Introduction, Chapter 1, p.1-18

p.2 The concept of emergence describes the properties, behaviors, and structure that occur at higher levels of a system, which are not present or predictable at lower levels.

p.2 Complex systems are distinguished from systems that are merely "complicated" by the possibility of emergence.

p.3 Emergent gameplay is made possible by defining simple, global rules; behavior; and properties for game objects and their interaction in the game world and with the player. Emergent gameplay occurs when interactions between objects in the game world or the player's actions result in a second order of consequence that was not planned, or perhaps even predicted, by the game developers, yet the game behaves in a rational and acceptable way.

p.5 Interactions in the game world are the foundation of the gameplay and the types of interactions depend on the game genre.

p.6 The key to creating emergent gameplay is to define a simple, general set of elements and rules that can give rise to a wide variety of interesting, challenging behaviors and interactions in varying situations... As with any emergent system, the fundamental set of rules and elements stay constant, but their situation and configuration change over time. The sensitivity of the elements to changing situations and the interaction of the elements with each other and the players are what create emergent gameplay.

[JLJ - Later Sweetser mentions that we also need to create a community of like-minded individuals to form a social world in which to "play" our emergent game.]

p.9 The ideal framework for facilitating emergent agent behavior is to have simple agents in a complex environment. The emergence comes from the interactions between agents, between the agents and the player, and the collective interactions of the agents with the game world. In order to achieve this, the agents must be given a way to sense and model their environment. Some common approaches to sensing game environments are probing, broadcasting, and influence mapping.

p.9 After the agent has sensed its environment and has an understanding of its situation, it must choose an action.

[JLJ - More importantly, an agent must have a scheme for operating in the environment, specifically for each predicament it is likely to encounter. An agent must be able to heuristically generate 'possible' actions, and construct intelligent diagnostic tests to estimate the potential of each posture for sustained action in the predicament, ideally for sustained action that is poised to take advantage of less-than-perfect play by an opponent, and allows for recovery from the unexpected.]

p.15 This book is primarily written for game developers and future game developers who hope to extend their games to enhance the player experience by allowing emergent behavior and gameplay.

p.15 Chapters 2 to 5 provide the fundamental theory of complex systems, game development, and computer science that later chapters will build upon.

Emergence, Chapter 2, p.19-42

p.19 it was Aristotle, over 2000 years ago, who first recognized the profound concept that a whole can be more than the sum of its parts.

[JLJ - Perhaps Aristotle based his thinking on yet older texts that have not survived to the present time.]

p.19-20 This chapter explores complex systems and their incarnation in physics, biology, and society... Emergence is the key to every system, theory, and methodology you will visit

p.20 If something is complex, it is so complicated or intricate that it is hard to understand or deal with... A complex system is a system that consists of many interconnected and interdependent parts. The parts themselves may be simple or complex, but the real complexity comes from their interaction.

p.21 A complex system is more than the sum of its parts, because it is as much about how the parts interact, interconnect, and affect each other as it is about the composition and behavior of the parts themselves. Simply looking at the parts in isolation and inferring the overall behavior of the system as an aggregate of these parts does not give an accurate description of the system.

  The interaction of the parts is so fundamental to the system as a whole that the system cannot be described without describing its parts and the parts cannot be described without describing how they relate to each other. The complex behavior of the system is said to be emergent, it cannot be simply inferred by the behavior of its components.

  The components of complex systems do not just interact, they are interdependent.

p.22 Most complex systems have a common set of properties. These properties include elements, interactions, formation, diversity, environment, and activities.

p.23 Physical systems are made up of the simplest elements - atoms, protons, electrons - obeying the universal laws of physics. Yet from the interactions of these base elements following the fundamental laws of the universe, come the most complex, emergent, and dynamic systems and behaviors

p.25 Life is composed of complex systems as it is made up of very simple elements reacting to their local environments in ways that create self-organized, complex structures with amazingly diverse properties and behaviors.

p.25 An interesting example of complex behavior in biological systems is via chemical signaling... Ant colonies have no hierarchical structure or chain of command... Instead, each ant reacts to stimuli in the form of a chemical scent... and leaves behind a chemical trail that provides stimulus to other ants... Despite the lack of direct organization, ants demonstrate complex behavior... and can solve geometric problems... The local interactions between ants give rise to the observed organizational behavior of an ant colony

p.26 Ants responding to chemical signals appear organized... Economies emerge from a multitude of people trying to satisfy their material needs by individual acts of buying and selling... flocks of birds are groups of individuals adapting to the movement of their neighbors

[JLJ - Economies emerge not only due to 'a multitude of people trying to satisfy their material needs by individual acts of buying and selling,' but also due to the scheming plans of many other individuals attempting to take advantage of patterns, trends, and new opportunities in such economies.] 

p.28 Complexity... is the amount of information that is required to describe a system.

[JLJ - Complexity involves the degree of intelligent effort required to understand a situation or configuration in enough practical detail in order to 'go on.'] 

p.28 Dynamic complexity relates to the computational effort required to describe the state of a system... A system can have a simple structure and yet have complex dynamical behavior... Complexity is a measure of the difficulty involved in understanding a system. For dynamic systems, the description includes how the system changes over time.

[JLJ - ...or how it might change over time. Certain things cannot be predicted exactly, yet we must make a practical attempt to do so, in order to determine how to 'go on.']

p.31 Collectivism identifies that the nonlinear feedback between the micro and macro levels of a system (the components and the whole) is essential to the dynamics of complex systems. A system is more than just a larger entity synthesized from smaller entities. In order to fully understand a complex system, the system must be viewed as a coherent whole, whose evaluation is continuously refined by nonlinear feedback between its macroscopic state and its microscopic components.

p.33 Systems that are poised at the edge of chaos are optimized to evolve, adapt, and develop emergent behaviors.

[JLJ - Why is this so? Perhaps the systems optimized to evolve, adapt, and develop emergent behaviors survive their present predicament and arrive in their next, and so on.]

p.40 Systems that exhibit emergence have a common set of elements (see Table 2.5) and adhere to a common set of rules:

• Global phenomena emerge from local interactions of many simple entities
• There is no evidence of the global phenomena at the local level
• Global phenomena follow a different set of dynamics

Table 2.5 Common Elements of Emergent Systems

   Complex systems are distinguished from systems that are merely "complicated" by the possibility of emergence. Entities in complex systems do not merely coexist, they are interconnected and interdependent.

Playing Games, Chapter 3, p.43-78

p.43 Games are a form of entertainment - developers make them for people to play and enjoy.

p.46 Table 3.2 Five Key Elements of Player Interaction in Games

p.54 Interactive fiction was the first step away from passive media, such as movies and books.

[JLJ - ...yet board games like chess have existed for centuries. Each player presents his opponent with a puzzle as to what to do next, how to create a sustainable pressure which strategically aims to produce good positions in the future, from slight mistakes or inaccuracies in predicting emergent effects from local and global piece interaction. The Avalon Hill board games involving wargaming and replaying historical battles have also been around a while, but the rules for these game have been so complicated as to virtually prevent any but the most diehard from sustaining a continuing interest. Even card games such as bridge, hearts, poker, and Pinochle, are cheap (unless you consider the money you lose playing poker...) and require only a deck of cards. Checkers too. Sweetser needs to re-think this line of thought.]

p.56 The next major step in gameplay involved creating persistent, continuous game worlds that players could walk around and explore.

[JLJ - I think there were many attempts to market game-playing software - the successful ones simply reflected what the game-buying public was most amused by, and the most willing to part with money in order to experience. Chess is virtually free - you just buy a board and pieces to practice with, a few books to read to learn strategy, and there are usually clubs in the area or online to play.]

p.61 Emergent gameplay is made possible by defining simple, global rules, behavior, and properties for game objects and their interaction in the game world and with players. Emergent gameplay occurs when interactions between objects in the game world or a player's actions result in a second order of consequence that was not planned, or perhaps even predicted, by the game developers, yet the game behaves in a rational and acceptable way.

p.61 Global emergent gameplay occurs when the simple low-level rules and properties of game objects interact to create new, high-level gameplay that alters how the game as a whole plays out.

p.63 Emergent games have high replayability

p.64-65 flow theory is based on the premise that the elements of enjoyment are universal, providing a general model that summarizes the concepts common to all when experiencing enjoyment... The general, broad nature of flow theory makes it the ideal construct for a concise model of player enjoyment in games... Flow is an experience "so gratifying that people are willing to do it for its own sake, with little concern for what they will get out of it, even when it is difficult, or dangerous" (Csikszentmihalyi, 1990).

Emergence in Games, Chapter 4, p.79-115

p.80 board games... are the long-standing ancestors of current computer games and much can be learned from their centuries of design, refinement, and play. Many traditional board games, such as chess, checkers, and go, contain perfect examples of emergent gameplay.

p.81-82 Understanding the state of a chess game is not a simple matter of adding the values of the pieces on the board. As with all complex systems, the whole is more than the sum of its parts. The pieces on the board interact to support one another... Chess is a game of emergence as the simple, low-level pieces and rules give rise to complex, high-level behavior and properties that are not present or predictable from the low-level components individually. Understanding the state of a chess game involves describing not only the pieces and their positions, but the emergent patterns and formations of the pieces.

p.83 Interactions in the game world are the foundation of the gameplay and the types of interactions depend on the game genre.... In real-time strategy games, interactions include... attacking, and defending.

p.84 The environment is the central component of an emergent game system as it defines the game world and the interactions that are possible within the world.

p.86 Objects in games are numerous and varied... Each type of game object interacts with the game environment and the player in different ways, which gives rise to interesting possibilities for action for the player

p.89 Situational awareness gives an agent a sense of what is happening in its current environment, what could happen next, what options there are for action and the possible outcomes of those actions. Situational awareness is the foundation for making decisions in complex operational environments.

[JLJ - Yet the next step beyond awareness is diagnosing a capacity to act and react, within a current predicament.]

  Giving an agent an awareness of its environment and a way to sense and model the situation is the most crucial step in creating reactive, dynamic, and emergent behavior.

p.89 The ideal framework for facilitating emergent agent behavior is to have simple agents in a complex environment. The emergence comes from the interactions... Some common approaches to sensing game environments are probing, broadcasting, and influence mapping.

p.90 Probing is quite fast and efficient if there are only a few specific things that the character is seeking.

p.90 In The Sims, the intelligence is embedded in the objects in the environment, known as "Smart Terrain". Each agent has various motivations and needs and each object in the terrain broadcasts how it can satisfy those needs.

p.91 Influence mapping, a technique used in many strategy games, divides the game map into a grid with multiple layers of cells, each of which contains different information about the game world... The values for each cell in each layer are first calculated based on the current state of the game and then the values are propagated to nearby cells, thereby spreading the influence of each cell. This influence propagation gives a more accurate picture of the current strategic situation... Influence mapping provides passive sensing of a continuous environment... allows the agents' situational awareness to evolve as a function of the environment, and gives rise to reactive and emergent behavior... the agent is continuously adapting its behavior to the environment... and its behavior is a function of its environment... the agent is presented with a single value (calculated using the weighted sum to combine all factors) instead of numerous messages being sent to the agent about the environment.

p.92 Influence mapping is commonly used in games that require coordinated group behavior, such as strategy games and sports games.

p.100 Player as creator narrative is emergent. Some, or all, of the story is a product of the player's interactions in the game world, interactions between objects or characters in the game world, and knock-on effects.

p.111 Uncertainty is an important property of emergent systems; it gives rise to new and unexpected behavior.

[JLJ - ...yet uncertainty can be minimized by expertly-constructed diagnostic tests.]

Techniques for Emergence, Chapter 5, p.117-168

p.118 Decision trees are algorithms with a tree-like structure that are used for learning, classification, and decision-making.

p.135 Decision trees are applicable in games where classification or prediction is required.

p.136 Our brains are made up of about 100 billion neurons, each with up to 10,000 connections to other neurons.

p.141 Artificial neural networks are flexible techniques that can be used in a wide variety of applications in games, including environmental scanning and classification, memory, and behavioral control.

p.144 A complex system is a system that consists of many interconnected and interdependent parts. The parts themselves may be simple or complex, but the real complexity comes from their interaction.

Game Worlds, Chapter 6, p.169-232

p.169 Game worlds are the possibility spaces of games... Interactions in the game world are the foundation of the gameplay. The gameplay is made up of how the player uses the basic interactions to solve problems, achieve goals, and advance through the game. The key to creating emergent gameplay is to define a simple, general set of elements and rules that can give rise to a wide variety of interesting, challenging behaviors and interactions in varying situations.

p.169 Game worlds can be divided into two fundamental components - environment and objects.

p.170 The environment is the central component of an emergent game system as it defines the game world and the interactions that are possible within the world.

p.171 An active environment, based on simple interactions between cells, provides a foundation for emergent behavior to occur in game objects and agents, as well as the environment itself.

p.171 strategy games are almost always divided into grids, called influence maps.

p.172 Game worlds only need to approximately model reality for the purposes of entertainment.

[JLJ - Perhaps, but if someone is better entertained by someone else's model of reality, your game and your model might become less popular. 'Approximately' only works when no one else is doing it specifically any better.]

p.201 Objects in games are numerous and varied... Each type of object interacts with the game environment and the players in different ways, which gives rise to interesting possibilities for action for the players

p.227 Players are dissatisfied with the static, unintuitive, and unrealistic worlds in current games and emergence provides the opportunity to enhance player enjoyment with game worlds that allow consistency, freedom, intuitiveness, and realistic physics... Emergent worlds provide far greater potential for interactions and complexity than emergent objects alone.

   Obviously, the behavior of the world alone is not a game, but it defines the potential actions and interactions of the players. The more flexible and reactive the game world, the more opportunities and freedom the players have in interacting and affecting the game world.

Characters and Agents, Chapter 7, p. 233-306

p.234 As agents are able to choose how to react to the environment, they are able to actively change the state of the world in ways that might not have occurred without their intervention... Characters and agents can be used to create emergence in games by being given an awareness of their environment and an ability to react to the changing state of the environment.

p.234 An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators.

[JLJ - An agent is anything capable of perceiving a current and future predicament in which it exists, needs which can be met, and which can use an intelligent, practical strategy to maneuver to meet these needs, ultimately making investments, actions and decisions, and taking various active and passive postures in the environment.]

p.235 Situational awareness gives an agent a sense of what is happening in its current environment, what could happen next, what options there are for action and the possible outcomes of those actions. Situational awareness is the foundation for making decisions in complex operational environments.

[JLJ - Yes, but awareness alone is not sufficient - one needs a capability to adjust or reconfigure a position or posture with potential, and an ability to take advantage of an opportunity, and an active capability to diagnose one's situation, in order for awareness to be of any value.]

  Giving an agent an awareness of its environment and a way to sense and model the situation is the most crucial step in creating reactive, dynamic, and emergent behavior... The ideal framework for facilitating emergent agent behavior is to have simple agents in a complex environment... Some common approaches to sensing game environments are probing, broadcasting, and influence mapping.

[JLJ - There is no better example than the ancient game of chess. The pieces themselves have simple behavior, but emergence develops from their interactions and the net, complex environment.]

p.236-237 There are some games in which the agents sense and react to other agents by actively probing the environment for information... The agents must actively check to determine whether they can sense something at given time intervals, unlike real vision and hearing, which arrive at the senses continuously. Depending on the agents' frequency of probing the environment, it is likely that events and actions will be missed.

Probing is quite fast and efficient if there are only a few specific things that the character is checking.

p.238-239 Influence mapping, a technique used in many strategy games, divides the game map into a grid with multiple layers of cells, each of which contains different information about the game world... The values for each cell in each layer are first calculated based on the current state of the game and then the values are propagated to nearby cells, thereby spreading the influence of each cell

p.239 This influence propagation gives a more accurate picture of the current strategic situation, because it not only shows where the units are and what they are doing, but also what they might do and the areas they potentially influence.

[JLJ - This idea is currently being used in computer chess programs, specially tuned so that the cost of calculating the influence does not impact the effectiveness in implementation.]

  Influence maps can be used for strategic assessment and decision making, because their structure makes it possible to make intelligent inferences about the characteristics of different locations in the environment.

p.240 Influence maps are commonly used in games for strategic, high-level decision making. However, it is also possible to use them for tactical, low-level decision-making, such as individual agents or units reacting to the environment.

p.241 Influence mapping provides a passive sensing of a continuous environment (as opposed to discrete entities), allows the agents' situational awareness to evolve as a function of the environment, and gives rise to reactive and emergent behavior.

p.242 Even if the agent has a sophisticated world model, if it fails to act or react appropriately, it will appear lifeless and unintelligent.

p.243 With enough information about their environment, agents can use a simple set of rules to decide how to act and react appropriately.

p.244 For an agent to react sensibly to the environment in the Active Game World, it is necessary for it to have two things. First, it must have a way to sense the environment and second, it must have a way to choose a suitable reaction, based on what it sensed.

[JLJ - For an agent to react sensibly to the environment, it must adopt a reasonable posture against both the known and the unknown, and possess a repertoire of practical actions from which one is able to sustain the present, follow-on and future predicaments. The agent must be able to imagine realistic scenarios of change in a competitive environment using tricks that work, effectively "diagnosing" the capacity of present and follow-on future positions, maintaining an adaptive capacity to change and reorganize, ultimately taking strategic advantage of opportunities as they emerge.]

p.245 The utility function for the agents in the Active Game World determines how comfortable each cell is for the agents and is therefore called a comfort function.

p.246 If the agents are not comfortable in their current cell then they must locate and move to a more comfortable cell. Each agent reassesses its situation each time step... In choosing a destination, the agents evaluate the comfort values of the cells in a neighborhood of a given size and choose the cell with the lowest comfort value.

p.248 The aim of the first round of tuning was to determine the best neighborhood size, in terms of agent performance and behavior, that agents should evaluate when choosing a destination... Each of the neighborhood sizes that were evaluated in the first round of tuning had various advantages and drawbacks.

p.284 Tactics involve a group or team cooperating and behaving in a coordinated way in order to achieve a group goal, such as securing an area, defeating the enemy, winning a match, or making a successful play.

p.294 With the simple steering behaviors of cohesion, separation, alignment, chase, and chase flag, the two teams can attack, defend, hold, and capture flags, giving rise to interesting gameplay.

p.298 The most effective and robust emergent systems are built on very few, simple, fundamental rules. Remember that the emergence in an agent-based system comes from the complexity created by the interaction of a large number of very simple entities. The more complex these entities are themselves, the more chaotic and confusing the system can become.

p.301 [Craig Reynolds] I was lucky that a simple combination of the three rules [JLJ - that produce a natural appearance of birds moving in a flock formation] produced interesting behavior since in retrospect it has a lot of shortcomings.

p.303 [Craig Reynolds] software design prior to implementation and test is often futile. Certainly for emergent systems, I know of no practical approach other than to repeatedly test and tweak. Even if one wanted to take a principled approach to emergent system design, the science just does not yet exist to support it. There is no underlying theory to describe emergence. Given a set of local rules, there is no reliable formal way to predict what global behavior will emerge from them.

[JLJ - ...yet one can develop a practical, competition-tested scheme for dealing with the consequences and opportunities made available due to emergence. A practical sensitivity to environmental cues that matter, combined with a behavior tree of situations and actions, including a sensitive tuning and adjusting of our sensitivities, is practically enough to inform decision on how to 'go on.' Exactly what will emerge is not critically important - exactly how we position ourselves in the current predicament for such emergence, is.]

Emergent Narrative, Chapter 8, p.307-356

p.307 A game's narrative is the story that is being told, uncovered, or created as the player makes their way through the game.

p.307 People in all cultures teach and learn through storytelling... This chapter identifies three narrative paradigms that can be used in games - player as receiver, player as discoverer, and player as creator of the narrative.

p.343 Awareness of the game world should be fluid and change over time. Events should be given weight by relevance, timing, and importance.

p.354 Gameplay should not be locked within the confines of the narrative and there are more ways to tell a story than through words and cinematics.

Social Emergence, Chapter 9, p.357-410

p.386 Social play is not just gaming; it is open-ended play without goals, progression, or even any potential for winning or losing. People engage in play in virtual worlds and online games in the same way that children in a playground engage in play. People play just for the sake of social interaction, rather than to reach their own personal goals.

[JLJ - In my own philosophical system, a 'reveling in the present,' guided by knowledge from past actions and wise, sensitive premonitions, forms the core guiding or orienting structure for an agent determining how to 'go on' in a lifeworld with needs (but without apparent rules or goals), a predicament of sorts in which we continuously find ourselves, and a ticking clock which threatens to degrade and ultimately take away from us, at a distant point in our future, whatever it is that we are and accumulate in the here and now. You can be rational in your life if that is what you choose to do. You only have so many heartbeats and so many breaths of air that you will be able to draw. Certain predicaments we are in call for a rational approach. In others, the situation will allow us more freedom, and in these situations and circumstances I prefer instead to revel.]

Conclusion, Chapter 10, p.411-420

p.413 Chapter 7 discussed the design, implementation, and tuning of reactive game agents in the Active Game World. These agents use influence maps, in conjunction with the cellular automata in the Active Game World, to dynamically respond to the environment.

p.414 In creating emergence, developers put the onus on the environment, making the information readily available for the agents to sense, through probing, broadcasting, influence mapping, or other methods. The more complex the environment and the more information is synthesized for an agent, the simpler the agent itself can afford to be. The emergence comes from many simple interactions with other agents and a complex world.