Video games are a vast medium enjoyed by all sorts of people. As this amazing medium expands its gameplay, literary, visual, and audio capabilities, the diversity and quality of video games will continue to improve. While many games have succeeded without stories in the past, the demand for quality storytelling in video games has been high since 1997. So, how do you tell a good story in a video game? First of all, a story is a story no matter the medium, so all the advice for writing novels, movies, etc. applies to writing video game stories as well. The idea of this page in particular, however, is to help you find the tropes and techniques for telling a good story in a video game (specifically). Most notably, it covers how to fully utilize the medium's interactive nature when telling a story.

!'''Necessary Tropes'''
Video games are a business. What this means is that, unless you can finish making a video game all by yourself (or have enough helpful pals willing to work for free), you'll need to frame your story in such a way that a businessman would want to invest in it. As far as stories go, this usually means adhering to what this troper calls business 101 - "copy that other product that made a bunch of money!" Hence, knowing VideoGameTropes inside and out will help get your video game story that much closer to publication. It's also worth noting that the VisualNovel medium shares many of these tropes, because the visual novel and video game industries have influenced one another over the years.

!'''Choices, Choices'''
As said above, all choices depend on the game's genre to some degree. For now, let's discuss VideoGameTropes that apply to all (or most) genres.
* Difficulty tropes are one of the easiest to decide. Do you want to design a game to be NintendoHard? Do you want the game to have an EasierThanEasy mode? Do you want to establish EasyModeMockery? This troper's advice is to consider the consequences of the player losing at this time as well (frequent save/check points, unlimited lives, etc.) as well as the target audience (easy games are usually designed for kids, and hard games are usually designed for adults).
** Note, additionally, that there are ''different kinds of difficulty''. Players might have difficulty grasping the overall picture--"What's that MostAnnoyingSound mean, and why can't I ignore it?" They might have difficulty grasping the particular nuances of ThatOneRule, or be overwhelmed by LoadsAndLoadsOfRules. They might have trouble with the ''physical motions'' of using the controller (SomeDexterityRequired). When designing, keep in mind which of these flavors of difficulty you happen to be good at, and make sure to get a second opinion on the difficulty level you've created.
* Do you want your game to be a {{Deconstruction}}? Do you want it to be meta? Even though such ideas are specific to the genre in which you want to use them, designing a game around it is best done as an idea from the start because making a game like that requires both a careful selection of tropes and intricate application of said tropes. For example, do you want to use GameplayAndStorySegregation or try to integrate the two? Do you want to use MissionControlIsOffItsMeds as a vehicle for the game's meta value? Do you want to use WhatTheHellHero and DesignatedHero to deconstruct the player's intuition? Does making the world a CrapsackWorld appeal to some of the meta concepts for which you're aiming (thus allowing you to show the effect of gameplay conventions as if they occurred in real life)?
** The meta game: These are games whose entertainment lays not in a unique world, but their relation to other video games. They do not tear apart ideas, they show how ridiculous they are. They can be tongue in cheek (VideoGame/CthulhuSavesTheWorld, VideoGame/TeamFortress2 has this going on with making the story fit the gameplay to a very odd degree), or they can be serious (VideoGame/BioShock, VideoGame/MetalGearSolid2, VideoGame/SpecOpsTheLine is so far in this end it's embedded in the wall), but they can be in between (VideoGame/{{Borderlands}}). What these games have, so far as meta value goes, is that they call attention to and sometimes even play with tropes you're expected to find in games. They can be as simple as a LampshadeHanging (VideoGame/CthulhuSavesTheWorld is very fond of this approach to {{JRPG}}s), or they can be important ([[spoiler:[[VideoGame/BioShock1 A man chooses, A slave obeys!]]]]) to the overall story. Done right, the game becomes a big hit because of how it makes the player think about how conditioned they are about the games or how they see everything in videogames.
** The deconstruction: Although meta-games do overlap with this, it's different. As with any medium, videogames have set traditions, and as with any medium they can sometimes be seen as negative. The best of these games are often meta because they have to draw you into the world and shred your fantasies before you. There is a blurred line at the end of serious meta games and deconstructions. ''VideoGame/EpicBattleFantasy 4'' is on the lighter end, deconstructing kleptomaniacs running around [[KleptomaniacHero who happen to care about the world and want to stop it being destroyed]]. Then you have VideoGame/SpecOpsTheLine, where the game is recommended on the basis that it shakes up the concept of FPS in a way that you end up disliking yourself. As far as popularity, they would ride mostly on the wave of meta-games to be popular, as for a design choice it's easier to make than a meta, although do not expect it to be easy still. The best point about these is that it can help base itself into a game world, and allow you to make a deeper story that is more unique to the video game genera on the merit of what you can do with it.

You should do some thinking about what platform you want your game to be on. The UsefulNotes/PCVsConsole argument has been going on for ages, partially because BothSidesHaveAPoint.
* Computer Games are played on a computer. Computers typically have some of the most powerful hardware available, capable of running VideoGame/{{Crysis}} at full settings... but they also have some of the worst, since you basically can't control what kind of hardware is in your consumer's tower. They come with a mouse and keyboard, which is a good interface for strategy-oriented games where you need a lot of things to be available at the press of a button.
* Video Games are played on a console. Consoles are easier to program for because the hardware is standardized: every UsefulNotes/PlayStation4 has the exact same things inside it as any other (with the sole exception of hard drive space). You know exactly what the console can do. However, this requires a fair bit more in terms of licensing fees, and a bit more bureaucracy to wade through, since most console manufacturers want to do at least a little bit of Quality Assurance before they let the game released on their machines.
* {{Mobile Phone Game}}s are played on cell phones, particularly smartphones these days--AndroidGames and {{iOS Games}} are proliferate. They benefit from extreme portability, as well as the (relative) ease of touchscreen controls, but most people don't have time to play a smartphone game for more than about 3 minutes at a time, so you'd better design the game accordingly. Additionally, whereas computers come with a 101-key keyboard and mouse, and consoles with a minimum of Thumbstick, D-Pad, 4 face buttons and 2 Shoulder buttons[[note]]and sometimes +2 shoulder buttons, +1 thumbstick, buttons _under_ the thumbsticks, and even a touch-sensitive interface if you're a [=DualShock=] 4[[/note]], a touchscreen phone has only... its touchscreen to display controls on. You will need to think hard about your GUI and how you want to display things. This is not to say that you ''can't'' have titles on a phone from genres that are normally dominated by computers (such as RealTimeStrategy title ''[[ Tactile Wars]]'') or consoles (HackAndSlash ''[[ Implosion: Never Lose Hope]]''; ActionRPG ''VideoGame/FinalFantasyXV: Pocket Edition''); it is simply to say that it's easier for the player to get in their own way on a phone.

You need a business model. As mentioned, video games are a business, so if you are smart you'll build profit possibilities into your game.
* The more traditional model has been described as the "Games As Product" model. You create a game, you sell it for a very large chunk of change ($60 per game for UsefulNotes/TheEighthGenerationOfConsoleVideoGames), and once it's on store shelves, you never touch it again. The game exists as it is, bugs and all. While you make fewer sales, you get larger chunks of money, and you can always create {{Expansion Pack}}s if your game is successful.
* The model pioneered by [=MMORPGs=] and Microsoft are the "Games As Service" model. You create a game, you release it, you update it frequently. There are multiple places you can make money: [=MMOs=] charge a monthly subscription, for instance, while most {{Mobile Phone Game}}s that use this model place limits, sometimes artificial ones, on gameplay and then offer "In-App Purchase" options to let you get around it--''VideoGame/FarmVille'' forces you to expend Energy on every action and regenerates it slowly (1 charge every 15 minutes, one charge every hour, etc), but allows you to purchase more for real money. They may also allow you to LevelGrind your way to certain bonuses or simply buy them for convenience and time-saving. This can verge into BribingYourWayToVictory, but the company's not likely to care, since they're the people you're bribing--and, in well-designed games, the fact that players ''can'' buy power will be worked into the CompetitiveBalance. Of course, in any situation where players can buy power, you ''also'' have to think about the game's overall economy -- how far you want the LensmanArmsRace to go. Suddenly you need to understand financial matters like [[ hyperinflation]]. The good news is that if it ''works'', you can find yourself enjoying the benefits of games like ''VideoGame/WorldOfWarcraft'' or ''VideoGame/GameOfWarFireAge'', which consistently earn their creators millions of dollars '''a day'''.
* The model pioneered by {{Collectible Card Game}}s and {{Card Battle Game}}s is the "Games As Collection" model: you buy ''pieces'' of the game. Such games typically incorporate a GottaCatchEmAll mentality to encourage continued purchasing. They require you to ''continue'' releasing {{Expansion Pack}}s in order to keep the game fresh. The upsides are that novelty is a very powerful factor, and a game that is constantly new, the {{metagame}} constantly changing, can be addictive on a "CrackIsCheaper" level. The downside is that it's ''very'' easy to release {{Game Breaker}}s on accident. You're also going to have to deal with [[NewRulesAsThePlotDemands Complexity Creep]], since you keep adding on new features and such. Players who leave the game will have trouble returning, because so many things may have changed in their absence. (All of this is true of the "Games As Service" model too, by the way.)

You do not have to choose only one of these models. ''VideoGame/WorldOfWarcraft'' requires (or required) you to purchase the game ''and'' pay a subscription fee. ''VideoGame/AngryBirds'' requires you to purchase the game (on [=iPhones=]; [[NoExportForYou Android players get the game for free]]) ''and'' has IAP options. Decide how you want to do it and go from there.

You should decide whether other players will be a part of the gameplay experience.
* A single-player game features just you, your skills and your abilities. This kind of design is discouraged because (it is believed) it offers fewer hooks for IAP: players who want to triumph over other players will happily shell out money to do so, but not when facing only the computer. Careful balancing of the difficulty curve could avert this. It also, well, lacks multiplayer. Two heads are better than one, and two people playing a game results in more interesting experiences. Having said that, a 1P experience is the absolute best platform for ''actually telling a story''; most multiplayer games that attempt to thatch their 1P campaign into the multiplayer experience (''VideoGame/AgeOfConan'', ''VideoGame/{{Titanfall}}'') do not succeed, and indeed have been mocked for how [[{{Narm}} poor]] the attempts actually were.
* CoOpMultiplayer is when you and other players work together to achieve a shared goal. Successful video games (''VideoGame/Left4Dead'', ''VideoGame/{{Borderlands}}'') and board games (''Forbidden Island'') have been created that utilize this model. Such games can be extra-vulnerable to {{troll}}s and {{griefing}}, so the developers need to work in countermeasures, but when done correctly they create FireForgedFriends from strangers and can result in chaotic, spectacularly fun experiences.
** The hybrid child of CoOpMultiplayer and Single Player is DropInDropOutMultiplayer, perhaps best illustrated by ''VideoGame/DeadSpace3''. During the 1P campaign, the first player controls {{protagonist}} Isaac Clarke; when a second player joins, an {{NPC}}, Sgt. John Carver, becomes their avatar, and fights alongside Clarke as he progresses through the plot. Visceral Games took pains to seed "trap doors" throughout the game's script, so that Carver could be PutOnABus (or have [[TheBusCameBack The Bus Come Back]]) at a moment's notice, without having any impact on the story. The ''Franchise/{{Halo}}'' games didn't even bother with a {{Watsonian}} justification; they just had extra characters show up and stand around when {{Cut Scene}}s happened.
* CompetitiveMultiplayer is when you and other players ''compete'' to achieve goals. The vast majority of video-game multiplayer, from FightingGames to FirstPersonShooter Deathmatches to sports games and more, take place in this space; they can use (theoretically) equal teams, or be giant free-for-alls.
** Symmetric Multiplayer is a situation where both teams have the same goal--"Capture the Flag," "Kill ## people," "Score goals," "checkmate your opponent's King." While this may sound boring, it should be pointed out that the vast majority of sports and games throughout history use this model. It's also way, ''way'' easier on the developers when it comes time to institute CompetitiveBalance.
** AsymmetricMultiplayer is where the two teams have ''different'' goals. Some ''VideoGame/UnrealTournament'' or ''VideoGame/TeamFortress2'' matches involve one side attacking a fixed position and the other defending it, which has significant impact on the strategies and tactics each side uses. The recent FirstPersonShooter ''VideoGame/{{Evolve}}'' revolves around this trope: all matches are 4v1, with human Hunters pitted against one very large alien Monster. Again, the downside of this is in balancing. Each character / ability / job class / whatever is probably stronger at offense than at defense (or vice versa), and yet it still needs to be viable when being used the "wrong" way, so that the StopHavingFunGuys don't make too much noise.

You should decide on the ''timing'' of your multiplayer experience. This is true even if you want a solely single-player campaign, as we shall shortly explain.
* Synchronous multiplayer is the traditional experience. Two or more people sit down and play the game at the same time, competing or cooperating in real time. Doing this requires a certain amount of infrastructure--servers, for instance, that the players can connect to so that their control inputs are thatched together properly--but provides the most thrilling experience. Almost all eSports involve synchronous multiplayer.
* '''''A'''''synchronous multiplayer is when only one person plays at a time. While this sounds ridiculous, it's OlderThanTheyThink: PlayByPostGames of TabletopGame/{{chess}} have been a thing for centuries. It can be used both for co-operative and competitive play.
** '''Competitive''': In ''VideoGame/ClashOfClans'', players can only be attacked whilst offline, with the AI controlling your defenses on your behalf. The Creator/Suda51 game ''VideoGame/LetItDie'' is a ''VideoGame/DarkSouls''-influenced permadeath {{roguelike}} where your slain character becomes an NPC enemy in a randomly-selected instance (yours, someone else's, whatever). If you kill a former PlayerCharacter this way, you get extra loot; if it kills ''you'', its owner gets bonuses. A lot of mobile games use this model because it allows you to "participate" (defensively) in battle even if you are not on your phone; additionally, because [[ArtificialStupidity AI typically isn't very good]], it means that most attacking players will win -- a thing most players enjoy doing.
** '''Co-operative''' asynchronous multi typically relies on SocializationBonus. In ''VideoGame/BraveFrontier'' you form a party of five characters, and are allowed to "borrow" a friend's character to serve as a SixthRanger. In ''VideoGame/FarmVille'', you can't complete certain tasks until you collect TwentyBearAsses... but said items can only be provided by ''friends'' who also play the game. (Or {{microtransactions}}.) This is why ''Farmville'' players are always sending you Facebook requests asking for help.

And finally, you should probably decide on what ''genre'' your video game is going to be in. There are a ''lot'' of these, a number of them already encapsulated by their own articles:
* SoYouWantTo/WriteAnAdventureGame
* SoYouWantTo/WriteADatingSim
* SoYouWantTo/WriteAFirstPersonShooter
* SoYouWantTo/WriteAMajorMMORPG, SoYouWantTo/WriteAMinorMMORPG
* SoYouWantTo/MakeAMetroidvania
* SoYouWantTo/WriteARealTimeStrategy
* SoYouWantTo/WriteAnRPG
* SoYouWantTo/WriteASurvivalHorrorGame
* SoYouWantTo/WriteAWesternRPG

The biggest pitfall for any game is to make the gameplay bad, but designing fun gameplay is much easier said than done. This can be seen in TheProblemWithLicensedGames, which (when based off of a movie with a good story) shows the pitfall of trying to attach a good story to a horrible game. This is the golden rule for all games at all - do not let the gameplay detract from your overall goal for the game.

This is another subject where each game genre has its own tropes, but there are a few overlying ideas you should use to avoid having your game's gameplay flop.

!!Keep It Simple, Stupid
The most important is to never let an idea stand undisputed for a game. Ever. Even if it's a solid idea, ends up making it into the game and being popular, during the creation process don't be afraid to put the idea aside for another one to serve your reasons. If you refuse to make changes to a or any part of the game, then you are doomed to failure. This can best be shown by the "Online FPS" example. After ''VideoGame/ModernWarfare'' made it big, every FPS had to have online multiplayer. The producers wouldn't budge on that one inch: had to have it to attract the crowd, and thus the devs needed to spend time, resources and energy on it. As a result, the other elements of the game(s) suffered in quality. Most people will tell you that most FPS games from around 2007 to 2010/11 were not that entertaining, from ''VideoGame/{{Singularity}}'' to games that didn't even get made because of this halfway switch. Now, that's not to say having a solid idea and doing as much as you can for it is a bad thing, if it's what you want to do, but giving absolutely no room for change in an idea has a high chance of causing problems.

Also note that taking gameplay elements ''out'' of the game can, believe it or not, actually improve the product. One of gaming's most recent rave successes, ''VideoGame/TheLastOfUs'', provides a compelling example. The entire duration of TheTeaser, you have extremely limited control over your characters: you can move your character, you can move the camera, there are a couple QuickTimeEvents, and ''that's it.'' "How could that be fun," you ask, "that's bordering on ControllableHelplessness." And the answer is, Yes, it absolutely is... and what else could be more compelling ''in a ZombieApocalypse''? Heck, you don't even have a gun! Sure, ''Joel'' has his little revolver, but the only time he fires it is in a CutScene, and after that he gives it to his brother Tommy to wield. You, ''The Player'', never have a gun. And that increases the sense of triumph when you reach the military perimeter: despite having literally nothing but your feet, you have not only escaped from zombies, but you have carried your daughter Sarah to safety. You are the MostTriumphantExample of an ActionSurvivor. ...And, in addition, this increases the impact of the PlayerPunch when [[FirstEpisodeSpoiler Sarah dies]]; all that hard work, all that desperation, all that sacrifice, [[DownerEnding for nothing]]. It's a brilliant EstablishingCharacterMoment for not only [[ShellshockedVeteran Joel]] but for the game as a whole, and it's accomplished by, essentially, ''not'' letting the player play the game.

This is even true in more complex games. Daniel Friedman, a writer for Polygon, [[ commented]] on how some of the simplest characters (in this case, Warwick from ''VideoGame/LeagueOfLegends'' and Reaper from ''VideoGame/{{Overwatch}}'') are the best in their titles, simply because their extremely limited gameplay styles foster the development and expression of skill. They ''force'' the player to "git gud," as gamers like to say it these days, and offer freedom and creativity by limiting choice. (That's actually the MO of almost ''all'' popular video games these days. {{MOBA}}s and Hero Shooters like ''Overwatch'' and ''VideoGame/TeamFortress2'' are very specific about what each of their characters can and can't do. Despite the inherent limitations, these games are amongst the most popular, the most played and the most financially lucrative in the world. Take a lesson accordingly.)

-->"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

!!Story Vs Gameplay: Fight!
Writing the story of a video game is tricky for the same reason that films are trickier to film, and songs tricker to write, than novels: there's more than one storytelling language being used simultaneously. In all of these media, there is a '''story''' -- who the MainCharacter is, what they want, why they can't have it, and why the audience should give a [PrecisionFStrike] about it. But in films there's also "cinematography," which involves the aesthetics of the moving image and how ''it'' can tell a story; [[TheNostalgiaChick Lindsay Ellis]] has an excellent analysis of how camera angles in Creator/MichaelBay's ''Film/{{Transformers}}'' [[ actually obscure]] the only CharacterDevelopment in the film. In a song, you have lyrics, but you also have the music, and the two can work at cross-purposes -- for instance, the LyricalDissonance of a jaunty, happy piano tune to which Music/EltonJohn sing, "[[ Think I'm Gonna Kill Myself]]." And in video games, there's not only the story being told by the, well, story, but also the one being told ''by gameplay''. And, just as in the other two examples, sometimes the two stories don't agree.

And this is a problem.

A quick foreward: to formalize vocabulary for this section, we are going to borrow some terms from James Howell's seminal work of games criticism, "[[ Driving Off the Map]]." Particularly, we are going to talk about '''Player Objectives''' -- the things which a human being, sitting around in RealLife playing VideoGames, hopes to achieve -- and '''Actor Objectives''' -- the things which the PlayerCharacter, an in-game entity controlled by the player, hopes to achieve. These two are not always the same; for instance, in the ''The Last Of Us'' example above, Joel has the Actor Objective of saving his daughter, while the player has the Player Objective of correctly manipulating the Dualshock 3 controller in a way that results in Joel navigating through the in-game world, avoiding obstacles and zombie attacks. These are being established because the two sets of objectives are not always in accord.

Sometimes Player Objectives beat Actor Objectives. ''VideoGame/MassEffect3'', and its notorious ending, is an example. Per WordOfGod, the CentralTheme of the story is, "[[TheChainsOfCommanding You can't save everyone]]." WarIsHell, and somewhere along the line you're going to have to choose [A] over [B] and watch [B] die a fiery, dramatic, slow-motion death with OneWomanWail in the background. In other words, there is no GoldenPath where you get absolutely everyone on your side. The salarians still believe that inflicting a SterilityPlague on the krogan, and resulting ChildlessDystopia, was justified? Then you have to pick between them and the krogan. The quarians won't stop fighting their RobotWar against the geth? Then you have to choose one or the other. This is a very effective Actor Objective, and the resulting game would have been awesome -- arguably, better than what we actually got (and what we actually got was pretty darn good, notorious ending notwithstanding). The ''problem'' is, Player Objectives mandate the inclusion of a GoldenPath. There's been one for the other two games in the series, and CentralTheme of ''the series'' is, "You can ''always'' TakeAThirdOption; there ''is'' a GoldenPath." For the third game to discard this would absolutely confound Player Objectives. So they kept the GoldenPath; it exists. You ''can'' get the quarians and geth to reconcile; and the salarians come around if you stick to your guns. Even worse, situations in which there genuinely ''was'' no Third Option--in which you must endure the PlayerPunch of condemning a NonPlayerCharacter to death, with no recourse whatsoever, as you did on Virmire--were DummiedOut. ([[spoiler:It was to have been on Thessia: Liara and the Virmire Survivor were going to be your mandatory squad members, and you'd only have time to save one when the temple floor collapsed.]]) Thus, Actor Objectives were defeated by Player Objectives. The story tells you one thing but gameplay lets you do the exact opposite. And, even worse, [[PoorCommunicationKills the writers weren't told about it]], with the result that there's no GoldenEnding even though there ''is'' a GoldenPath. (That disconnect is part of why the ending was so notoriously ill-received.)

''VideoGame/BioShock1'' had a similar issue: Actor Objectives mandate that you spare the Little Sisters, {{Heartwarming Orphan}}s who are victims of a heartless system, but Player Objectives encourage you to kill them, because if you don't, ''you can't buy new magic powers.'' And powers are kind of important in Rapture, not just as a plot point (the game takes place AfterTheEnd was brought about by ''abuse'' of said "Plasmids") but because your character, Jack, is barely one step up from an ActionSurvivor, and needs all the help he can get. Ken Levine at least had the wit to include MultipleEndings depending on which decision you made, but it still eroded the {{escapism}} that video games often offer as one of their prime selling points, because killing Little Sister resulted in the ''bad'' ending. Player Objectives were defeated by Actor Objectives because ''the game punishes you for playing all of it.'' This conflict was so egregious that someone actually coined an entire new term, "ludonarrative dissonance," to describe Actor Objective / Player Objective conflict.

Like any other trope, ludonarrative dissonance can be employed deliberately; ''[=BioShock 1=]'' did so, as did ''VideoGame/SpecOpsTheLine''. You have to be really careful about doing so, though, because the only thing it can possibly do is piss The Player off. ''VideoGame/MetalGearSolid2SonsOfLiberty'' is the poster child for this subversion -- especially because it's the game whose critical analysis gave us the Player Objective / Actor Objective terminology. In ''[=MGS2=]'', the two were constantly at odds; Raiden might defeat a boss, but would never get to deal the finishing blow; and succeeding at sneaking aboard Arsenal Gear would result in Raiden getting captured (and having to escape [[MaleFrontalNudity butt-nekkid]]) and the destruction of the Plant which he had worked so hard to save. People didn't like playing as Raiden, because he never seemed to succeed at what he was trying to do. This was ''very'' much intentional; the whole point of Raiden as a character was to make fun of, or perhaps deconstruct, ''the player'', and their Player Objective of "Relive ''VideoGame/MetalGearSolid'' and get to be a badass like Solid Snake." Through Raiden, series creator Hideo Kojima was able to point at players and laugh: "You wanted to be Solid Snake. ''You are''. Contemplate ThePerilsOfBeingTheBest. Look at what a wreck Snake is. ThisLoserIsYou." This upset players [[SarcasmMode for some reason]]. Alienating your audience is a ''very'' dangerous thing to do, even if on purpose. So DoNotTryThisAtHome, unless you're 100% sure you know what you're doing.

The point is this: games are supposed to be fun, so make sure the story is encouraging you to have the same kind of fun that gameplay is. Don't punish the player for accessing basic features of the game. But likewise, don't come up with story reasons for the player to be locked out of basic features either. Games work best when the Player Objectives and Actor Objectives are the same: You Get To Do [X], and there's no reason, narrative or mechanical, why you can't.

!!Choices, Choices and More Choices
You'll need to consider player agency. Video games are an interactive medium, where players are given choices--or, at least, the ''illusion'' of choice--and expect to see those choices respected and reflected in how the game proceeds. Sometimes this is merely a gameplay aspect--"I chose 'Burning Fist' instead of 'Frost Punch,' so I better be able to use Burning Fist when I press Circle-Circle-Square"--and if you're having problems you need to talk to your programmers or your Quality Assurance team. But sometimes it's a story choice. So if you give players choices over the events of your game's story, they ''have'' to play out over the course of the rest of the game. This is why {{Railroading}} is so decried as a trope: it not only renders the player's choices moot, but it pokes holes in the WillingSuspensionOfDisbelief. ''VideoGame/MassEffect3'' had some bad examples of this. In the first game, you made a choice whether to wipe out an alien who was the LastOfItsKind or not. In ''[=ME3=]'', that alien reappears in a specific mission... regardless of what you chose. It was [[TropesAreTools kind of cool]] to have said alien appear no matter what, but--once again--this writing decision made the choice in the first game [[TropesAreTools retroactively meaningless]]. (And it was one of the most significant emotional beats of the first game, so having the writers just throw it out was a little disrespectful.)

Writing a game means making sure you give players choice. And that can be difficult, because every option The Player has? You had to decide to give it to them. In other words, (the illusion of) choice is something you have to ''create''. DevelopersForesight needs to be ''mandatory'' for your process, because if you don't, there's no game. It is your job to decide what actions are available. And that means you need to sit down and think about as many possible actions that a player ''could'' want to take, for fear of spiking WillingSuspensionOfDisbelief via {{Railroading}} or other silly obstacles (InsurmountableWaistHighFence, WhyDontYouJustShootHim, etc).

Beware of {{Moon Logic Puzzle}}s, but also beware of the opposite: AcceptableBreaksFromReality. Video gamers are GenreSavvy enough to know that everything they can do is something you gave them the option to do. They have also played a ''lot'' of video games where they tried to TakeAThirdOption and were unable to because ''you'', the '''programmer''', didn't realize they'd want to. The combination of "most developers are stupid" and "But I'm not" can be some ''serious'' GuideDangIt moments. Two free examples: in ''VideoGame/GodOfWarII'', there's a puzzle that you solve by raising a timed platform, which is shaped like a T, and then wedging it in the air using a pushable block. The problem is that you can only do this if the T-shaped platform has the collision physics ''of'' a T-shaped platform, instead of just being a giant pillar, the way most programmers would do it to save time. The puzzle is only solvable if an object obeys the laws of physics. The mere fact we have to ''say'' that is an indicator of just how egregious "Acceptable" Breaks From Reality have become at times. The other is from the seminal ''VideoGame/SpecOpsTheLine''. Late in the game, one of your NPC friends is strung up by a civilian lynch mob, with your characters coming across the process too late to stop it. The game suggests either letting them go or slaughtering the civilians; the [[TakeAThirdOption Third Option]], FiringInTheAirALot to scare them off, works pretty well in RealLife but might not in a video game because the civilians might not be programmed to be intimidated that way. These are just two examples where DevelopersForesight actually caused more problems than they solved. The trope is ''nowhere'' near as prevalent as it could be, and ''players know that''. So never forget: players know that their choices are artificially limited by ''your'' decision-making capabilities. It will take a lot of coaching, and a lot more excellent gameplay design, before this fact ceases to hold sway over gamers.

!!'''Controls vs. Complexity vs. Speed'''
This is a tricky one because it's not a sliding scale; it's a ''triangle'', where gaining points in one means sacrificing points in two others. But, to get on with things:

In the ideal game, you can do 1) Lots of cool things 2) easily and 3) in real time. In reality, you will often have to sacrifice at least one of those ideals. The reason for this is simple: the human being is a limited creature which can only absorb, and react to, limited amounts of information. There is only so much a single player can do without getting overwhelmed.

Ultimately, many video games can defined by which of them they sacrifice.
* Real-time action games like ''VideoGame/SuperMarioBros1'', ''VideoGame/Tetris'', ''VideoGame/{{Portal}}'' and ''VideoGame/{{Doom}}'' focus on Controls and Speed, resulting in a game with the qualities of Real-Time and Ease Of Play. The Player Characters in those games can't do a whole lot, in the grand scheme of things, but it's relatively easy to make them do those things. The upsides to this style of game are the (relatively) low barrier to entry; the downsides are the (relatively) limited movesets available to player(s).
* Turn-based games focus on Controls and Complexity, resulting in something that's easy to play (InterfaceScrew and GuideDangIt notwithstanding) and gives the player tons of options, but doesn't move very quickly. Think about TabletopGames/{{chess}}, or FourX games, or ''VideoGame/{{Pokemon}}'', or even ''TabletopGames/DungeonsAndDragons'': there are a ''lot'' of things you can do in these games, but you cannot do them in anything even approaching real time, and sometimes you can't even do them efficiently! The result is a cerebral, strategic style of gameplay that will appeal to certain people and bore others to death.
* RealTimeStrategy and FightingGames have Complexity and Speed, but require the player to commit a great deal of information to memory. They have big {{metagame}}s, from TacticalRockPaperScissors to control inputs with SomeDexterityRequired to even remembering what the hotkey is for a specific action. This kind of game is good for people who can absorb a lot of information quickly, but bad for people who just want to pick up and play.

Various games have attempted to merge more towards that Platonic Ideal of a video game -- complexity, speed ''and'' simple controls -- with varying levels of success. What's worth studying is the ways games have invented to ''get around'' these limitations.
* In Real-time Action games with NonPlayerCharacters, said [=NPCs=] may be controlled by AI. This can verge into a hair-tearing EscortMission, so the AI needs to be either smart ([[ArtificialStupidity yeah right]]), helpful in other ways besides combat, or have GameplayAllyImmortality. ''VideoGame/BioshockInfinite'' was praised for capturing the latter two elements, and finding ways to justify the latter within the story. ''FinalFantasyXII'' implemented PausableRealTime, as well as the "Gambit" system, which allowed you to program your non-controlled characters to take (real-time) actions when certain criteria were fulfilled ("'''if''' [any party member] '''is''' [below 25% HP], '''hit them with''' [a Heal spell]"; "'''if''' [any enemy] '''is''' [flying], '''hit them with''' [anything Earth-elemental]").
* In Real-time Action games, responsibilities can also be split up. Consider ''VideoGame/TeamFortress2'', ''VideoGame/GunsOfIcarus'' and {{MOBA}}s like ''VideoGame/Dota2'': players work together to achieve several goals (namely, "1) Don't lose, 2) Win") but are limited in what they, personally, can contribute to that victory (defense, healing, offense, psychological warfare, etc).
* In Turn-based games, the question starts to depend on the scope of the decisions being made every turn. ''VideoGame/FinalFantasyX'' created a fluid battle system that only involved one character moving per turn (as opposed to ''Civilization'' where you might have to give orders to five or ten cities at once, not to mention your military); FourX games often implement a notification system, creating UI elements that remind you to do one of the (many, many) things you might want to do.
* Complex Real Time games often let you use (or create your own!) keyboard shortcuts to do things quickly. They also focus on the UI, providing you information that you can both absorb and ignore, depending on what you're trying to do at the moment. (Ignore it at your own peril, of course.)

!!'''Some Other Considerations'''
!!! Graphics
Graphics are always a big thing in video games these days. Everyone wants good ones... but creating good ones takes a lot of time and effort. It can also require a great deal of processing power in terms of the hardware necessary to run your game. Even worse, graphics ''age''. Games that were considered to have stellar, cutting-edge graphics ten years ago (''VideoGame/TheElderScrollsIVOblivion'', ''VideoGame/TheSims II'', ''VideoGame/{{Battlefield}} II'') look dated today. One simple workaround is to look at games which ''don't'' look dated--''VideoGame/TheLegendOfZeldaTheWindWaker'', ''VideoGame/TeamFortress2'', ''VideoGame/{{Limbo}}'', ''VideoGame/SpiralKnights''. What do these games have in common? Simple: they don't try to be photorealistic. Instead, they have an ''art style'' with graphics that aren't ''supposed'' to look like "reality" and instead like... well, whatever they're trying to achieve (cel shading, sliding silhouettes, etc). And, since they achieve it, their graphics become timeless. Of course, to achieve this, you need (someone to provide) very strong art direction, which is also a rare commodity. It's TechnicianVsPerformer for graphics, and the question is which one you decide to throw money at. But there's a Sliding Scale Of Photorealistic Vs. Artistic, and the simple fact is that the second one ages better.

!!! Minimum Viable Product
Beware, ''beware, '''beware''''' the trap called the "Minimum Viable Product." As the term suggests, this is a benchmark that you and/or your team sets, representing the absolute most bare-bones version of the game that can be released to consumers. Exactly what this benchmark consists of -- what the core loop looks like, how many extras are available, how much content you have, if there is multiplayer, etc -- is going to depend on the nature of your product itself. For instance, for Creator/TelltaleGames, the MVP is "An engine and 20% of the content" because their games are episodic and the "Expansion Packs" consist solely of data that is slotted in later. But if you're on the team that made the original ''VideoGame/FinalFantasyVII'' your minimum is "the engine, ''all'' the content, and every bell and whistle we decide to add (including some extremely-well-hidden option that [[UrbanLegendOfZelda lets you revive Aerith]])." This can vary even within your genre; the creators of the [[MultiplayerOnlineBattleArena MOBA]] ''VideoGame/LeagueOfLegends'' decided to ship their game with 40 characters, whereas the competing ''VideoGame/{{Demigod}}'' went out with a mere ''eight''. (And that's why you've never heard of ''Demigod''.) Additionally, it's going to go up and down as the product evolves -- this is done, that is not; we can't implement this feature for various reasons; ExecutiveMeddling requires us to add [this], whether or not it fits. ''And'' it's prey to the current climate of gaming, specifically the "Games As Service" model that dominates.

Because games can be, and are, updated on a regular basis, it's become increasingly acceptable to take an ObviousBeta, declare it meets your MVP, and ship it, often by MovingTheGoalposts to accommodate the product that currently exists. '''''Whatever you do, don't do this.''''' Very few games that shipped half-finished were financial successes, because the simple fact is that if players are going to spend a full game's worth of money, they want to receive a full game's worth of content for it, ''today'', not tomorrow. Even worse, because of the way people play games these days, they're gonna go through content fast. People who make smartphone games can tell horror stories about how they shipped games which, they thought, had months of content, only to have players get through it in days or even ''hours''. When this happens, players lose interest, and fast. The fate of games like ''VideoGame/FalloutShelter'' and ''VideoGame/{{Titanfall}}'' are examples of games that ''could'' have gotten huge... had they been released with sufficient content. But no: someone took a half-finished version and declared it the Minimum Viable Product, even though it couldn't hold people's attention. And didn't.

Now, the flipside is that MovingTheGoalposts is a common feature of game development, as artistic, technical and scheduling limitations fall into place. Eventually you will have to compromise. YouTube's [[ Design Doc]] gaming-analysis channel gives [[ an example]] of an adapation of ''Film/ANewHope'' in which Luke can decline the CallToAdventure and spend the rest of the game in a FarmingSimulator. This is something that, almost certainly, would get cut during production, because its return-on-investment is dismal. Goalposts will move; goalposts ''have'' to move. The key is to know which of your goalposts are critical to the game you want to create.

!'''Potential Subversions'''
Subversions to a video game are impossible, but as discussed in meta-game and deconstruction, actual gameplay expectations can be subverted. A good example is the moral choice system. Most players expect either A) The whole thing would have a major, or at least notable, impact on the game and being in the middle ground is pointless in terms of bonuses, maybe with the possibility that your alignment decides things, not what you have done to get it. (''VideoGame/InFamous'', ''VideoGame/KnightsOfTheOldRepublic'') or that B) It's just a small setup to give the player an illusion of choice and doesn't really matter in the end, except maybe for moral choice system having some impact on gameplay. ''VideoGame/DantesInferno'' never really went anywhere with moral choices, but they did buy you upgrades. ''VideoGame/BioShockInfinite'' makes a small point in its gameplay on moral choices without actually having a moral system.

Working from that gameplay convention, there could be a few subversions to the expected style. A subversion for type A could be a major shock for people when what they are so used to seeing gets pulled out from under them. Maybe the moral choices actually affect the storyline and characters in an organic way, maybe they affect how people treat you but actually doesn't influence the main story when a sudden story swerve comes out of nowhere and rendered the system pointless. Done well, it can be very meaningful and show players on how morality comes into play on some things, but not so much so for others. Done poorly it can look like a half-hearted attempt at removing the system midgame. This shows the impact of storytelling conventions in games and how players can look at them.

Type B shows how players can expect gameplay and story to be separated and simply another system. A subversion for type B could easily include moral choices having no effect on gameplay, seemingly, and then suddenly spring up as being important and recognized by other characters. A fantasy RPG has a moral choice system that seems to only effect what type of spells or skills are unlocked for the PC, and suddenly in the middle of a game a character mentions how the forces responsible for magic are actually paying attention to the player, and are granting him spells based on how he acts and solves problems. Done well, and followed up upon so it doesn't just look like a HandWave, it can actually be a surprise to the player about how this thing they had mentally placed as gameplay is touched upon by the world it happens in and has actual meaning. Done poorly it will still look like a HandWave, or maybe even a VoodooShark, and annoy the player that such things were being justified when it was just fine as a gameplay feature.

!'''Writers' Lounge'''
!!'''Suggested Themes and Aesops'''
!!'''Potential Motifs'''
!!'''Suggested Plots'''
!!'''Set Designer''' / '''Location Scout'''
!!'''Props Department'''
!!'''Costume Designer'''
!!'''Casting Director'''
!!'''Stunt Department'''
!'''Extra Credit'''
!!'''The Greats'''
!!'''The Epic Fails'''