manage a character. You manage a committee.

Commercially, this lands in the most crowded corner of PC gaming. The tactical deckbuilder space now includes Slay the Spire II‘s long shadow, Wildfrost, Balatro‘s gravitational pull on the entire roguelike conversation, and a dozen grid-tactics hybrids. Ground Shatter’s differentiator was never card design — it was physics. Shoving a bandit into a brazier is a verb no other deckbuilder owns. Knights leans harder into that, and when it sings, it sings.

The problem is that the sequel grew in the wrong direction. The party system buries the original’s clarity under deck clutter and class-stacking math, and the friction shows up exactly where the game asks you to spend the most time.

It is a Windows PC release on Steam — no eShop, iOS, or Android version. That’s a notable retreat from the original’s multi-platform reach, and it matters more than it sounds. More on that below.

• Shared-Hand Party Combat: Up to three heroes draw into one hand. Any card you hold belongs to a specific character, so the tactical question is “who is even in position to play it.”
• Environmental Kill Physics: Walls, spikes, braziers, and other bodies are all weapons. Knockback chains — shoving A into B, sending B into a hazard — remain the most satisfying thing in the game and the clearest expression of its design identity.
• Team-Up Attacks: Position characters correctly and they combo off one another, letting a rogue finish what a knight started. This is the mechanic that justifies the party system, when it fires.
• Large Card Pool: A substantially larger upgrade and synergy space than the original, spread across class identities you shape through recruitment, shops, and events.
• Branching Run Structure: Node-map traversal through fights, shops, and events, punctuated by bosses, with permadeath restarts.
• Long-Tail Modes: Difficulty modifiers, a Daily Play leaderboard, and an Endless mode for players who want the run to keep going.

The Gameplay Loop

Combat happens on small gridded arenas — typically enough room for a dozen tiles of meaningful movement and not one more. You draw a hand, spend momentum, move, strike, and end turn. Enemies telegraph their intent. Everything is legible before you commit. That legibility is the game’s finest quality, and it deserves credit: at any moment you can see exactly what is about to happen to you, which turns every loss into your loss rather than the RNG’s.

The core tension is position over damage. Raw attack cards are the weakest tool in your kit. The strong play is almost always a shove that arranges the board — pushing an axeman into a spike pit for far more damage than your sword would have done, or lining up three enemies so a single knockback bowls them into each other like a shopping-cart pileup. When a multi-body chain reaction clears a room without direct damage, Knights delivers something no other card game does. That is a real accomplishment, and it’s why the game is worth playing at all.

The Three-Deck Problem

Here’s where it gets complicated, and here’s where the design falters.

In the original, your hand was a set of options. In Knights, your hand is a set of options attached to people who may or may not be where those options are useful. You hold a devastating sorcerer card while your sorcerer is pinned in the wrong corner. You draw two knight cards, one archer card, and a rogue card, and the archer is the only one with a viable line. The hand stops reading as a toolkit and starts reading as a logistics puzzle — not “what’s my best play” but “which of my three employees can actually do their job this turn.”

This is not automatically bad design. Managing dead cards is a legitimate tactical burden. The problem is that the card economy wasn’t rebalanced to absorb the added variance. Draw pools repeat basic attacks too frequently, so a hand can easily be three vanilla strikes belonging to two characters who are out of range. In Fights, a dud hand meant a defensive turn. Here it means a wasted turn for a party of three, which is triple the tempo loss.

The obvious workaround is telling: stack two or three of the same class. Run three knights. Run double sorcerer. When your party shares an identity, the shared hand stops fighting itself — cards become more or less interchangeable across bodies, synergies concentrate, and the whole system clicks. That the optimal path in a game selling you a varied adventuring band is homogeneous parties is a design failure worth naming plainly. The fantasy is a mixed party. The optimal play is a monoculture.

Combat scale compounds it. These are tight spaces, per the title. In cramped rooms, melee dominates. Archers want range they rarely get; sorcerers want setup time the grid doesn’t grant. So the class balance quietly funnels you toward the same answer the class-stacking math already suggested: bring knights.

Run Structure and RNG

The node map does what node maps do — competent, familiar, unremarkable. Shops and events give you real build decisions, and the large card pool means late runs can produce distinct engines. Recruitment is the best structural idea here: picking up a new companion mid-run reshapes your whole hand, which is a bigger decision than any single card draft in the genre.

The weak point is spawn RNG on certain missions. Some encounters place enemies in configurations that punish you before your first turn, and in a game where positioning is everything, a bad spawn isn’t a challenge — it’s a tax. Difficulty modifiers help, and shipping meaningful sliders rather than a single punishing default respects players. But sliders are a workaround for encounter design that occasionally doesn’t respect its own rules.

The art direction is the least controversial thing here, and rightly so. Stylized silhouettes, flat bold color, exaggerated impact effects that punch well above what a turn-based game needs — every kill lands like a stunt in an action movie. The electronic score is driving and unfashionably confident for a fantasy setting, and it works precisely because it refuses to be lute music. Ground Shatter understood that the appeal of Fights was kinetic presentation on a static system, and they didn’t dilute it for the medieval reskin.

A 2D game of this scope is light on hardware, and the modest rendering budget should translate well to handheld-class machines.

The UI is the weak link. The End Turn button is small, and the indication of which hero is currently active is unclear. Those sound like nitpicks. They are not. In a game where the entire skill expression is knowing whose card you’re holding and where that person stands, an ambiguous active-character indicator attacks the core loop directly. Adding two more characters tripled the interface’s job, and the interface didn’t get three times the design attention. The information display itself is otherwise excellent — enemy intent, damage previews, hazard telegraphing are all clear. It’s the character-attribution layer that’s underbuilt.

Steam, Windows only — and that’s a step backward from the original’s reach. Fights in Tight Spaces found a real audience on Switch and mobile, where a turn-based grid game with no reflex demands is close to an ideal fit. There is no eShop, iOS, or Android release, which leaves the “play a run on the couch or the train” use case entirely to Steam Deck.

On Deck, the pick-up-and-put-down run structure suits short sessions, and a turn-based game of this scope is a natural fit for handheld play. Mouse and keyboard remains the precision option — drag-targeting a shove across a crowded grid is faster with a cursor, and the small UI targets that strain on a handheld are trivial with a mouse. If you’re buying this to play in bursts, Deck. If you’re buying it to optimize a Daily Play leaderboard run, desktop.

Knights in Tight Spaces is a sequel that mistook more for better. Every individual addition is defensible — a party adds strategic texture, team-ups add expression, a bigger card pool adds depth. Together they crowd out the thing that made the original special: the total clarity of one person solving one room.

And yet. The knockback physics remain unmatched, the presentation is thrilling, and once you accept the game’s real rules — stack your classes, favor melee, treat the shared hand as a positioning puzzle rather than a toolbox — there’s a lot of sharp tactical play here. At $19.99, though, the friction is priced in, and we can’t recommend it for most people.

We recommend it only to players who want the best environmental-destruction tactics on Steam and don’t mind fighting the interface for it. Everyone else should wait for a sale — or for the original’s elegance, which this sequel never quite recaptures. Either way, this is a studio doing something nobody else is doing, executing it well short of its potential — which is still more interesting than most of the genre firing on all cylinders.