Killer Bean was released in early access, but the developer promises to support their project for at least two more years. The project already has a core gameplay loop, but it repeats every subsequent level. This raises the main question: what exactly are they going to do with the project for two more years? After all, we already have a roguelike shooter with procedural missions, random weapons, factions, vehicles, and primitive leveling.
John Wick at home
Killer Bean is built on a simple idea: a coffee bean destroys other beans, moves rapidly, shoots in slow motion, uses finishers, and participates in a ridiculously serious criminal story. In the first few minutes, it looks amusing.
But such a character is better suited for a short arcade action game. Killer Bean could work as a set of quick shootouts with a simple task: enter a level, beautifully clear a room, move on. But the game tries to be more. It wants to combine roguelike, open world, and procedurally generated story.
The combat system is the most understandable part. The hero shoots with two hands, jumps, dodges, uses slow motion, masters melee techniques, and special abilities. In short skirmishes, it works. The game gives the feeling of a chaotic shootout where the player constantly moves and shoots enemies on the go.
However, after a few hours, it becomes clear that the gameplay is primitive. Most encounters are built on the same scheme: the player goes to a point, captures an object, kills enemies, gets a new target, and moves on. Sometimes bosses, vehicles, or altered mission conditions appear, but the overall rhythm remains the same.
Slow motion quickly ceases to be an effective technique. It becomes a mandatory button because the game often throws too many enemies at the player. Melee finishers and special attacks look amusing but are not very useful. And although the guns on the levels offer bonus effects, the runs themselves do not become more diverse.
This is a serious problem for a roguelike. The genre relies on the desire to start a new run because next time the game might play out differently. In Killer Bean, a new run more often promises the same set of actions in a different order.
The battles themselves take place with large groups of opponents. Enemies often appear in waves and quickly surround the player. This should create pressure, but more often it highlights the weakness of artificial intelligence.
Opponents poorly read the situation, react strangely to the player, get stuck, or simply wait for death. In such conditions, skirmishes rarely become tactical. The player does not adapt to smart enemies but clears another group of identical targets in slow motion.
There is also a problem with visual readability. Enemies, civilian characters, and other beans often look too similar. In a chaotic scene, it is not always immediately clear who poses a threat and who should be spared.
Bean Roguelike
Killer Bean uses procedural generation of the world, missions, and events. In theory, this should make each playthrough unique. Allies can become enemies, objectives change, and islands are reassembled.
In practice, procedural generation does not eliminate the routine from the gameplay. Missions boil down to monotonous tasks, the world remains empty, and the story turns into a set of excuses for shootouts.
Because of this, the roguelike orientation seems erroneous. It prevents the game from becoming a dense story-driven action game, and the developer himself seems not to understand the meaning of the genre. After all, even leveling up does not give a noticeable sense of progression, and random weapons do not change the playstyle.
The game features open islands with cities, bases, and warring factions. For a one-developer project, this looks impressive, but in Killer Bean, scale often works against the concept.
Locations are empty and are mainly needed for moving between tasks. Exploring them is uninteresting, and biomes, although visually different, rarely change the approach to gameplay.
Vehicles speed up movement but do not make trips pleasant — cars and boats perform a purely functional role. At the same time, they are presented only in two missions and in the segment with purchasing upgrades. As a result, the open world is perceived as an unnecessary layer between action segments.
The Charms of "Early Access"
Killer Bean has enough technical problems. There are bugs with missions, animations, interface, vehicles, and enemy behavior. But the technical state does not explain the game's main weakness. Bugs can be removed, controls can be improved, artificial intelligence can be refined.
The structure itself is tiring. The main loop is already clear, the roguelike formula is already revealed, and the open world is unlikely to be grandly reworked. If the game starts to repeat itself after a few hours, mere fixes and patches will not be enough.
The developer announces long-term development in early access. Plans include new features, co-op, customization, mod support, and other updates. All this can make the game more stable and larger.
But it no longer looks like a draft that can be easily rebuilt — it's an almost assembled game that lacks polish, content, and a focus on mechanics rather than scale.
Verdict
Killer Bean is a paradox game.
It has a fun foundation, interesting combat mechanics, and an attempt to provide fun for fans of the source material. But this is not enough for a big open-world roguelike, which they inexplicably decided to cram in here.
After a few hours, the game becomes repetitive, sending the player again and again on missions that are as alike as two drops of water, capturing points and mowing down crowds of enemies.
If early access comes down to bug fixes, new modes, and expanding the same formula, rather than refining fundamental mechanics — we will lose a potentially fun project in a silly universe. And, judging by the fact that we already have an almost finished product — that's exactly what will happen.