In game design, balance is the concept and the practice of tuning a game's rules, usually with the goal of preventing any of its component systems from being ineffective or otherwise undesirable when compared to their peers. An unbalanced system represents wasted development resources at the very least, and at worst can undermine the game's entire ruleset by making important roles or tasks impossible to perform.
Any good computer game must be totally fair. It must be possible for the player to reach the objective and win. This is not to say the game cannot be complicated or random or appear unfair.-- Tim Barry of InfoWorld, 1981
An important trait of any game is the illusion of winnability. If a game is to provide a continuing challenge to players, it must also provide a continuing motivation to play. The game must appear to be winnable to all players, beginners and experts, but it must never truly be winnable or it will lose its appeal.
Dani Bunten was once asked how to play-balance a game. Her one word answer was "Cheat." Asked what to do if gamers complained, she said, "Lie!"
Chris Crawford wrote in 1982 of the importance of a game's "illusion of winnability"; Pac-Man is popular because it "appears winnable to most players, yet is never quite winnable". The illusion, he said, "is very difficult to maintain. Some games maintain it for the expert but never achieve it for the beginner; these games intimidate all but the most determined players", citing Tempest as an example.
A fair game is winnable but, InfoWorld stated in 1981, can be "complicated or random or appear unfair". Fairness does not necessarily mean that a game is balanced. This is particularly true of action games: Jaime Griesemer, design lead at Bungie, said that "every fight in Halo is unfair". This potential for unfairness creates uncertainty, leading to the tension and excitement that action games seek to deliver. In these cases balancing is instead the management of unfair scenarios, with the ultimate goal of ensuring that all of the strategies which the game intends to support are viable. The extent to which those strategies are equal to one another defines the character of the game in question.
Simulation games can be balanced unfairly in order to be true to life. A wargame may cast the player into the role of a general who was defeated by an overwhelming force, and it is common for the abilities of teams in sports games to mirror those of the real-world teams they represent regardless of the implications for players who pick them.
Player perception can also affect the appearance of fairness. Sid Meier stated that he omitted multiplayer alliances in Civilization because he found that the computer was almost as good as humans in exploiting them, which caused players to think that the computer was cheating.
Video games often allow players to influence their balance by offering a choice of "difficulty levels". These affect how challenging the game is to play, and usually run on a general scale of "easy", "medium", and "hard". Sometimes, the difficulty is set once for the entirety of a game, while in other games it can be changed freely at any point.
In addition to altering the game's rules, difficulty levels can be used to alter what content is presented to the player. This usually takes the form of adding or removing challenging locations or events, but some games also change their narrative to reward players who play them on higher difficulty levels or end early as punishment for playing on easy. Difficulty selection is not always presented bluntly, particularly in competitive games where all players are affected equally and the standard "easy/hard" terminology no longer applies. Sometimes veiled language is used (Mario Kart offers "CC select"), while at other times there may be an array of granular settings instead of an overarching difficulty option. An alternative approach to difficulty levels is catering to players of all abilities at the same time, a technique that has been called "subjective difficulty". This requires a game to provide multiple solutions or routes, each offering challenges appropriate to players of different skill levels (Super Mario Galaxy, Sonic Generations).
Balancing goals shift dramatically when players are contending with the game's environment and/or non-player characters. Such player versus environment games are usually balanced to tread the fine line of regularly challenging players' abilities without ever producing insurmountable or unfair obstacles. This turns balancing into the management of dramatic structure, generally referred to by game designers as "pacing".
Pacing is also a consideration in competitive games, but the autonomy of players makes it harder to control.
The simplest game balancing technique is giving each player identical resources. Most competitive games feature some level of symmetry; some (such as Pong) are completely symmetric, but those in which players alternate turns (such as chess) can never achieve total symmetry as one player will always have a first-move advantage or disadvantage.
Symmetry is unappealing in games because both sides can and will use any effective strategy simultaneously, or success depends on a very small advantage such as one pawn in chess. An alternative is to offer symmetry with restrictions. Players in Wizard's Quest have the same number of territories, but choose them in alternating order; the differing combination of territories causes asymmetry. Most single-player games are asymmetric; humans get resources that rely on their superior planning ability, while computers get resources that compensate for their lack of intelligence.
The brute force approach to balancing is the mathematical analysis of game session results. With enough data, it is possible to identify unbalanced areas of a game and make corrections.
Randomization of starting conditions is a technique common in board games, card games, and also experimental research, which fights back against the human tendency to optimise patterns in one's favor.
The downside of randomization is that it takes control away from the player, potentially leading to frustration. Methods of overcoming this include giving the player a selection of random results within which they can optimise (Scrabble, Magic: The Gathering) and making each game session short enough to encourage multiple attempts in one play session (Klondike, Strange Adventures in Infinite Space).
Many games become more challenging if the player is successful. For instance, real-time strategy games often feature "upkeep", a resource tax that scales with the number of units under a player's control. Team games which challenge players to invade their opponents' territory (football, capture the flag) have a negative feedback loop by default: the further a player pushes, the more opponents they are likely to face.
Many games also feature positive feedback loops - where success (for example capturing an enemy territory) leads to greater resources or capabilities, and hence greater scope for further successes (for example further conquests or economic investments). The overall dynamic balance of the game will depend on the comparative strength of positive and negative feedback processes, and therefore decreasing the power of positive feedback processes has the same effect as introducing negative feedbackback processes. Positive feedback processes may be limited by making capabilities some concave function of a measure of raw success. For example:
(1) In role playing games using a level structure, the level attained is usually a concave transformation of experience points - as the character becomes more proficient they can defeat more powerful adversaries, and hence can earn more experience points in a given period of playtime - but conversely more experience points are required to 'level up'. In this case, the players level and perhaps also power does not improve exponentially, but approximately linearly in playing time.
(2) In many military strategy games the conquest of new territory only gives a marginal increase in power - for example the 'home province' may be exceptionally productive, whereas new territories open to acquisition might only have by comparison slight resources, or may be prone to revolts or public order penalties which reduce their ability to provide significant net resources, after resources are allocated to adequately suppressing revolts. In this case, a player with initially impressive successes may become 'overextended' attempting to hold may regions which provide only marginal increases in resources.
(3) In many games there is little or no advantage in acquiring a large horde of some particular item. For example, having a large and varied cache of equipment or weapons is an advantage, but only weakly over a somewhat smaller horde with a similar degree of diversity - for example only one weapon can be used at a time, and having another in an inventory with very similar capabilities offers only marginal gain. In more general terms, capabilities may depend on some bottleneck where there is no or only weak positive feedback.
Strongly net negative feedback loops can lead to frequent ties. Conversely, if there is on net a strong positive feedback loop, early successes can multiply very rapidly, leading to the player eventually attaining a commanding position from which losing is almost impossible. See also dynamic game difficulty balancing.
Often dynamic balancing is crudely addressed by game designers by some add-hoc rule, which may cause player annoyance. For example there may be some hard upper limit to power imposed by some arbitrary rule which does not make sense in the context of the rest of the game mechanics. In this case though the player may be playing well and making successes, these are not reflected in any advantage, which may make their efforts seem pointless.
A game can be balanced dynamically by a gamemaster who observes players and adjusts the game in response to their emotional state.
Although gamemasters have historically been humans, some videogames now feature artificial intelligence (AI) systems that perform a similar role by monitoring player ability and inferring emotional state from input. Such systems are often referred to as having dynamic difficulty. One notable example is Left 4 Dead and its sequel Left 4 Dead 2, cooperative games that have the players fight through hordes of zombie-like creatures including unique creatures with special abilities. Both games use an AI Director which not only generates random events but tries to create tension and fear by spawning-in creatures to specific rule sets based on how players are progressing, specifically penalizing players through more difficult challenges for not working together. Research into biofeedback peripherals is set to greatly improve the accuracy of such systems.
A gimp is a character, character class or character ability that is underpowered in the context of the game (e.g., a close range warrior class equipping a full healing boosting armour set, despite having no healing abilities). Gimped characters lack effectiveness compared to other characters at a similar level of experience. A player may gimp a character by assigning skills and abilities that are inappropriate for the character class, or by developing the character inefficiently. However, this is not always the case, as some characters are purposely "gimped" by the game's developers in order to provide an incentive for raising their level, or, conversely, to give the player an early head-start. An example of this is Final Fantasy's Mystic Knight class, which starts out weak, but is able to become the most powerful class if brought to a very high level. Gimps may also be accidental on the part of the developer, and may require a software patch to rebalance.
Sometimes, especially in MMORPGs, gimp is used as a synonym for nerf to describe a rule modification that weakens the affected target. Unlike the connotatively neutral term nerf, gimp in this usage often implies that the rule change unfairly disadvantages the target.
A nerf is a change to a game that reduces the desirability or effectiveness of a particular game element. The term is also used as a verb for the act of making such a change. The first established use of the term was in Ultima Online, as a reference to the Nerf brand of toys whose bullets are soft and less likely to cause serious injury.
Among game developers, MMORPG designers are especially likely to nerf aspects of a game in order to maintain game balance. Occasionally a new feature (such as an item, class, or skill) may be made too powerful, too cheap, or too easily obtained to the extent that it unbalances the game system. This is sometimes due to a method of using or acquiring the object that was not considered by the developers. The frequency and scale of nerfing vary widely from game to game, but almost all massively multiplayer games have engaged in nerfing at some point.
Nerfs in various online games, including Anarchy Online, have spurred in-world protests. Since many items in virtual worlds are sold or traded among players, a nerf may have an outsized impact on the virtual economy. As players respond, the nerf may cause prices to fluctuate before settling down in a different equilibrium. This impact on the economy, along with the original impact of the nerf, can cause large player resentment for even a small change. In particular, in the case of items or abilities which have been nerfed, players can become upset over the perceived wasted efforts in obtaining the now nerfed features. For games where avatars and items represent significant economic value, this may bring up legal issues over the lost value.
A buff is the opposite of a nerf: namely, a change to a game's rules which increases the desirability or effectiveness of a particular element. It likely came from the bodybuilding term of "getting buff" in which the person is taking action to develop their muscles towards the idea of improvement - thus "buffing" themselves.
Overpowered (often abbreviated to OP) is a common term referring to a perceived lack of game balance. It is often used when describing a specific class in an RPG, a specific faction in strategic games, or a specific tactic, ability, weapon or unit in various games. For something to be deemed overpowered, it is either the best choice in a disproportionate number of situations (marginalising other choices) and/or excessively hard to counter by the opponent compared to the effort required to use it.
Underpowered is a common term referring to a perceived lack of game balance, but unlike overpowered, refers to when describing a specific class in an RPG, a specific faction in strategic games, or a specific tactic, ability, weapon or unit in various games as far weaker than average, resulting in it being always one of the worst options to pick in most situations. In such way, it is often marginalized by other choices because it's inherently weaker than similar options or it's much more easily countered by opponents.
This section needs expansion. You can help by adding to it. (June 2018)
A revamp (or rework) is a term for improving or modifying items, skills, abilities, or stats, as opposed to direct nerfing or gimping.