The save was the design choice

Early in Alien: Isolation, Creative Assembly, 2014, the save was going to be the ordinary kind: invisible checkpoints, progress kept as you crossed a line. Cheap, legible, and fatal to the fear, because a checkpoint is a promise that the last minute did not count. The team threw that promise out. You save at a machine, in the open, where the creature can find you.

That is what thinking like a designer is. Not a talent for ideas. A habit of asking what a borrowed mechanic is actually promising, and whether this game can afford the promise. A checkpoint, a mini-map, a regenerating bar, a companion who talks: each one arrived from some other game, carrying that game priorities. Import it whole and you may import a game you did not mean to make.

Take saving. Early in development, Alien: Isolation saved the way most games do: invisible checkpoints scattered through the level, progress cached automatically as the player crossed them. Cheap to build, instantly legible to anyone who had played anything else.

The longer the team worked, the more wrong that familiarity looked. Death cost a couple of minutes, players knew it, and they walked the station like people with nothing much to lose. In a military shooter that is fine. In a game built on terror and vulnerability it quietly cancels the point.

None of which is an argument against borrowing. Designers borrow constantly, then evolve, combine, and remix what they take. The failure mode is narrower: lifting a mechanic without asking why it worked where it came from, what it will make your players do, and whether doing that helps your game.

The MDA framework splits a game into three connected layers - mechanics, dynamics, and aesthetics - and that split is what makes the question above answerable.

Mechanics are the rules and systems themselves. What each button does, how much ammo fits in the player's pockets, the numbers behind movement, how saving works, how enemies decide what to do. Whatever the code allows.

Dynamics are what players actually do once those rules are in front of them. Hand them ammo without limit and they will charge in and shoot at everything that moves. Make ammo scarce and the same players start dodging fights, lining up shots, picking over rooms for supplies, thinking before they fire.

Aesthetics here has nothing to do with graphics. It means the emotion the player ends up with. The full pockets read as reckless, powerful, unstoppable. The empty ones read as frightened, careful, small.

Put briefly: the mechanics sit in the code, the dynamics happen in what the player does, and the aesthetics happen in what the player feels.

You cannot reach into a player's head and install an emotion. You can change a rule. Rules shape behavior, behavior produces emotion, so a change made in the code travels down the chain and comes out the other end as a feeling.

That chain is what makes MDA good for analysis. "Is this mechanic fun?" is a dead end. "What does this mechanic make players do, and what does doing that make them feel?" is a question with an answer.

Weapon durability in The Legend of Zelda: Breath of the Wild is the standard case. Ask why swords break and the honest answer is behavioral. Players stop leaning on the frontal attack. They sneak, they set up ambushes, they cycle through a bag of odd weapons, they keep scanning the world for the next one. A blade snapping mid-fight is its own small piece of drama.

Those behaviors add up to a specific emotional posture: underpowered, resourceful, inventive, always out looking, in a world that is ancient and broken and dangerous. Durability is not really a meter. It is the thing that puts the player in that relationship with the world.

The difficult part is vocabulary. These feelings usually sit below the level a player can articulate, and "this is fun" is far too coarse to design against. Reach for specific words instead: powerful, creative, sneaky, tense, intimidated, curious, deceitful, cooperative, flustered, relaxed, afraid.

With the word chosen, a mechanic becomes testable. It earns its place when the behavior and emotion it generates run alongside the experience you are after. It falls under suspicion the moment it starts producing something the rest of the game is working against.

Flower shows this cleanly. It once carried a level-up system, spells, resource management, and time limits - all perfectly ordinary video game furniture. Every one of them pulled against the calm and peace the game existed to deliver, so all of them went. Familiar and functional is not the same as emotionally right for this game.

A vision is the overarching feeling, fantasy, or experience you intend the player to come away with, and its practical value is that it turns every later mechanic into a yes-or-no question.

It can be a phrase. Subnautica was built on the "thrill of the unknown", which meant its rules had to protect curiosity, danger, and discovery. Resident Evil Village worked from "struggle to survive", and that idea gave the team something to hold combat up against when playtesters said it felt messy.

It can equally be a fantasy: be Batman, be an assassin, run a country, command a starship, be lost somewhere that wants you dead. FTL started from wanting to reproduce what it feels like to captain a ship. Once the fantasy is stated, choosing mechanics becomes a matter of which ones deliver it.

Mechanics are not the only things being judged this way. Art style, music, animation, story, color, and where the camera sits all generate emotion too. Dead Space was first given a conventional sci-fi score, and it made players feel heroic when the game needed them scared. The music changed because the emotion was wrong.

Games read as sharp and confident when their parts all lean in one direction. DOOM in 2016 is the demonstration. The vision was "push forward combat", and it is difficult to find a piece of that game not obeying it.

The rules shove you at the enemy: you move fast, you recover health by getting close and aggressive, and the enemies themselves behave in ways that reward closing distance. Everything outside the rules says the same thing - the weight of the music, the brutality of the animations, a protagonist framed as something the demons should be running from.

The strength of that is alignment. Mechanics, dynamics, aesthetics, presentation, audio, and fiction all serve one emotional goal, so the player is never being told to act one way while the game insists they feel another.

In practice this is far less tidy than the framework suggests. No game is one mechanic. It is hundreds or thousands of them, most of them small, all overlapping, interacting in ways nobody sat down and planned.

A platforming character alone might be described by dozens of numbers: acceleration, jump height, gravity, coyote time, air control, terminal velocity, friction, and on it goes. Nudge a few of them and the character travels from sluggish and fragile to quick and expressive.

Mechanics can also work against one another. The Callisto Protocol restricts ammo so the player feels outmatched and afraid, then offers a smooth one-button stealth kill that makes them feel like a predator. Each rule is defensible alone. Side by side they transmit two different emotions.

Which is why the real answer comes from putting the build in front of players and watching. Educated guesses are worth making, but players will ignore a system, misread it, exploit it, or turn up something nobody anticipated. Occasionally that discovery is a gift - the flying cars in Rocket League still serve the game's vision. Other times it hardens into a degenerate strategy and takes the intended feeling with it.

One emotion for the whole running time is not a requirement. The feeling you want often shifts as the story or the structure moves. A character who starts out weak and green and finishes capable and dangerous should have mechanics that track that arc.

Practically, that can mean rewriting rules, resources, abilities, enemy pressure, or the cost of failure as the game proceeds. The question you ask does not change, only where you ask it: what is this mechanic making the player do right here, and what does doing that feel like at this point?

One more wrinkle: emotion is subjective. A scoring system reads to one player as a challenge worth another run, and to the next as a verdict they did not ask for. A timer that thrills one person makes another anxious enough to quit.

Skill filters all of it. A rhythm-action game may be built to make you feel like a rock star, but a player with unreliable timing will feel clumsy instead. The mechanic still produces the behavior it was designed to produce - the emotion on the far side is filtered through that player's body, ability, expectations, and mood.

This is not a reason to give up on designing for feeling. It is a reason to be precise about your intent, to watch real people play, and to hold your conclusions loosely. Rules are emotional tools, but they are not the same tool in every pair of hands.

The checkpoints went. What replaced them looked more like older survival horror: a handful of manual save stations that the player has to locate and use. Saving is not instant, and while the machine works the alien is still out there and can still kill you.

The dynamics turned over immediately. Players started hunting for stations, plotting routes between them, weighing whether the trip was worth it, and sitting exposed and helpless through the wait. Saving your game became one of the most frightening things in it.

And that fear is exactly the terror and isolation the whole game was built to produce. The mechanic was not picked because it was fashionable or because other games did it. It was picked because of how it made people behave, and how behaving that way made them feel.

Borrow freely. Just never stop at "does this mechanic work?" The questions that matter are what behavior it produces, what emotion follows from that behavior, and whether that emotion belongs in the game you are making.

MDA earns its keep by giving you a path to walk: rule, then behavior, then feeling. Where the feeling at the end complements your vision, the mechanic has a case. Where it fights the vision, the mechanic needs changing or cutting, however well it performs.

So there is no universal verdict on any mechanic. Whether it belongs depends on your game, the emotions you are chasing, and the experience you want your players to walk away with. The way to find out is to get it running and play it yourself.

In the Flockbay app, build a short stretch of horror with automatic saves, play it, then move the save onto an object the player has to choose to use. Play again. Keep the version where you hesitated. Hesitation is the information.

A stealth horror template and the AI horror game maker are the right fear. One corridor and one save.