The spark is the easy half

People treat the spark as the scarce thing. It is not. A premise is cheap: a city, a curse, a job. The scarce thing is a verb you can do with your hands in the next hour, and a reason to do it again. An amazing idea that cannot be played on Thursday is a wish. A plain idea that can be played, judged, and thrown away is a project.

The filter is almost rude. Who is the player, what do they do most, and what goes wrong when they do it badly? If you cannot answer those without inventing lore, you do not have a game yet. You have a setting looking for one. Settings can wait. The verb cannot, because everything expensive, the art, the story, the content, is a bet that the verb was worth it.

Every game begins as an idea, but finding the spark is the easy half. The harder question is whether that spark can carry months or years of work and still be a game somebody wants to play.

A useful idea does two things at once. It points the creator somewhere, giving them a place to start on mechanics, systems, characters, worlds, and levels. It also gives the player something they can grasp, attempt, and care about. An idea that reads well in a notebook has not yet proved it can become playable.

Seeds come from all over. You can modify a game that already exists, work inside a genre, or start from a single mechanic, a control scheme, an activity, a theme, a fantasy, or something that happened to you. Whichever route you take, the seed then has to survive four questions: can you build it, will anyone notice it, does it have any pull, and is it any fun once it exists outside your head?

The most reliable starting point is a game you already know, with one important thing changed. Satisfactory started from the thought of putting Factorio in first person. The early pitch for Subnautica was Minecraft underwater. Among Us took its shape from social deduction games like Werewolf. Stardew Valley exists in part because Eric Barone felt the spirit of Harvest Moon deserved better than it had been getting.

None of these are reskins. The earlier game supplies a foundation, and then a shift in perspective, theme, medium, structure, or context opens up ground nobody has walked on. Factory logistics feel different when you are standing among the machines. Survival crafting feels different at the bottom of an ocean. Deception plays differently when it is carried out by small animated astronauts calling emergency meetings.

Influence also mutates when it passes through a new person. The Japanese indie developer Ikiki made games that hit designer Jonatan Soderstrom hard. He built his own take on Hakaiman and never put it out. When he later joined forces with musician and artist Dennis Wedin, that prototype ran through a different set of tastes and moods, and it fed into Hotline Miami. You can see the lineage, and the finished game is still its own thing.

The pattern repeats everywhere you look. Infiniminer, Zach Barth's blocky sandbox, shaped Minecraft. Vampire Survivors leaned heavily on the Android game Magic Survival. The battle royale mode in PUBG made the Fortnite version possible. Games talk to the games that came before them.

So the lesson is not to copy without thinking. It is to drop the pretense that ideas arrive from nowhere. Begin with the games you love, the ones that annoy you, and the ones you keep replaying in your head. Then ask what changes if you move the camera, swap the theme, rewrite the goal, redo the controls, or simply let the idea pass through your taste rather than someone else's.

A genre is a template, which is exactly what makes it a good place to start. It hands you a set of mechanics, verbs, expectations, and promises to the player. Roguelike implies runs, randomness, risk, resources, and death. City builder implies placement, growth, tradeoffs, and space. Platformer implies movement, jumps, hazards, and mastery.

One way in is to find something the genre gets wrong and fix it. Ryan Clark felt that traditional roguelikes often played unfairly because so much rested on the roll. That complaint led toward Crypt of the NecroDancer, a rhythm roguelike where your timing counts for as much as the dungeon does. Islanders came out of a similar irritation with city builders: the wish to make something beautiful without the sprawling complexity those games tend to pile up.

Combination is another route. Spelunky came from Derek Yu loving platformers and roguelikes and being frustrated by both. Replaying identical platformer levels got repetitive. Roguelikes could feel arbitrary and opaque. Fusing them produced a game where procedural space, platforming skill, risk, and discovery all reinforce each other.

You can also take something away instead of adding it. The Captain Toad levels in Super Mario 3D World strip the jump out of a platforming world. That single subtraction rewrites the pace and the logic of every level. A gap you would normally leap becomes a slower, more spatial puzzle.

Sometimes the sharpest move is to keep the mechanics and change what they represent. A first-person shooter is largely about aligning a reticle with a target, judging distance, and picking the moment. For most of the medium's life that has meant guns and enemies. Point the same skill through a camera and you get Pokemon Snap. Point it through an industrial pressure washer and you get PowerWash Simulator.

Hold genre loosely, then. Do not let every convention that makes a game qualify as a member of its category dictate your design. Reduce the genre to the few pieces that actually matter, and then bend one, delete one, or re-skin one with something nobody expects.

When games and genres feel like a closed loop, look at the world instead. Ordinary activities make good mechanics because they already come with goals, rules, tools, rituals, and tension built in. Lucas Pope passed through a lot of passport control while moving between Japan and the United States. Watching officials hunt for discrepancies in documents fed into Papers, Please, a game about who gets to cross a border.

Shigeru Miyamoto's gardening fed into Pikmin. Will Wright has talked about ideas surfacing from reading, from study, from bumping into a subject that suddenly looks playable. The point is not to stare at a whiteboard demanding originality from an empty room. It is to keep gathering material until some corner of it starts behaving like a system.

Controls generate ideas too. Luke Muscat was thinking about the ways a finger meets a touchscreen and realized that a swipe felt like a blade. Attach that to the image of chopping fruit and Fruit Ninja is the obvious game for the device. When Nintendo put analog shoulder buttons on the GameCube controller, Yoshiaki Koizumi connected the feel of squeezing one to a water pistol, and that pointed toward Super Mario Sunshine.

A single mechanic lifted from another game can seed an entire project. Braid grew out of rewinding time, which players had already met in Prince of Persia: The Sands of Time. Yacht Club Games loved the downward sword thrust in Zelda 2 and built Shovel Knight around that one motion. A magnetic glove power-up can imply a whole side-scrolling platformer as soon as the designer starts asking what movement, hazards, puzzles, and levels would grow out of it.

This is bottom-up design. Start from rules, verbs, and the things a player physically does, then go looking for the story, theme, and look that suit them. Nintendo works this way often. Splatoon was a mechanic about characters hiding inside blobs of ink long before it was a world of Inklings, fashion, music, and street style.

Top-down design runs the other direction: name the experience you want the player to have, then hunt for mechanics that produce it. Hideo Kojima tends to start from images, moods, characters, music, and concepts while the mechanical form is still open. Death Stranding grew from one arresting mental image into a game about connection, delivery, terrain, isolation, and movement that could break.

Fantasy is a strong top-down source. Which role should the player step into? FTL wanted you to feel like the captain of a ship in a hostile galaxy, holding it together while everything failed around you. Thronefall started from the fantasy of raising and defending a small kingdom, a compressed version of the ruler power fantasy.

Theme works as well. Spiritfarer began with the wish to handle death and dying gently, in a cozy register. Into the Breach came from thinking about superhero stories where whole city blocks get flattened while the monsters brawl. What happens when a tactics game makes that collateral damage the actual subject?

Your own life can be the source. That Dragon, Cancer is autobiographical, built around parents facing their son's death. But a story on its own is no more a finished game idea than a lone mechanic is. Either way, the starting point still has to turn into something the player does.

A helpful way to force that turn is four questions. Inside one level or scene, what counts as winning for the player? What is pushing back and making that hard? What failure are those obstacles driving them toward? And what actions do they take to get past the obstacles and win?

Crazy Taxi answers all four cleanly. Winning means earning enough money by getting passengers where they are going. The obstacles are the clock, the traffic, and everything in the road. Failing means the timer runs out before the money arrives. The actions are fast, reckless driving decisions. That is an idea with enough structure to start building on.

When none of the larger categories catch, smaller drills can get things moving.

One, break games into their smallest parts. Keep a mental periodic table of mechanics and systems so that picking, mixing, and altering them gets easier. Rather than deciding to make a platformer, think in rules and verbs: jump, wall slide, carry, bounce, fall, chase, hide, trade, reveal, rotate, rewind.

Two, play like a critic. What is each game doing well? What grates? What problem does its genre keep stepping around? One game's flaw is another game's opening.

Three, attack a convention. Board game designer Rob Daviau noticed that nearly every board game wipes itself clean between sessions, as though the last one never occurred. Asking what it would mean for one night to change the next led toward Risk Legacy, where players tear up cards and write permanently on the board.

Four, play widely. Staying inside one genre or one medium narrows what you can reach for. Undertale would be a different game if Toby Fox had looked only at JRPGs and never dragged bullet-hell dodging into its enemy attacks.

Five, set yourself a limit. An empty page is intimidating, while a boundary gives your imagination something to push against. After Sayonara Wild Hearts, Simogo decided the next project would be slow, quiet, and black and white. A restriction can harden into an identity.

Six, do game jams, ideally ones with a firm theme, since a theme acts as prompt and constraint together. The Ludum Dare theme Limited Space got Tim Fitzrandolph asking what normally demands a lot of room but might get interesting with less. That became Parking Garage Rally Circuit, a racing game crammed into the ramps of a multistory car park.

Seven, open the editor and mess around. Some ideas only appear while you are making something. A Short Hike started when Adam Robinson-Yu built a small cozy scene in Unity and liked it enough to keep going.

Eight, keep music and art around you. A song, a photo, a painting, a sleeve, or a piece of animation can join ideas that a formal brainstorm never would. Kyle Gabler heard Tango Apasionado and pictured a rainy town at dusk where people hauled chairs and tables out to build an enormous tower. That picture worked its way into World of Goo.

Nine, go do something unrelated. Forcing a brainstorm usually beats out worse ideas than letting your attention drift. Solutions tend to show up in the shower, on a walk, on a commute, or halfway through a chore, because the mind finally has room to link distant things.

Ten, keep it small. Do not try to specify a whole game on paper before anything gets tested. Treat the idea as a seed that grows through prototyping, production, feedback, and revision. Polishing the daydream of a game you may never make is time you do not get back.

With one idea in hand, or ten, the next job is deciding which of them deserves your time. Seizing the most exciting one and committing several years to it without testing anything is possible, but it is a bet rather than a method.

Feasibility comes first. Can you build this with what you actually have? Every developer has a different mix of time, people, money, skills, experience, tools, and patience. The same idea can be a fine project for one team and a catastrophe for another.

Jonas Tyroller offers a rough gauge: if the gameplay prototype takes a day or two, the game may take a year or two. If the prototype alone eats two weeks, the full game is likely to become a very long haul. The arithmetic is not meant to be precise. What matters is that a prototype tells you the real scope faster than optimistic planning ever does.

Scope can also come down without the idea dying. Firewatch was going to have other characters out on the trail. Modeling, texturing, and animating them was beyond a tiny team, so the story moved onto the walkie-talkie instead. The game kept its identity and found the version that fit the people making it.

Feasibility is not purely technical, either. You need the appetite for it. If the project runs for months or years, something has to carry you through the dull stretches: bug fixing, iteration, content, store assets, platform requirements, and the long middle where the novelty has burned off.

Then ask whether anyone will see it. Your game arrives in a sea of other games, and if you only want to make something for the joy of it, that hardly matters. If you intend to sell it, the idea needs some way of catching an eye.

The hook is one lens on this. Ryan Clark describes a hook as a piece of information about a game interesting enough to make people try it or tell someone else. Darkest Dungeon shows the difference. A dungeon crawler is barely a hook at all. A Lovecraftian dungeon crawler about the psychological cost of adventuring is instantly more specific.

A quick test is to write the headline. If a site covered your game, what would the story be called? When the best you can manage is new platformer or new dungeon crawler, the idea probably has no visible shape yet. When the headline pairs a known genre with an unexpected pressure, rule, fantasy, or conflict, the idea has become pitchable.

Hooks can also overshoot. Chris Zukowski warns that a game can be original enough that nobody can tell what it is. If the pitch sounds risky, muddled, or hard to place, players may admire it and keep scrolling because they cannot predict what they would be buying. Arco mixed a Mesoamerican fantasy RPG with real-time turn-based combat that verged on bullet hell. Critics and players liked it, and the unusual pitch still made it a harder sell.

That is where the anchor comes in. The hook is the unfamiliar part; the anchor is the familiar part that lets a player work out how the game functions. A known genre, camera, control scheme, structure, or goal can make the surprising element feel safe to try.

Zachary Richman puts the formula plainly: simple, with something unexpected. A tower defense game where your own bullets can hurt you is easy to picture and hard to forget. Simple gets people through the door. Unexpected keeps them there.

A hook is one form of pull, not the only one. Jonas Tyroller prefers to talk about appeal, meaning everything that draws a person in before they have played a second of it. Appeal can come from a screenshot, a clip, a line of description, a fantasy, a toy-like interaction, a challenge, a joke, a memory, or the promise of a place worth walking through.

Fantasy appeal makes someone think, I want to be that. Exploration appeal makes them think, I want to go there. Toy appeal makes them think, I want to poke that and find out what it does. Other games run on challenge, comedy, horror, coziness, or nostalgia.

This is the argument for thinking about marketing earlier than feels natural. You do not need final store art before you write a line of code, but you should be able to picture how the idea would be shown. Lucas Pope has described asking whether a concept could be turned into a trailer. When he cannot imagine a good one, he treats that as a warning.

Tom Francis has made a related point about names and capsule art. Working out the title, the capsule, and the first public image can tell you whether the idea has a proposition worth selling before you sink real resources into it. Neither exercise is only marketing. Both are design tests, asking whether the idea has a visible center at all.

Not every game reduces to one perfect sentence, and plenty of successful ones resist an elevator pitch entirely. Even so, the creator should be able to name the kind of appeal the game carries and describe how it could reach somebody who has never touched it.

The last question is whether it is fun, and this is where imagination is least trustworthy, because people are terrible at simulating games in their heads. Almost anything sounds fun while it stays abstract. The problems only surface once a player has to move, choose, wait, aim, fail, retry, and understand.

A prototype converts the daydream into evidence. It can show you that the idea is dull, overcomplicated, sluggish, illegible, or expensive in ways nobody spotted on paper. It can also show the reverse: one small interaction that feels far better than expected and deserves to sit at the center of the whole game.

The prototype does not have to justify the entire project. It has to test the one thing that makes the idea worth building. If the pitch rests on a strange movement rule, build the movement. If it rests on psychological strain in a dungeon, build one decision under that strain. If it rests on a small kingdom, build the loop of raising and defending it. An idea that only becomes fun after twenty hours of content is not ready to be an idea yet. Get a build running, play it yourself, and let your own judgment about what you just played settle the question.

A good game idea, then, is more than a premise, a story, a mechanic, or a genre mash-up. It is a playable direction. It has a goal, an obstacle, a fail state, and actions. It has a hook, or at minimum a form of appeal. It is anchored enough to read and small enough to test. Above all, it holds up when somebody plays it.

That is the whole route from inspiration to something worth building. Take influence without embarrassment. Run it through your own point of view. Look outside games for mechanics and fantasies. Boil the result down to a playable structure, check that it can be noticed, and build the smallest version that can tell you the truth.

In the Flockbay app, refuse the premise for the first sentence. Describe only the verb and the failure. Approve a brief for one room. Play it. If you are bored, the idea was the setting. If you want another room, you have something.

An AI game idea generator will offer premises. Use it to throw them away. A start-here template is the room.