1666: Amsterdam Turns Witchcraft and Possession Into a Dense Stealth Sandbox

1666: Amsterdam is a supernatural stealth game built around a premise that is immediately stranger, and more promising, than a conventional historical adventure. Set amid witch hunts in 17th-century Amsterdam, it casts players as Noa, a witch who has one calendar year to hunt figures known as the Originals. These are humans bound to demonic forces, and finding them is less a matter of following a highlighted trail than patiently working through neighborhoods packed with homes, businesses, institutions, alleys, canals, secrets, and suspicious guards.

The game comes from the creator of Assassin's Creed, but its most distinctive qualities lie less in rooftop action or open-world sightseeing than in its tightly constructed infiltration spaces. Its Amsterdam is designed as a dense, interconnected place where individual buildings matter. A lead may only provide a name and a broad part of the city, leaving players to investigate on their own terms. That approach makes the early missions feel closer to a classic social-stealth sandbox: observe routines, find a way inside, create an opening, and use whatever identity or route best gets Noa closer to the truth.

That foundation is especially compelling because Noa is not built to simply overpower every obstacle. She can use magic, but much of the game's tension comes from avoiding the consequences of looking like a witch in a city hostile to them. Her conspicuous red robes can draw unwanted attention, so a quieter outfit is often safer. Even seemingly minor acts, such as lingering at an open window or entering an area where she does not belong, can raise suspicion. Guards respond quickly, and a situation can escalate to the point that Noa has to flee.

When that happens, the game does offer a more dramatic escape. Noa can take to a broom and speed along Amsterdam's canals to lose pursuers. Yet the more memorable solutions come from stealth, observation, and a highly unusual companion.

For useful background on this topic, read Hideo Kojima Says He Is 'Envious' of Backrooms Director Kane Parsons' Early Film Breakthrough.

A cat familiar changes the rules of infiltration

Noa's familiar is Aaron, a cat inhabited by the spirit of a confused historian from 1999. The explanation behind that arrangement is intentionally part of the game's layered narrative mystery, but the immediate gameplay impact is clear: Aaron is central to nearly every interesting break-in.

Players can directly control Aaron, allowing the cat to climb, jump, squeeze through narrow gaps, steal a single item, and investigate areas Noa cannot safely reach. He can also distract people with a meow, a small but useful tool when timing matters. More importantly, Aaron can access the interiors of buildings and unlatch doors from the other side. A locked entrance that stops Noa can become an invitation once the cat finds a way through a window, over a wall, or along an overlooked route.

This feline perspective makes Amsterdam's environment feel more tactile. Rather than treating a house as a sealed mission prop, the game encourages players to look at its windows, back rooms, side passages, and internal layout as a set of possibilities. A cat can turn a suspicious-looking entry attempt into a clean infiltration route. That alone would make Aaron valuable, but his most consequential ability is possession.

Possession is the game's most exciting stealth tool

Aaron can scratch certain frightened characters and allow his spirit to enter them. Once perched on a host's shoulder, he can take control of that person and make them carry out Noa's wishes. The result is a flexible body-hopping system that changes infiltration from a question of "Can Noa get through that door?" to "Who can get Noa through that door without raising alarms?"

A servant, for example, can be used to move through the spaces where they naturally belong, open interior doors, and eventually unlock the front entrance for Noa. The host can then accompany her inside, helping her avoid attention from guards who might otherwise question her presence. A baker can provide access to supplies in a storeroom. A scholar encountered in the city can create an entry point into university laboratories. A scoundrel with lockpicks can be made to handle barriers that Noa cannot. Even guards can become assets, with the potential to use their own truncheons against fellow guards.

That is the heart of what makes 1666: Amsterdam stand apart. Possession is not a limited scripted gimmick reserved for a handful of set pieces. It is presented as an ongoing tool for social manipulation, exploration, and problem-solving. The player is encouraged to look for the right person rather than merely the right key. Each potential host represents a different relationship to the city's rules, giving Noa indirect access to spaces that would be dangerous or inaccessible on her own.

The system evokes the body-switching ideas of Omikron: The Nomad Soul, but applied to a more open-ended stealth structure. It can lead to elaborate, player-driven scenarios: Aaron slips into a building, possesses someone inside, opens it from within, and guides Noa through an environment that had initially seemed off limits. The game's best moments appear to emerge from this kind of improvisation rather than from a prescribed sequence of objectives.

Clues, restricted spaces, and an Amsterdam worth learning

Noa's hunt for the Originals depends on gathering information. Public notices and overheard conversations at inns can provide useful direction, but they are not the complete solution. The decisive clues needed to identify a target and the demon connected to them are tucked away in restricted places. That means a successful investigation requires more than listening to gossip; it requires breaking into the places where people hide records, possessions, and uncomfortable truths.

The early-access content includes a first map spanning two primary missions and three targets, with more linear missions set in 1999 appearing between the historical chapters. The 1999 material follows Clio, a present-day scholar interpreting texts left by her ancestors, Noa and Aaron. That framing gives the story an unusual, multi-period structure, combining the witch-hunted Amsterdam of 1666 with an investigation taking place centuries later.

Even after roughly ten hours with the available material, there is still plenty of city left unexplored around Noa's safe haven. That sense of unfinished personal knowledge is important to the game's appeal. The streets are not merely scenery between objectives. They are spaces players can gradually understand: which doors open, who belongs where, where a cat can slip through, and which person might be useful for gaining access to a particular building.

Crafting also supports this investigative loop. Searching rooms can yield materials and potion recipes. One discovered potion can allow Noa to move through restricted zones without disturbance for a short time, offering another answer when there is no convenient person to possess. Hidden coffers left by Noa's coven can provide upgrades as well, rewarding thorough exploration beyond the immediate search for a target.

Magic upgrades add uncertainty, while combat needs work

Noa's progression uses a roguelike-style system that offers a choice of three upgrades. This gives each build an element of chance and flexibility, while upgrades that have not been permanently locked in can be rerolled after death. Some abilities are easier to understand at a glance than others. Options include hiding in bushes and pulling guards into confined spaces, while other unlocks can be less clear until they are tested in practice.

That partial uncertainty fits a game about learning occult abilities through experimentation, though it can also make the early-access experience feel underexplained. 1666: Amsterdam does not always do enough to teach the full range of tools available for slipping into forbidden spaces. Players willing to experiment may find that opacity intriguing, but clearer guidance could help ensure that the game's best systems are not missed.

The weaker side of the current build is its boss combat. Investigations are periodically broken up by fights, but these encounters have been described as generic, with uninteresting attack patterns. Noa can attack using her broom, and the basic combat is functional, but it does not carry the same creativity as the possession and infiltration mechanics. In a game whose identity is rooted in fragile, deliberate stealth, straightforward boss battles feel like a mismatch. This is an area where Early Access leaves ample room for revision.

An intriguing Early Access experiment with rough edges

1666: Amsterdam is not being presented as a polished finished package. Its Early Access state shows in the uneven tutorials, the need for a stronger combat approach, and elements of production that have drawn criticism. The use of AI in the game's production has been a particular point of concern, though that use has subsequently been scaled back to some degree.

Still, the game's roughness does not erase what is unusual about it. Its central ideas are strong because they create stories players can tell through their own actions. A quick search for a supernatural suspect can become an extended detour involving a cat, a frightened servant, an unlocked door, a stolen identity, a carefully timed escort, and a cupboard full of potion ingredients. That chain of decisions is more memorable than a checklist of map markers because it grows from the player's interpretation of the world.

For now, 1666: Amsterdam's greatest spell is its willingness to let players be nosy, vulnerable, and inventive. The city's dense streets reward patience. Aaron's abilities encourage players to think through spaces from a new angle. Possession turns ordinary residents into possible pathways. If its combat and onboarding improve during development, the game could become a distinctive stealth experience precisely because it makes investigation feel messy, supernatural, and deeply personal.