History SoYouWantTo / WriteAWesternRPG

* How do the enemies and AI followers decide whom to attack next and how? Is friendly AI customizable by the players? Do allied [=NPCs=] generate [[UsefulNotes/RolePlayingGameTerms "threat"]] that makes them priority targets, and if so, how is it calculated? Which in-game means do the players have to manipulate threat?

* License a commercial MediaNotes/GameEngine. If you're on a tight budget, ''UsefulNotes/{{Unity}}'' is in great demand among indie developers; beyond that, Website/TheOtherWiki helpfully provides an [[http://en.wikipedia.org/wiki/List_of_game_engines extensive list]].
* Use a GameMaker geared towards [=RPGs=], such as the UsefulNotes/RPGMaker. UsefulNotes/RenPy is pitched as a VisualNovel engine, but several Western developers managed to create engaging [=RPGs=] in it.

Be sure to read SoYouWantTo/WriteAnRPG for more general advice. Other guides, particularly SoYouWantTo/WriteAStory, SoYouWantTo/WriteAVideoGame, SoYouWantTo/WriteAHeroicFantasy, SoYouWantTo/WriteAHighFantasy, and SoYouWantTo/WriteACyberpunkStory, may also be helpful, since this guide will focus less on giving you ideas for plots and characters and more on providing a general overview of important considerations when designing an RPG. Lastly, make sure you're familiar with common UsefulNotes/RolePlayingGameTerms.

* '''Color'''. The graphics (and sound) of your game, ranging from the interface design to the [[ConceptArt visual style of characters, levels, and enemies]]. Usually implemented within a UsefulNotes/GameEngine.

''The seven core tropes are thus: PlayerCharacter(s), CharacterCustomization, WorldBuilding, TheQuest, UsefulNotes/GameEngine, GameSystem, and TakeYourTime.''

VariousVideogameViews affect PlayerAndProtagonistIntegration, so it is important to consider their advantages and limitations and choose or develop the UsefulNotes/GameEngine accordingly. Some would argue that IsometricProjection is the one and true view in classic Western [=RPGs=] but that is [[http://www.joystiq.com/2012/09/21/what-makes-a-classic-rpg-everything/ not the whole truth]]. The three most common views in role-playing video games are:

* License a commercial UsefulNotes/GameEngine. If you're on a tight budget, ''UsefulNotes/{{Unity}}'' is in great demand among indie developers; beyond that, Website/TheOtherWiki helpfully provides an [[http://en.wikipedia.org/wiki/List_of_game_engines extensive list]].

* {{Expansion Pack}}s run on the same UsefulNotes/GameEngine and GameSystem as the core game and provide additional content. Some are standalone adventures set [[{{Prequel}} before]], after, or in an AlternateTimeline of the main campaign; others are integrated into it. You can usually produce an expansion pack or two for every successful core game without much thought.

Finding true "[[DarthWiki/SoBadItsHorrible epic fails]]" among published and somewhat well-known examples of the genre is actually quite hard, because ever since mid-90s, {{Western RPG}}s have not been mainstream enough to flood the market with shoddy hack jobs. Most horrible [=RPGs=] either never leave production, or pass into obscurity immediately, unnoticed and unplayed by anyone, so it's much easier, in fact, to list games that have turned out alright but still failed to live up to their full potential for different reasons:

** A {{BFG}} for when you absolutely have to kill everyone in the room

* BonusDungeon. Of particular interest is the BrutalBonusLevel variety with a BonusBoss at the end. It is the one place (besides the HarderThanHard difficulty) where having players RageQuit in frustration is a good thing.

* Personal Drama Endings. In these games, the endings differ mainly in which PlayerParty members, including [[TheHeroDies the PC]], survive them and [[FateWorseThanDeath in which state]]. In some games, they can be rated from [[EverybodyLives Good]] to [[CharactersDroppingLikeFlies Bad]], while others [[BittersweetEnding defy an easy]] [[AmbiguousEnding assessment]] but may include a clear GoldenEnding that requires going an extra mile to unlock. What ending the player gets is usually determined by several key choices spread throughout the game, although the last one (or the [[BigFirstChoice first]]) may be [[LastSecondEndingChoice disproportionately momentous]].

In the latter two systems, CharacterLevel becomes superfluous but is often included, nonetheless, for the ease of PowerLevel assessment. This is usually done by dividing the total number of skill points/levels gained since the character creation by a two-digit number (often 10). In some games, {{Level Up}}s occur automatically; in others, they have to be triggered manually (often by {{resting|Recovery}}), so the players can postpone the irreversible character build decisions that come with it.

* Are drops subject to LevelScaling? If so, make sure that [[NominalImportance unique named items]] are worth the effort of getting them; in other words, that they will not be outperformed by random level-scaled "junk" loot after the next LevelUp.

At this point, you also need to figure out how to avoid EmptyLevels. Players like agency and have the most fun in acquiring and trying out new powers that change how they play the game, so most of them will feel cheated if their hard-earned LevelUp gives them little besides a basic stat increase. "New powers" aren't necessarily limited to SkillScoresAndPerks -- for instance, a piece of previously restricted LevelLockedLoot can be one, too -- and it is imperative to give meaningful rewards to characters going up in level. Possible solutions to the EmptyLevels problem include:

* Giving the player a new power at every level (regardless whether it's predetermined by their class or chosen freely from a list), while discreetly increasing the basic stats in the background. In this case, you may want to have [[AbsurdlyLowLevelCap fewer total attainable character levels]] to make each LevelUp feel more significant, and also because having too many powers can easily make the players feel lost and risks padding the [[SkillScoresAndPerks skill tree]] with unnecessary variations of identical powers. This is the recommended option for story-driven games.

Classes are character templates that more or less rigidly define which skills the character can improve with each LevelUp. Very rigid class definitions, where certain skills and equipment are exclusive to certain classes, induce CompetitiveBalance and enforce inter-class cooperation within the PlayerParty. More lax class restrictions, e.g. where only certain classes gain bonuses from certain equipment/skills but others can still use them, enable more [[JackOfAllStats rounded builds]] and are thus better suited for single-character games.

* In sandbox [=RPGs=], loot must be, first and foremost, consistent with the game world. Which means that most of it will be junk, with tons of non-functional FlavorEquipment lying around or being carried by enemies for practical or sentimental reasons. Powerful named items still benefit from a [[FlavorText flavor narrative]], but don't need to outshine RandomDrops and player-produced inventory.

* [[WeaponOfChoice Weapons]]. See below for more info.

* ''VideoGame/DragonAgeII'', as a result of its extremely short development time, suffered from incredibly repetitive level design, from CutAndPasteEnvironments to endless {{Teleporting Keycard Squad}}s, which couldn't be offset even by a [[http://www.gamasutra.com/view/news/264322/Opinion_Why_did_Dragon_Age_II_leave_some_fans_so_cold.php novel storyline]].

* ''VideoGame/DragonAgeII'' suffered from extremely lazy level design, from CutAndPasteEnvironments to endless {{Teleporting Keycard Squad}}s, which couldn't be offset even by a [[http://www.gamasutra.com/view/news/264322/Opinion_Why_did_Dragon_Age_II_leave_some_fans_so_cold.php novel storyline]].

* License a commercial UsefulNotes/GameEngine. If you're on a tight budget, ''UsefulNotes/{{Unity}}'' is in great demand among indie developers; beyond that, Wiki/TheOtherWiki helpfully provides an [[http://en.wikipedia.org/wiki/List_of_game_engines extensive list]].

* Personal Drama Endings. In these games, the endings differ mainly in which PlayerParty members, including [[TheHeroDies the PC]], survive them and [[FateWorseThanDeath in which state]]. In some games, they can be rated from [[EverybodyLives Good]] to [[KillEmAll Bad]], while others [[BittersweetEnding defy an easy]] [[AmbiguousEnding assessment]] but may include a clear GoldenEnding that requires going an extra mile to unlock. What ending the player gets is usually determined by several key choices spread throughout the game, although the last one (or the [[BigFirstChoice first]]) may be [[LastSecondEndingChoice disproportionately momentous]].

* [[WideOpenSandbox Open World]] is a Hub Network taken UpToEleven, where the players travel back and forth between multiple towns and dungeons on a [[LoadsAndLoadsOfSidequests highly nonlinear web of quests]]. This is pretty much the only way to make a sandbox RPG.

Naming is important for both expansions and sequels. Most RPG titles follow the convention of "[series title] [installment number, unless it's the first game]: [subtitle]". Subtitle is optional for core games; omitting it helps avert ColonCancer when making expansions, since they are usually identified only by their subtitle. If you have trouble coming up with a story-relevant subtitle, you can always use [[ThePlace the name of the geographic region the installment is set in]] (this also works well for entire series' titles) or just take a StockSubtitle.

* [[DungeonCrawling Dungeon Crawler]]. The players' motivation in these games is to master the game system by improving their character's attributes, skills, abilities, and equipment until they can beat every challenge the game throws at them. Such games commonly feature randomized and respawning enemies, {{random|lyGeneratedLoot}} but [[SocketedEquipment customizable]] loot, intricate {{Skill Tree}}s, and {{Absurdly High Level Cap}}s. The story is used [[PlayTheGameSkipTheStory mainly as a mood-setting backdrop]], and the exploration is kept within strict bounds. Typical examples include all {{Roguelike}}s, ''VideoGame/{{Wizardry}}'' and ''VideoGame/{{Diablo}}'' series, and the latter's many clones.

* Giving the player a new power at every level (regardless whether it's predetermined by their class or chosen freely from a list), while discreetly increasing the basic stats in the background. In this case, you may want to have [[AbsurdlyLowLevelCap fewer total attainable character levels]] to make each LevelUp feel more significant, and also because having too many powers can easily make the players feel lost and risks padding the SkillTree with unnecessary variations of identical powers. This is the recommended option for story-driven games.
* The basic stat increases serving as stepping stones towards unlocking new powers, either via StatGrinding or in a complex SkillTree where stat increases lead up to new abilities. This lets the players set their own goals and revel in the feeling of achievement upon reaching them -- as long as it doesn't take too long to get from one new power to the next. This is the preferred option for dungeon crawlers, since it allows for a much finer control over the character build, as well as for sandbox games, as it allows for more customization.

* Is there a SkillPointReset option? While [[LaserGuidedAmnesia utterly]] [[InstantExpert unrealistic]], this feature lets the players [[AntiFrustrationFeatures fix messed up builds]] and [[RuleOfFun try out new things]] without restarting the game. It doesn't have to be readily available, of course -- for instance, it can cost increasing amounts of in-game currency, or only be available at certain opportunities (e.g. upon unlocking a PrestigeClass SkillTree, or when a NonPlayerCompanion gains their first level after joining, so players can re-spec them after trying out their predefined stats).

* VendorTrash. Generic items that are only good for selling in shops for extra cash (e.g. non-functional gems) and an occasional CollectionSidequest.

A Personal Call can segue into Epic Villainy after the Twist, but that is in no way a requirement. In fact, while SavingTheWorld is a popular gaming objective, its capacity to motivate a remotely GenreSavvy player is next to zero. Players who have not yet lived in the world you've created and don't [[VideoGameCaringPotential care about its residents]] will most likely [[EightDeadlyWords feel absolutely nothing]] about its catastrophes. Not to mention that by playing your biggest card right from the start, you bar yourself from any kind of meaningful SequelEscalation later on. Therefore, instead of [[TheWorldIsAlwaysDoomed putting the world in peril again]], start off with a personal threat and spin a central conflict around the main actors' desires and flaws -- while [[SaveTheWorldClimax saving epic threats for the sequels]].

Similarly to how classes are an attempt to demarcate the role a character plays in exploration and combat, CharacterAlignment was an attempt by early tabletop [=RPGs=] (particularly, ''TabletopGame/DungeonsAndDragons'') to define their role in dialogue and other non-violent interactions with {{Non Player Character}}s, such as {{Quest Giver}}s and party members or followers. Since ''VideoGame/UltimaIV'' and ''VideoGame/{{Fallout}}'', digital [=RPGs=] have gained a more dynamic tool of assessing a character's moral standing in form of the KarmaMeter.

* ''VideoGame/{{Fallout 1}}'' and ''VideoGame/{{Fallout 2}}'' have merged open world and narrative elements in a way that brought out the best of both worlds and wouldn't be replicated until ''VideoGame/TheWitcher3''.