Difference between revisions of "Randomness"

From gdp3
Jump to: navigation, search
Line 158: Line 158:
[[Perfect Information]],
[[Perfect Information]],
[[Performance Uncertainty]],  
[[Performance Uncertainty]],  
[[Player Balance]],
[[Predetermined Story Structures]],   
[[Predetermined Story Structures]],   
[[Predictable Winner]],  
[[Predictable Winner]],  

Latest revision as of 13:22, 6 April 2018

Effects or events in the game cannot be exactly predicted.

Randomness is the process of making effects and events unpredictable in games. It does not necessarily make games totally unpredictable, as Randomness usually has a structure where players can know the chances for certain effects and events. This bounded uncertainty can provide suspense since players may know in advantage which advantageous and disadvantageous results are possible before they happen. Besides the effects of uncertainty, Randomness can create variations in games so that each game instance can be different. This in turn can make games interesting to play several times independent of how much variety comes from the players' actions in the game.


Very few card games do not randomize the cards by shuffling them before giving the players their cards. Not doing so would ruin nearly all games, and especially games that include some form of gambling. Contract Bridge, Texas Hold'em and Blackjack all examples of popular card games where the cards are shuffled before play and Magic: The Gathering and the Pokémon Trading Card Game are similar examples from Collectible Card Games. Card games with other card sets also make use of Randomness, especially games such as Race for the Galaxy and Dominion where the draw stacks need to be reshuffled several times during the games.

Roleplaying games such as Dungeons & Dragons and GURPS use Randomness to determine the outcome of battles and skill tests. Other common uses of Randomness in these games are in the generation of player characters (Dungeons & Dragons and Warhammer Fantasy Roleplay) and random encounters to spice up the game world (Dungeons & Dragons and Mutant) and give the players the impression that there is more to the game world than they experience.

Some computer games where exploration constitutes a major part of gameplay, e.g. such as Minecraft, NetHack, and Diablo, create dungeons using pseudo-random system. By doing so they can be replayed several times with different challenges each time.

The Left 4 Dead series mixes Randomness with decisions by the AI director to vary when players will encounter special types of infected or hordes of common infected. Likewise the location and types of weapons and other equipment found in randomized. Left 4 Dead 2 expands this by also randomizing which paths are open and which are blocked along the route the players need to move.

See the category of Games of Chance on this wiki for additional examples.

Using the pattern

Randomness can be used both before gameplay begins to create Asymmetric Starting Conditions or to affect ongoing gameplay. One example common to Roleplaying Games such as Dungeons & Dragons, Hârnmaster, and Traveller is to include Randomness in the process for constructing Player-Created Characters (but see GURPS and the Storytelling System for examples not using Randomness). In Card Games using Trumps, this may also be randomly determined before gameplay begins. For affecting gameplay as it happens, Randomness can be used to determine what effects an event or action should have, who should be effected by an event or be allowed to choose an outcome, and when an effect should begin and for how long it should persist. Basically all types of actions and events can be modified by Randomness but the common ones include any that involves Skills (including Critical Failures and Critical Successes), Combat (including creating Critical Hits or Critical Misses), Movement (for both Agents and Moveable Tiles), Rewards (including Free Gift Inventories), and Penalties. Randomness can also be used to make the order in which Turn Taking is done unpredictable. The specifics of goals may also affected by Randomness, e.g. to create Varied Gameplay in Quests or related to overcoming Enemies, and for Ephemeral Goals this may include when they appear and how long they exist.

In computers and video games, Randomness is typically generated through pseudo-random sequences that are not random according to scientific definitions but seem random to casual observations. In Self-Facilitated Games, Paper-Rock-Scissors are often used for the purpose of generating Randomness (even though the game itself has national and international competitions[1]) since this requires no additional game elements.

Dedicated game elements used to create Randomness are Dice, Cards and (less commonly) Tiles; in the two latter cases this is often to create randomized Hands through Drawing Stacks or Decks. Bag Building allowed Tokens to be used in a similar fashion. The main difference between Dice and the other options is that of memory and static distribution. The roll of Dice is unaffected by previous rolls, so the randomization process can be seen as having no memory and the chances for any result are exactly the same as they were the previous time. Cards, on the other hand, use outcome like a form of Non-Renewable Resources, so Cards can support Memorizing since players who have done this have better chances of predicting what cards will appear next. The use of Deck Building allows players influence on which Cards will be randomized from in Drawing Stacks or Decks. Tiles can be used to create Randomness in the same fashion as Cards, in this case most typically for Tile-Laying (e.g. Carcassonne or Drakborgen). One option for generating these types of Drawing Stacks is to use Stack Seeding, putting specific Cards or Tiles in specific regions of the Drawing Stacks, since this can help give a form of Predetermined Story Structure to the gameplay. By doing so the game Pandemic distributes the outbreak of diseases evenly throughout the game while keeping the exact moments random, and Thunderstone does the same with Guardians and the Thunderstones themselves. The second version of the board game Drakborgen uses the same mechanic to place the dragon's lair somewhere near the center of the board.

The randomness of a random outcome can be influenced in several ways. Using several Dice or Cards together to generate an outcome allows distributions to have higher likelihood for some results and less for others (i.e. have distributions closer to normal or bell curve distributions). Generating many outcomes but letting players choose one provides Freedom of Choice while at the same time making Perceivable Margins apparent since the distributions are radically shifted depending on the number of outcomes (e.g. Bloodbowl for use of this pattern). Drafting Spreads and other ways of combining Drafting with Freedom of Choice (as in Small World or Ticket to Ride but also present as an optional rule in Race for the Galaxy) are variations of this, and can create Asymmetric Information and Uncertainty of Information among players when the choices or outcomes are not Public Information. How already used cards are reshuffled into the Drawing Stack can also be used to change the distribution of outcomes. This is used in Pandemic, where after an epidemic card is drawn all cards in the Discard Pile are shuffled and placed on the top of the Drawing Stack, increasing the likelihood that diseases will break out in already affected cities.

The above options can be used uniformly through out game sessions or be influenced by the current game state. Letting what happens in the game influence how the randomization is done allows the construction of Positive Feedback Loops or Negative Feedback Loops. Racing games such as Mario Kart and Super Monkey Ball use the latter to create Balancing Effects by giving leading players more passive and less aggressive Pick-Ups while giving trailing players the opposite. Bloodbowl is an example of both, where the number of Dice used is part of the skills of the Characters, which do not change during the game, and their positions in relation to each other, which does.

When games have too much Randomness, the use of Fixed Distributions can regulate this. While Fixed Distributions is one form of Misfortune Mitigation, other forms of Misfortune Mitigation can be used to focus more specifically at removing negative effects of Randomness.

Diegetic Aspects

Rather than being something that may be difficult to describe thematically, Randomness in the real world is easy to translate into a game setting although the outcomes and their likelihoods may not be comparable with the real world. Randomness can also be used as a way of getting around having too fine granularity in a game system when trying to model a world, instead of providing detailed deterministic rule a randomized outcome can simulate this. Examples of real-world concepts that are often modeled in games partially through Randomness include Combat, and especially Damage or the Variable Accuracy real-life aiming has, and the starting skills and conditions for Characters. The use of Skills that represent many actions or actions done in hazardous situations also make use of Randomness to abstract complex situations into a manageable situation in game terms.

On a more concrete plane, Randomness can be used to change the specifics of Game Worlds or Levels by simply making the locations or appearances of game elements different for each game instance - or change the actual worlds if they are Reconfigurable Game Worlds. While this may create Varied Gameplay, randomly created Game Worlds or Levels may cause problems with Player Balance or the possibility of having Predetermined Story Structures, Randomness can go a step further and create fully Procedurally Generated Game Worlds. If done with moderation, Randomness can also be used to provide Unpredictable Behavior to a Game System Player or Agents based upon Algorithmic Agents, and can thereby help make it less easy to deduct the underlying rules. It can also provide variation to Generic Adversaries while still maintaining the commonality among these.

Interface Aspects

The way players generate Randomness in a game can be vital in determining if they feel they have direct influence on events or if they feel that Luck matters most in the game. Although this may vary significantly between players, Dice and Cards are physically handle by the players themselves are may thereby promote player agency more than other means such as computer generated results.


Probably the most basic consequence of using Randomness in games is that it introduces an Uncertainty of Outcome. Randomness has a close but fickle relation to players' perception of how they can influence a game. When players feel that they have some influence on how the randomization is done, e.g. by physically shaking and rolling a die, Randomness can provide them with an Exaggerated Perception of Influence if they believe in Luck and can thereby promote enough illusion of influence to support games with No Direct Player Influence. However, if they do not believe in it the effect might be completely opposite. As long as players feel Exaggerated Perception of Influence, or are emotionally affected by the outcome (e.g. through the use of Betting), Randomness can create Tension but may equally well ruin already existing Tension if it is based upon Gameplay Mastery. Regardlessly, the use of Randomness does work against the influence of players' own performance, i.e. it works against the use of Performance Uncertainty and can lessen feeling of Further Player Improvement Potential. When players have the choice to perform or not perform actions based upon Randomness this gives them the opportunity to take a Leap of Faith. Randomness can support Spectacular Failure Enjoyment when it create extreme or impossible odds but which still take time to play through, and random events that are either very unlikely or have significant impact on the gameplay are likely to be seen as Exceptional Events. Explicit Random Seeds allows players to reuse the same random starting positions several times but doesn't remove Randomness since different players actions from the same starting position quickly leads to very different game states typically.

Depending on how Randomness affects events or actions in a game, its presence can have many different consequences. Randomizing can modify Delayed Effects of events and create Predetermined Story Structures from several different events (potentially Never Ending Stories); it can make Experimenting more difficult but also make it necessary (e.g. testing potions in NetHack). Randomness for which players actions are available can force Varied Gameplay; players of Dominion are for example limits to the actions of the cards they have drawn. When the exact outcome of events that follow players actions is uncertain they may still have a Determinable Chance to Succeed and make a Risk/Reward choice that does not depend on any skill they have, and thereby reduces pressure on their own performance. Using Randomness to create variation in what is required to succeed with Puzzle Solving activities is a way to create Replayability in puzzle-based games.

More generally, Randomness makes it impossible to have Perfect Information about actions and events in the game. This can be true even if players have complete access to the game state (e.g. through Game State Overviews) since future events cannot be calculated. When introduced in games it thereby makes Predictable Consequences more difficult and give players Limited Foresight and Limited Planning Abilities, and can be used to avoid the presence of a Predictable Winner (and thereby also Kingmaker). This can make Strategic Planning difficult as well, although knowing the probabilities involved can still allow Strategic Planning on a higher level of abstraction. This is usually done for three reasons: to generate Asymmetric Resource Distribution, ensure Imperfect Information, or lessen the risk of Analysis Paralysis in the games. By doing so a game can thematically simulate events in the real world that are chaotic and unpredictable but can also may Replayability more likely to be interesting since each game instance will have variations. Given that players may have some information about the Randomness distribution the patterns often supports a bounded case of Uncertainty of Information. This can encourage players to perform Memorizing of the distributions to gain Strategic Knowledge that can be used to make better Risk/Reward choices. In the case of Cards, players may also do Memorizing of the cards played or revealed to be able to know the distribution of the remaining cards. When Randomness is used to add actions to Collaborative Actions, this can make it easier to make the others actions into Anonymous Actions.

Randomness can in itself create an unstable form of Player Balance. It is unstable because all players may have equal chances to receive what is randomized, but as soon as the outcome becomes apparent, players may feel disadvantaged. Similarly, the presence of Randomness typically provides a weak form of Balancing Effects, as the outcome is not affected by players' skills and affects all players impartially. These Balancing Effects become stronger the more Randomness is introduced if the outcome produces Arithmetic Progression rather than Geometric Progression. However, this tends to lessen the chance for Perceivable Margins depending on players' skills in games and makes Gameplay Mastery less noticeable. It can also ruin possibilities for players to feel Value of Effort regarding their actions when the Randomness has bigger impact on outcomes, or when Randomness can tip outcomes from being successful ones to being unsuccessful ones. Randomness can mask Balancing Effects by having separate by overlapping distributions for players depending on their position when providing Penalties or Rewards; one example when this is done is when players pick up Power-Ups where the effect is decided through Randomness.

The use of Randomness can limit the influence of Dedicated Game Facilitators but is usual only interesting if the facilitation is done by other humans. A special case when this is more likely to be interesting is in roleplaying games (such as Dungeons & Dragons or GURPS), since the Game Masters there wield life and death powers over the Player Characters and is in some sense a player also. Randomness can of course due to its unpredictability kill Player Characters when neither Game Masters nor players wish this to occur; Fudged Results offer an escape from this.


Can Instantiate

Asymmetric Starting Conditions, Balancing Effects, Critical Failures, Critical Hits, Critical Misses, Critical Successes, Exaggerated Perception of Influence, Exceptional Events, Experimenting, Imperfect Information, Limited Foresight, Limited Planning Ability, Luck, Memorizing, Player Balance, Procedurally Generated Game Worlds, Replayability, Risk/Reward, Spectacular Failure Enjoyment, Strategic Knowledge, Tension, Uncertainty of Information, Uncertainty of Outcome, Unpredictable Behavior, Variable Accuracy

with Exaggerated Perception of Influence

No Direct Player Influence

with Enemies, Game Worlds, or Levels

Varied Gameplay

with Ephemeral Goals or Quests

Varied Gameplay, Replayability

Can Modulate

Agents, Algorithmic Agents, Asymmetric Resource Distribution, Balancing Effects, Betting, Characters, Collaborative Actions, Combat, Damage, Dedicated Game Facilitators, Delayed Effects, Determinable Chance to Succeed, Ephemeral Goals, Free Gift Inventories, Game Masters, Game System Player, Game Worlds, Generic Adversaries, Hands, Levels, Moveable Tiles, Movement, Never Ending Stories, Penalties, Player-Created Characters, Power-Ups, Predetermined Story Structures, Predictable Consequences, Puzzle Solving, Reconfigurable Game Worlds, Tile-Laying, Trumps, Turn Taking, Quests, Quick Games, Rewards, Skills

Can Be Instantiated By

Bag Building, Cards, Decks, Drawing Stacks, Dice, Paper-Rock-Scissors, Tiles, Pseudo-random number generation

Can Be Modulated By

Deck Building, Dedicated Game Facilitators, Drafting, Drafting Spreads, Explicit Random Seeds, Fixed Distributions, Freedom of Choice, Fudged Results, Game Masters, Misfortune Mitigation, Non-Renewable Resources, Stack Seeding

Possible Closure Effects


Potentially Conflicting With

Analysis Paralysis, Dedicated Game Facilitators, Exaggerated Perception of Influence Experimenting, Further Player Improvement Potential, Game Masters, Gameplay Mastery, Kingmaker, Perceivable Margins, Perfect Information, Performance Uncertainty, Player Balance, Predetermined Story Structures, Predictable Winner, Strategic Planning, Value of Effort


A rewrite of a pattern that was part of the original collection in the book Patterns in Game Design[2].


  1. Wikipedia entry for Paper-Rock-Scissors.
  2. Björk, S. & Holopainen, J. (2004) Patterns in Game Design. Charles River Media. ISBN1-58450-354-8.


Jose Zagal