• a Chunk
    a Chunk

    The Science of Level Design: Design Patterns and Analysis of Player Behavior in First-Person Shooter Levels - Kenneth M. Hullett

     

    This article is a portion of a dissertation by Kenneth Hullet. The source writing is over 250 pages in length.  We share this in hopes that it will further the learning of level designers.  The Table of Contents is listed directly below, within a spoiler.  The parts marked in Orange are included here.  Follow the link at the end of this article to read the full writing, as it contains much value.

     

     

    TABLE OF CONTENTS - In Spoiler

    Spoiler

     

    Table of Contents

    List of Figures

    List of Tables

    ABSTRACT

    Dedication

    Acknowledgments

     

    Chapter 1: Introduction

    1 The First-Person Shooter Genre

    2 Level Design

    3 Design Patterns

    4 Research Questions

     

    Chapter 2: Related Work

    1 Design Patterns

    2 Level Design

    2.1 Pacing

    2.2 Tension

    2.3 Challenge

    2.4 Segmentation

    2.5 Spatial Configurations

    2.6 Navigation

    3 Player Behavior

     

    Chapter 3: Design Patterns

    1 Single Player First-Person Shooter Levels

    1.1 Patterns for Positional Advantage

    1.1.1 Sniper Location

    1.1.2 Gallery

    1.1.3 Choke Point

    1.2 Patterns for Large-scale Combat

    1.2.1 Arena

    1.2.2 Stronghold

    1.3 Patterns for Alternate Gameplay

    1.3.1 Turret

    1.3.2 Vehicle Section

    1.4 Patterns for Alternate Routes

    1.4.1 Split Level

    1.4.2 Hidden Area

    1.4.3 Flanking Route

    1.5 Level Analysis Example

    1.5.1 Discussion

    2 Combat Non-Player Characters in First-Person Shooter Games

    2.1 Elements of a Non-Player Character
    2.2 Pattern Collection

    2.2.1 Soldier

    2.2.2 Aggressor

    2.2.3 Carrier

    2.2.4 Tank

    2.3 Example Analysis

    3 Weapons in First-Person Shooter Games

    3.1 Aspects of Weapon Patterns

    3.2 Pattern Collection

    3.2.1 Sniping Weapon

    3.2.2 Close Blast

    3.2.3 Assault Weapon

    3.2.4 Projectile

    3.2.5 Power Weapon

    3.2.6 Melee Weapon

    3.2.7 Placed Weapon

    3.3 Effects of Weapon Patterns on Level Design

     

    Chapter 4: Understanding Player Behavior

    1 Sources of Data

    1.1 Internal Testing

    1.1.1 Developers

    1.1.2 Quality Assurance

    1.2 External Testing

    1.2.1 Usability Testing

    1.2.2 Beta Tests

    1.2.3 Long-term Play Data

    1.3 Subjective Evaluations

    1.3.1 Surveys

    1.3.2 Reviews

    1.3.3 Online Communities

    1.3.4 Post Mortems

     

    Chapter 5: User Tests

    1 Levels

    1.1 Elements of Test Levels

    1.1.1 Enemy NPCs

    1.1.2 Weapons

    1.1.3 Items

    2 Data Collection

    2.1 Telemetry

    2.2 Video Capture

    3 Pilot Study

    3.1 Lessons Learned

    4 Full Study

     

    Chapter 6: Results 

    1 Key Metrics

    2 Data Analysis

    2.1 Sniper Location

    2.1.1 SniperLocation-1

    2.1.2 SniperLocation-2

    2.1.3 SniperLocation-3

    2.1.4 SniperLocation-4

    2.1.5 SniperLocation-5

    2.2 Arena

    2.2.1 Arena-1

    2.2.2 Arena-2

    2.2.3 Arena-3

    2.3 Choke Point

    2.3.1 ChokePoint-1

    2.3.2 ChokePoint-2

    2.3.3 ChokePoint-3

    2.4 Stronghold

    2.4.1 Stronghold-1

    2.4.2 Stronghold-2

    2.4.3 Stronghold-3

     

    Chapter 7: Discussion

    1 Research Questions

    2 Differences Between Observed and Intended Effects

    2.1 Sniper Location

    2.2 Arena

    2.3 Choke Point

    2.4 Stronghold

    2.5 Summary

    3 Threats to Validity

    4 Future Work

     

    Appendix A: Additional Design Pattern Collections

    1 Multiplayer First-Person Shooter Levels

    1.1 Key Concepts

    1.1.1 Conflict Points

    1.1.2 Player Respawn

    1.1.3 Level Balancing

    1.2 Patterns in Multiplayer Game Types

    1.2.1 Location-based Control (Control Point, Domination, King of the Hill)

    1.2.2 Capture the Flag (CTF)

    1.2.3 Deathmatch (Arena, Slayer)

    1.2.4 Delivery (Demolition)

    1.3 Patterns in Multiplayer Levels

    1.3.1 Arena

    1.3.2 Vehicles and Vehicle-oriented Geometry

    1.3.3 Alternate Routes

    1.3.4 Battlements and Cover

    1.3.5 Items and pickups

    2 Open World Games

    2.1 Mission State Patterns

    2.1.1 Chase/Evade

    2.1.2 Escort

    2.1.3 Acquisition

    2.1.4 Elimination

    2.1.5 Following

    2.2 Non-Mission State Patterns

    2.2.1 Populated Areas

    2.2.2 Interstitial Areas

    2.2.3 Side Objective

    2.2.4 Collection

    2.2.5 Marker

    2.3 Discussion

     

    Appendix B: Data Analysis in Project Gotham Racing 4

    1 Dataset

    1.1 Features

    1.2 Subdivisions

    1.3 Subsets

    2 Analysis and Results

    2.1 Regular players play more multiplayer

    2.2 Regular players play more career mode

    2.3 Many options were underused

    2.3.1 Game Modes

    2.3.2 Event Types

    2.3.3 Routes

    2.3.4 Vehicles

    2.3.5 Vehicle Classes

    2.4 A- & F-Class vehicles most popular

    2.5 C-Class & B-Class equally popular

    3 Recommendations

    3.1 New players should be encouraged to play career mode

    3.2 New players should be encouraged to use F-Class vehicles in multiplayer

    3.3 Development time and costs could be reduced by having fewer available options

    3.3.1 Reduce number of vehicle classes from 7 to 5

     

    References: List of Cited Works and Games

     

     

     

    ABSTRACT

    Level designers create gameplay through geometry, AI scripting, and item placement. There is little formal understanding of this process, but rather a large body of design lore and rules of thumb. As a result, there is no accepted common language for describing the building blocks of level design and the gameplay they create. This dissertation presents a set of level design patterns for first-person shooter (FPS) games, providing cause-effect relationships between level design patterns and gameplay. These relationships are explored through analysis of data gathered in an extensive user study.

     

    This work is the first scientific study of level design, laying the foundation for further work in this area. Data driven approaches to understand gameplay have been attempted in the past, but this work takes it to a new level by showing specific cause-effect relationships between the design of the level and player behavior.

     

    The result of this dissertation is a resource for designers to help them understand how they are creating gameplay through their art. The pattern collection allows them to explore design space more fully and create richer and more varied experiences.

     

     

    INTRODUCTION

    Level designers create gameplay through geometry, AI scripting, and item placement. There is little formal understanding of this process, but rather a large body of design lore and rules of thumb. As a result, there is no accepted common language for describing the building blocks of level design and the gameplay they create. This research creates a science of level design based on design patterns for first-person shooter (FPS) levels and data analysis to show cause-effect relationships between level design patterns and gameplay.

     

    Level design is often viewed as an artistic endeavor, so the applicability of purely scientific approach may be considered controversial. This research argues that level designers employ design patterns while creating FPS levels, whether advertently or inadvertently. Furthermore, analysis of gameplay data can show distinct patterns of behavior in different situations. If we control for all factors besides the design of the level, we can claim that significant observed differences are due to the level design.

     

    To show these cause-effect relationships, we conducted a user study and performed analyses of the collected data. The user study explores what effects the patterns, and variations within the patterns, have on players’ in-game behavior. Based on deviations from the expected results, we are able to adjust the theory, improving our understanding of the relationships, and increasing the usefulness of the taxonomy as a tool for level designers. For each pattern explored in depth, we created multiple instances of the pattern, each with a different set of affordances – for example, with a sniper location, some instances were high, some low, some with good cover, some without, etc. Based on our surveys of existing FPS level design, we expect a lower sniper location to have less of an effect on the level’s pacing; we should observe less of an effect than we would when subjects encounter a higher sniper location. These instances are placed in the user test levels played by the subjects. From the data collected during the user study we can determine how gameplay is affected by the pattern, and if this is different from what we expect.

     

    This research is necessarily reductionist in its approach. In practice, design patterns are rarely distinct, instead overlapping with other patterns or elements to create varied effects. Nonetheless we will argue that design patterns provide a useful analytic framework for thinking about level design in a scientific way. The lowest possible segmentation of level design elements, the actual placement of individual walls, floors, items, and entities, is far too granular to elicit any understanding of designer intent or to observe an effect on player behavior. The highest level, a complete level, is far too coarse, as FPS levels generally contain multiple subareas with different gameplay objectives. Design patterns are a small enough unit that a clear distinct purpose can be elicited, but not so small as to be overwhelmed with details of pixel by pixel placement of objects and geometry.

     

    THE FIRST-PERSON SHOOTER GENRE

    FPS games are combat-oriented games where the player engages other characters with a variety of projectile and melee weapons. The player navigates a 3D world while looking through the eyes of the main character (i.e., a first-person point of view), though some games where the camera follows the player character (third person shooter or TPS) have similar gameplay and are generally considered to be in the same genre.

     

    FPS games are one of the most popular genres of commercial digital games, with many published titles on multiple platforms. Seven of the top-ten all-time best-selling games for the Xbox 360 are FPS games. Due to the processing power needed to render realistic-looking 3D environments, FPS games are often credited as a driving force behind technological advancement in personal computers and gaming consoles.

     

    Beyond entertainment, FPS games have been used for a variety of training and other serious game applications. One of the most notable is America’s Army, a training and recruitment game released by the US Army. Its intent is to provide a realistic simulation to familiarize recruits with modern Army combat procedures. The platform has been used as the basis for more advanced Army training programs.

     

    As a popular and broadly relevant genre, any research that improves our understanding of FPS games is likely to have significant impact. There is also a large body of in-depth analysis which can be drawn upon, including books and articles on FPS design in general and level design specifically. While the results of this study are specific to FPS games, the techniques we propose are generalizable to other game genres.

     

    FPS games are also a desirable genre for this study as the level design is a major component of the game and has a significant impact on the player's experience. Levels in commercial games are designed largely by hand and play tested extensively by designers to create specific gameplay effects. It would be difficult to conduct research of this nature on a genre of games where the level design was not as impactful. Furthermore, while the player's experience is by the level design, the mechanics of the game allow for enough variation in individual choice that these impacts are apparent.

     

    For this research, we have chosen to focus on single-player levels, though multiplayer is increasingly becoming the dominant gameplay mode. In multiplayer, players are generally playing against other players, rather than environmental challenges created by the designer. For this reason, it would be more difficult to conduct an experiment like the one described here for multiplayer levels. However, it is likely that level design does have an impact on gameplay in multiplayer FPS. Early exploration of patterns specific to multiplayer level design is described in Appendix A.

     

    LEVEL DESIGN

    The precise definition of a level varies by game and genre, but it is generally thought of as a subdivision of a game. Specifically, it is a space where gameplay occurs. While the mechanics of the game define the choices available to the player, the design of a level defines what the player experiences at any given point. It is through level design that level designers craft gameplay experiences for players.

     

    Levels for FPS games are generally designed for single- or multi-player play, but not both. Single player levels tend to be a linear sequence of challenges the player must overcome to reach the final goal, whereas multi-player levels are designed to create areas for player-vs.-player combat to occur.

     

    While level geometry is the most noticeable aspect of the level designer’s work, other considerations are important in the creation of gameplay. Level designers place objects in the world, including weapons, ammunition, and power-ups. They must be sure to provide enough so the player can complete the level, but not so much as to remove all challenge. They also place Non-Player Characters (NPCs), both friendly and enemy, and use AI scripting to control their behavior.

     

    When designing an FPS level, there are many factors the designer must consider, including challenge, pacing, and ease of navigation. Though many FPS games have been made, and numerous books have been written on the subject, there is little formal understanding of their level design. The existing literature conveys design lore and industry practice without exploring how levels create gameplay.

     

    Experienced level designers draw from their extensive knowledge of existing games when they create a level. They have an intuitive feel for what features they should include in a level to create different types of gameplay. They may imitate and adapt elements they’ve observed in other levels. Presently, there is no structured way for experienced designers to pass on this knowledge to less experienced designers. A more formal framework would improve designers’ abilities to communicate design ideas as well as provide a reference for possible features to incorporate into levels.

     

    For example, one of the design patterns identified is a sniper location. This is an elevated position from which a character can engage other characters in relative safety. There are numerous variations on sniper locations, including their height, amount of cover available, and whether it is intended for use by either the player or an enemy NPC. The effect of an enemy NPC-occupied sniper location is to slow the pace of the level – the player must move slowly and be more cautious to avoid taking fire. While we can predict this behavior based on our understanding of FPS gameplay, it is unknown if the effect is consistent in all cases, or how it is affected by variation within the pattern. Would the effect be less if the sniper location was lower, as it would be easier for the player to engage the enemy NPC? User tests where a number of subjects play levels with different instances of sniper locations will provide empirical evidence of these relationships.

     

    The taxonomy of design patterns is a useful tool for improving designers’ abilities to communicate design ideas and as a reference for possible features to incorporate into levels. However, the process by which it was created is necessarily subjective. Designers’ intentions in using certain features may vary, and how players react to the patterns may vary.

     

    DESIGN PATTERNS

    As described above, our user studies are focused on single player levels. While we have explored design patterns in both multi- and single player levels, level design necessarily has a greater impact on single player gameplay, as the players' only interaction is with the environment, rather than with other players. As such, this research is primarily focused on the design patterns developed from analysis of single player levels. The patterns are described in terms of their intended use, effect on gameplay, and variations within the pattern. Examples from popular commercial games are given.

     

    The use of design patterns to describe levels is inspired by design patterns used in both software engineering and architecture (the latter of which also inspired the former). A set of design patterns form a language for describing design practices in the domain. Duffy et al. characterize patterns in software engineering by the following:

    • “Noticing and naming the common problems in a field of interest,
    • Describing the key characteristics of effective solutions for meeting some stated goal,
    • Helping the designer move from problem to problem in a logical way, and
    • Allowing for many different paths through the design process.”

     

    This research adapts these characteristics to the domain of level design in FPS games. For level designers the problem is creating an entertaining and engaging experience for the player, and the solution is in how they design the level. We adapt the above to define characteristics of a pattern language for the domain of level design, described in detail below:

     

    • Noticing and naming common structures that produce specific types of gameplay

    The taxonomy presented in this dissertation was created by identifying design patterns in levels and the gameplay they produce. Examining existing levels and inferring the intended gameplay is the most common means of identifying design patterns, but other methods were employed, including interviewing designers about how they design to elicit certain types of gameplay and reading books and articles that describe common practices.

     

    • Describing the key characteristics of these structures and how they affect gameplay

    In identifying the patterns, we noticed that significant variations exist within any given pattern, and those variations have an impact on the gameplay produced. As examples of patterns are identified, variations and their effects are noted, resulting in a more complete detailed view of the pattern and its parameters.

     

    • Helping the designer address level design concerns in a logical way

    Armed with knowledge of level design patterns, the designer can tailor a level to the desired gameplay. For example, if a designer wants to change the pace of a level, they can add or alter instances of patterns that are known to affect pacing. If, during gameplay tuning, they discover a problem in a level, they can use the taxonomy to modify existing patterns to address the issues.

     

    1. Allowing for different approaches to create the desired gameplay

    The taxonomy identifies different design patterns that will affect gameplay in similar ways. If the designer wants to create a certain type of gameplay, they can identify multiple elements in the taxonomy that would be suitable, and pick one that is appropriate for that instance. They are not limited to repeatedly using the same patterns in the same ways; they can use different patterns, or variations with patterns.

     

    RESEARCH QUESTIONS

    The goal of this research is to use data analysis to develop the science of level design through a deeper understanding of FPS level design and how it creates gameplay. The research questions can be broken down into questions about design patterns, player behavior, and the applicability of the work.

     

    • RQ1: Are level design patterns useful for developing levels, communicating ideas, and teaching about level design?

    We have already identified level design patterns to create a language for describing levels. The application of design patterns to FPS levels and the patterns themselves are described in Chapter 3. These descriptions provide insight into the designer’s intent and the gameplay that will result.

     

    It should be possible to take an existing level and describe it extensively in terms of design patterns. We give an example with a level from Bioshock, a popular commercial FPS. Such description often reveals sections of a level that are not describable with the existing taxonomy, leading to the elicitation of a previously undescribed pattern. Through study of FPS levels we can improve and expand the pattern collection.

     

    Besides expanding the pattern collection, it is important to validate the effects of the patterns. The results of this study have helped close the loop and improve the descriptions of the patterns and their gameplay effects. The end result of the study is a set of patterns that has been shown to create specific behavior in the player.

     

    • RQ2: Can we use data analysis to understand player behavior in FPS levels?

    To test the cause-effect relationship of the patterns and their variants on gameplay, it is necessary to understand player behavior. What exactly does it mean, for example, when the tension of a level is increased? How is this reflected by the player’s in-game actions? Can this be observed and reported? While previous user studies provide some guidance, it was necessary to develop methods for identifying and classifying player behavior.

     

    How this was done in this research is described in Chapter 5. Subjects’ in-game behavior was studied in the video recordings of their level play-throughs and the logged gameplay data. This was correlated with the pattern variants that the subjects encounter to see what the effects are.

     

    • RQ3: Do the identified design patterns and their variants create the intended gameplay effects?

    Patterns are used in levels to affect gameplay – for example, when a player encounters a choke point where they have an advantage over enemy NPCs, the expectation is for increased pace and reduced challenge. This should be reflected in the player’s behavior by traits such as engaging enemy NPCs more aggressively, using weapons more frequently, making less use of cover, and moving more quickly. In validating these relationships, we are developing the science of level design. Chapter 5 describes the user study we ran to explore these cause-effect relationships and Chapter 6 explains the results of the analysis.

     

    If the expected behavior occurs when a player encounters a design pattern variant in a level, then the theory is validated. In the example above, when the player encounters the choke point, their behavior should be close to our expectations. If for some variation of the choke point, they instead begin moving more slowly and playing cautiously, then there is something about that instance that is creating different gameplay. We can identify what affordances of the pattern vary from other instances and adapt the pattern description to match the observed results.

     

    To fully explain the impact of this research, this document is broken into multiple chapters. Chapter 2 covers related work in the existing literature on level design and data analysis in games. Chapter 3 presents the taxonomy of design patterns that we have developed for this research. Chapter 4 explains the major sources of data in games and their impact on game development. Chapter 5 describes the user tests performed, and Chapter 6 details the results. Chapter 7 summarizes the findings and the contributions of the research.

     

     

    RELATED WORK

    There are three broad streams of work related to this research. First, previous work on applying design patterns to games in general and level design specifically. Second, previous work on exploring, understanding, and communicating about level design in general, mostly from an industry perspective. Third, previous work on understanding player behavior and how data analysis can be used to identify such behavior. These three areas are described below.

     

    DESIGN PATTERNS

    The use of design patterns to better understand levels is inspired by their use in software engineering, which were in turn inspired by design patterns in architecture. Kreimeier was among the first to adapt the concept of design patterns to the domain of digital games by identifying game design patterns.

     

    Björk et al. extend this work by studying how players interact with games and how entities in a game interact with each other. They identify over 200 patterns in game design ranging from the basic building blocks of games, such as the game world, to abstract concepts like player collaboration and immersion. The patterns are organized in broad categories such as “Patterns for Goals” and “Patterns for Narrative Structure.” Patterns are described in terms of how they are used, the choices a designer must make when using them, their consequences and relationships to other patterns. These patterns do not specifically deal with level design, but do relate to some level design concerns, such as balancing, goals, locations, and objects.

     

    For example, one pattern identified by Björk et al. is Pick-ups, described as “elements that exist in the game world and can be collected by players.” They go on to describe how pick-ups are used in a variety of games and the considerations a designer must make when choosing whether to include them or not. They describe general consequences of pick-ups, but they do not describe the immediate effects they have on a player’s behavior or the flow of a game. The level design patterns presented in this dissertation address these considerations.

     

    Björk et al. suggest four ways patterns can be used to support game design: idea generation, structured development, solving design problems, and communication. The level design patterns identified in this dissertation support these same uses.

     

    Another application of design patterns to games is Plass et al.’s study of educational games. They identify common patterns in educational games that increase enjoyment and engagement in players. These are high-level conceptual goals for designers to pursue, not patterns of mechanics as in Björk et al.’s work, or patterns of level design as presented in this dissertation. Examples include “Constructing things is fun and helps learning” and “Time and resource constraints make games fun and can improve learning.” These patterns were discovered through observational studies and interviews with children playing educational games.

     

    LEVEL DESIGN

    There are many books on level design written from an industry perspective. They discuss common practices and provide instruction on tools for aspiring level designers. In his book, Co takes the reader through the process of designing an FPS level, from brainstorming initial ideas, building the level using Unreal Editor, to testing and improving the level [6]. While useful references, neither this work, nor similar books by Bryne, Clayton, or Feil et al. present deep analysis of how level design creates gameplay.

     

    For example, Feil et al. describe the importance of overall pacing in a level. They discuss how a rhythm of rising and falling tension can contribute to the overall flow of a level without providing methods for creating these effects. Similarly, they discuss strategic considerations of terrain, such as access and height advantage, but do not discuss how they create gameplay. In contrast, the work presented in this dissertation provides specific, concrete idioms of level design described in terms of their direct impact on gameplay.

     

    Several shorter works examine single aspects of level design, from both academic and industry perspectives. The aspects examined can be broadly categorized as relating to gameplay – pacing, tension, and challenge – or space – spatial configurations and how the player navigates.

     

    PACING

    Pacing is the density of actions taken by the player in a level. Coulianos proposes methods to analyze and improve level pacing. Designers can plot the expected pacing as a sequence of gameplay elements. Playtesting can then be used to see how closely the player’s experience matches the designer’s expectations, leading to a series of iterative changes until the designer is satisfied.

     

    Davies also explores aspects of level pacing and suggests techniques designers can use to control pacing. For example, the player’s impetus to move is a key aspect of game pace, which the designer may want to increase or decrease. Movement impetus can be increased by elements such as a time limit or a threat from behind, or decreased by an obstacle or NPC interaction.

     

    TENSION

    Tension is the mental strain a game can create in the player as they struggle to survive or complete objectives. Level designers use tension to affect pacing. For example, NPCs can create tension by urging the player to move through the level faster. Its use is examined in depth by Wright, who conducted a study with subjects playing one of three levels that used NPCs to create tension differently. Completion times as well as the subject’s subjective impressions were compared to evaluate the methods. He found that urgency imparted from a friendly NPC was the most effective method, while chasing or being chased by enemy NPCs were less effective.

     

    CHALLENGE

    In his study of what makes games fun, Malone identified three main elements: challenge, fantasy, and curiosity [18]. All three of these are useful to level designers, but challenge is the most critical. Malone found the best way to create challenge is to provide clear goals whose attainment is uncertain. If the goal is unclear, the player will become frustrated. If the goal is too easy to attain, the player will become bored. Furthermore, if the goal is long range, there should be feedback given to the player that communicates progress towards the goal.

     

    SEGMENTATION

    Segmentation is a broad concept that can be applied to the examination of levels both in terms of gameplay and space. It refers to methods for breaking down aspects of the game into smaller elements. Zagal et al. describe three types of segmentation: temporal, spatial, and challenge. Temporal segmentation is closely related to pacing, as increasing or decreasing the length of time allowed for gameplay can affect tension and challenge. In terms of spatial segmentation, levels themselves are a form of this, but they can be segmented internally as well. As a player moves into a distinct section of a level, their behavior may be affected. For example, moving into a large arena with enemy NPCs will increase tension and difficulty. The third type of segmentation, challenge, also relates to pacing. Breaking up the challenges presented to the player allows the designer to control the level pace.

     

    SPATIAL CONFIGURATIONS

    Within spatial segmentations, the configuration of the environment is also a key concept in level design. Chen et al. compares level design to the architectural design that is used in real world buildings. When designing a building, the architect includes architectural devices to create specific effects, such as customizing a space to a particular use. The authors identify some architectural principles that level designers can apply to create spaces for gameplay, including having a clear path through the level, how to use different spatial organizations such as linear or hub and spoke, or including unique elements to break up the design.

     

    An examination of how space is used in team-based multi-player FPS levels was presented by Güttler et al. They identified common spatial configurations and how they contribute to gameplay. The key elements they studied are collision points and tactical choice. In a team multi-player level, the designer provides multiple routes through the level, allowing players the chance to make a strategic decision. The choice of route determines where in the level the two teams will eventually clash; these collision points are the major contested spaces where the game is played.

     

    There are some significant empirical studies that evaluate the effects of level design on gameplay. Gee studied the use of dead-ends in FPS levels. He identified ways in which dead ends are used and built example levels that included them or not. Subjects were observed playing levels and their preferences and playing time were reported. Results indicated that dead ends did not negatively impact FPS levels.

     

    An empirical study by Gonzales explored directional choices in FPS levels. Similar to the Gee study, they identified different techniques for presenting alternate routes and performed user studies on a set of representative levels. Survey responses and subject observations contributed to their conclusion that choice improves player immersion, as the lack of choice in a linear level can break the illusion of being in large, dynamic world.

     

    NAVIGATION

    A key use of spatial configuration in levels is in providing navigational cues to the player. This is particularly true in FPS levels as they are generally large, complex environments. Nerurkar examines some means level designers use to aid player navigation. Some, such as maps and navigation markers, are separate from the level design, but many are a function of the level design. Examples include features that attract the player’s attention, use of light and contrast, and directions from NPCs. Hoeg performed an empirical study of player navigation and player types in FPS levels. He identified elements that designers use to influence pathing decisions, including lighting, sound, and resistance, and formed a theory about how Bartle’s player types would react in each case. He constructed a level with multiple decision points, using different navigation cues. Subjects’ player types were determined by a survey, and their routing choices were recorded while playing the level. The results were compared to see if the theory was consistent with the player’s behavior. They found that some elements, such as placement of doors and motion, had strong correlation, whereas other factors had weak or no correlation.

     

     

    DESIGN PATTERNS

    While our user study is primarily focused on the effects of design patterns in single player levels, we explored design patterns in multiple aspects of FPS games. Of particular relevance are the patterns for combat NPCs and for weapons. Weapon and NPC design in FPS games fall into a grey area between game design and level design. While they are aspects of the game mechanics, and therefore game design elements, they are greatly influenced by the work of the level designer. Tuning of weapons and NPCs generally occurs late in the development process, and is a function of the constructed levels. As the final tuning of these elements are dependent upon their placement and use by level designers, they can be considered an aspect of level design. As such, patterns for these elements are described here along with the single player patterns. Other pattern collections are presented in Appendix A.

     

    SINGLE PLAYER FIRST-PERSON SHOOTER LEVELS

    The descriptions of the patterns explain how they can be used, the concerns designers must address, and the gameplay created. The fields are listed below:

    Description – A high level description of the pattern and the major design considerations.

    Affordances – Aspects of the pattern that can be varied by the designer.

    Consequences – A description of the gameplay the pattern creates.

    Relationships – Some examples from popular commercial games that illustrate the pattern.

     

    The use of the term "affordances" in this research is a bit idiosyncratic. In the field of design, the word typically means "the perceived or actual properties of the thing, primarily those fundamental properties that determine just how the thing could possibly be used." For example, the presence of a doorknob is an affordance that signals that a door may be opened. For this research we modify this definition slightly, so affordances are aspects of a pattern that can be varied by the designer ("perceived or actual properties") to alter the effect on gameplay ("how the thing could possibly be used"). Essentially, affordances are the knobs a designer can twist within a pattern to dial in different gameplay effects.

     

    The patterns are grouped into one of four following categories based on the type of gameplay produced. The categories are Positional Advantage, Large-scale Combat, Alternate Gameplay, and Alternate Routes. These distinctions are not mutually exclusive, a pattern might be perceived as being in one category or another based on its affordances. Furthermore, specific patterns may overlap, resulting in different effects and described in the relationships sections of each pattern.

     

    Positional Advantage – Spaces where one entity has an advantage over another.

    • Sniper Location – A protected, elevated location that overlooks some portion of the level.
    • Gallery – An elevated area parallel and adjacent to a narrow passageway.
    • Choke Point – A narrow area with no alternate routes, causing entities to be exposed to engagement as they move through.

     

    Large-scale Combat – Areas designed to facilitate combat involving large numbers of entities.

    • Arena – An open area or wide corridor.
    • Stronghold – A confined area with good cover and limited access points.

     

    Alternate Gameplay – Introduce new elements that break from the established mechanics of the game.

    • Turret – An area with a high-powered weapon where one side has a clear advantage.
    • Vehicle Section – Sections of alternate gameplay where the player drives or rides in a vehicle.

     

    Alternate Routes – Create alternatives for the player in how they approach the level.

    • Split Level – A corridor with an upper and lower section, where those on the upper section can attack those on the lower section.
    • Hidden Area – A small area off the main route that contains items for the player.
    • Flanking Route – An alternate path that allows characters to gain positional advantage.

     

     

    PATTERNS FOR POSITIONAL ADVANTAGE

    These patterns all result in one entity gaining an advantage in position over another entity. A positional advantage usually affords opportunities to attack other entities without being exposed to counter attack.

     

    SNIPER LOCATION

    Description:

    Sniper locations are one of the most common patterns. A character in a sniper location can attack other characters with long-range weapons while remaining protected. Any elevated position that overlooks some portion of the level is potentially a sniper location. They may be intended for use by either players, NPCs, or both.

     

    Creating a sniper location for use by an enemy rather than the player requires additional consideration. Enemies positioned in the sniper location may require special scripting to create the desired behavior; they should remain in place, using cover if available, and engage the player with long range weapons.

     

    Affordances:

    • The height of the sniper location over the main part of the level
    • How large of an area is available for the sniper
    • The amount of cover available for the sniper
    • The size of the area that the sniper can cover from the sniper location
    • How accessible the sniper location is from the area overlooked

     

    Consequences:

    When confronted with an enemy sniper location, the player is forced to make careful use of cover or seek alternate routes to avoid being exposed to fire. This can increase the tension and slow the pace of a level while creating a challenge for the player.

     

    A player sniper location generally slows the pace of a level while lowering tension as the player is able to engage enemy NPCs without being exposed to enemy fire. However, if the sniper location is not isolated from the rest of the level, the player will have to defend the access point as well, increasing tension.

     

    Relationships:

    Sniper locations interact with many other patterns. They may be placed to cover an arena or a choke point. Most stationary turrets are also sniper locations. A shooting gallery is specialized type of sniper location. A sniper location with access may be a type of stronghold.

     

    Examples:

    In the level “Route Kanal” of Half-Life 2, the player encounters an enemy sniper location, shown in Figure 1. It is high above the player’s position, but has very little cover. The player can engage the enemy NPCs, but is exposed and needs to be cautious.

     

    a.png.a6a595b43cfd9c9efce0b1a9ffc02569.png

    Figure 1: Sniper location in Half-Life 2

     

    There is a sniper location in the level “Corinth River” of Killzone 2. The player is on an elevated walkway overlooking a medium-sized area containing enemy NPCs.  Both the player and enemy NPCs have cover, but by looking down from above, the player is able to locate the enemy NPCs and engage them.

     

     

    PATTERNS FOR LARGE-SCALE COMBAT

    These patterns provide areas for combat gameplay, with the player either engaging large numbers of enemy NPCs or a single powerful enemy NPC (a boss fight).

     

    STRONGHOLD

     

    Description:

    A stronghold is a confined area, generally with good cover. Characters in a stronghold can defend against attackers while remaining protected. A stronghold has limited access points so the defending characters can cover them easily.

     

    Affordances:

    • The size of the stronghold
    • The amount of cover available in the stronghold
    • The number and type of access points
    • If defending/capturing the stronghold is a level objective

     

    Consequences:

    Generally a stronghold would be designed as a defensible location for the player. The effect is usually to reduce the pace of the level, but in some cases, a large number of entrances or advancing enemy NPCs can have the effect of increasing tension and challenge.

     

    Relationships:

    A stronghold can be considered a specialized type of arena or sniper location. Entrances to the stronghold may be choke points.

     

    Examples:

    The Halo 3 level “The Covenant” contains a stronghold. The player is in a large open area and engages enemy NPCs entering through multiple entrances. These entrances are choke points that help keep the player from being swarmed by enemy NPCs, but it is challenging to cover them all at once.

     

    There is an instance of a stronghold in the level “Fish in a Barrel” of Gears of War, shown in Figure 2. The player and friendly NPCs are in a central area with minimal amounts of cover while being engaged by enemy NPCs from multiple directions. The effect is challenging and high tension combat.

     

    2.png.9d436a2eb8c18b08575dad24f4b00f8a.png

    Figure 2: Stronghold in Gears of War

     

     

    PATTERNS FOR ALTERNATE ROUTES

    These patterns provide players with choices about how they want to engage the level.

     

    SPLIT LEVEL

     

    Description:

    A split level is a corridor with an upper and a lower section. Characters on the upper section can attack characters on the lower level. Players can choose the upper or lower route, or switch between them.

     

    Affordances:

    • The difference in height between the levels
    • The degree of openness between the levels, in terms of empty space
    • The number of paths between the levels

     

    Consequences:

    Allows for different strategies and can increase the pace of a level as the player moves back and forth between levels.

     

    Relationships:

    If the corridor is narrow, the upper section could be a gallery. Using one section to avoid enemy NPCs in the other section makes it a type of flanking route.

     

    Examples:

    There is a split level in the “Lowlife” level of Half-Life 2: Episode 1, shown in Figure 3. The player is moving through a large open area with elevated passageways. The player must switch back and forth between the two paths to avoid the most powerful enemy NPCs.

     

    3.png.f1d8e0181f85932a1030df9c62688ff9.png

    Figure 3: Split level in Half-Life 2: Episode 1

     

    The Halo 3 level “Crow’s Nest” features a long split level section. The player may stay on the upper level and engage enemies on the lower level, or use the lower section and engage them directly.

     

     

    COMBAT NON-PLAYER CHARACTERS IN FIRST-PERSON SHOOTER GAMES The work presented in this section is based on material originally developed in collaboration with Gabe Rivera.

     

    The patterns presented in this section are for the enemy NPCs in FPS games. Enemy NPCs are controlled by the game engine and are the main source of conflict during gameplay. While they could be considered aspects of game design rather than level design, they are placed by designers and their tuning and behavior are highly dependent on how they are used. Designers can control not only where the NPC is placed but also the NPC’s scripted behavior, how they are equipped, their level of health, their level of armor, and other variables.

     

    For this research we explored elements that pertain to all NPCs within the shooter genre and then analyzed various games to see if NPCs consistently fell into patterns. Patterns were identified by observing NPC behavior and discerning which elements were combined in the same way within a number of games. Each pattern is accompanied by our observations about how it’s used by designers to create gameplay, as well as a list of elements that define the pattern.

     

    ELEMENTS OF A NON-PLAYER CHARACTER

     

    Below is a list of elements that make up a NPC as well as a brief description of how they can be used by a designer to create gameplay during combat. These will be used in the pattern collection to categorize the specific patterns.

     

    Movement Type – This describes the way the NPC will typically move in a combat situation. Many NPCs employ multiple Movement Types and can switch between them depending on the situation.

    • Flanking Intensive – The NPC will move to attack from unexpected directions, i.e. the NPC tries to approach the player from a different side than where the player’s attention is directed.
    • Passive – The NPC will not move when attacking. Never straying too far from that location and available cover.
    • Slow Push – The NPC will slowly advance on the position of the opposing force, usually in a straight line. This can be without the need for cover, but it is possible for the NPC to utilize cover while making its way forward. This main difference between this and Cautious is that it will constantly try to close the distance from its target and not try to stay away.
    • Rush – The NPC will make a dash at a specific target without any regard for their safety, typically in a straight line. However, the main aspect of this movement type is that they will attack very fast and often try to close the distance between themselves and their target as fast as possible
    • Cautious – When used, it means that the NPC is opting to move around the battlefield but tries to maintain a distance from its target. Often trying to utilize cover when possible and not closing the distance when possible. This is different from a slow push because this NPC tries to maintain a specific radius around its target, without advancing.

     

    Movement Range – This is how far the NPC will move during an engagement. This can be Low, Medium, or High.

     

    Movement Frequency – This is how often the NPC will change their position during an engagement. This can be Low, Medium, or High.

     

    Attack Frequency – This describes how often the NPC will initiate an attack. This can be Low, Medium, or High.

     

    Weapon Type – The patterns include the following. They are described in more detail in the following section:

    • Sniping Weapon
    • Close Blast
    • Assault Weapon
    • Projectile
    • Power Weapon
    • Melee Weapon

     

    Weapon Damage – A general indicator on how much damage the NPC will do to the player’s Health, Shields, or Armor. This can be Low, Medium, or High.

     

    Armor/Health – This denotes how much damage the NPC can take before being killed. This will typically be linked to how hard the NPC is to defeat. This can be Low, Medium, or High.

     

    Motive – This is an indicator of what type of combat encounter the NPC would create and shows its purpose to the designer. This hinges on three main factors that an NPC can affect:

    • Challenge – The degree of difficulty within a combat encounter.
    • Tension – The degree of mental stress the player experiences during a combat encounter.
    • Pacing – The degree of movement that the player will engage in during a combat encounter

     

    A pattern can affect each of these three factors by creating a situation where they can be at Low, Medium, or High.

     

    PATTERN COLLECTION

    Below is a list of all the patterns that we have collected during our research. Each base pattern specifies the primary function of that general type, while each sub pattern denotes how that function is carried out.

     

    Soldier – An NPC that pressures the player from range.

    • Grunt – A weak enemy that attacks from a medium distance, often in groups.
    • Elite – A strong enemy that works to contain the player from a medium distance.
    • Grenadier – A weaker enemy that maintains long distance to encourage players to move forwards.
    • Sniper – An enemy that deals high damage from a long distance to force players to move carefully.

     

    Aggressive – An NPC that attempts to close the distance between itself and its target in order to increase pressure.

    • Suicide – An enemy that immediately rushes at the player, at the cost of its own life.
    • Swarm – An enemy that rushes the player in groups, but deals low damage individually.
    • Berserker – A strong NPC that deals a high amount of damage over a prolonged amount of time.

     

    Carrier – An NPC that will spawn more NPCs during an encounter.

    • Sacrifice – An NPC that creates more NPCs in the case of its own death.
    • Summoner – An enemy that spawns more NPCs at a distance

     

    Tank – An NPC that poses a significant singular threat and prevents the player from proceeding

    • Stationary Tank – A slow-moving NPC that deals high damage at a long range.
    • Shield – An NPC with a large amount of armor, but only in a single direction.

     

    The following sections detail all of the base patterns and at least one of their sub patterns.

     

    SOLDIER

    Soldier is a NPC that will pressure the player from long range. Its main strategy is to control the available space in the encounter. NPCs of this type make up the majority of units during an encounter. They are primarily used to control pacing by forcing the player to take particular paths through the environment. These NPCs will have a weapon type that is an Assault, Close Blast, Sniping, or Projectile.

     

    Grunt

    Description:

    The Grunt is a weak NPC that will try to maintain a medium distance away when attacking. The main function this serves is to draw the player to forward through the level and increase the player's confidence. This pattern is distinguished by always having medium movement range, medium movement frequency, and light armor. The motive of the Grunt pattern is to create a situation with low tension and low challenge.

     

    Affordances:

    • Movement type can be Slow Push, Flanking Intensive, or Cautious.
    • Attack frequency can be either Low or Medium.
    • Weapon damage can be either Low or Medium.

     

    NPC Relationships:

    The grunt has a special relationship with the Suicide pattern, because sometimes a grunt may change to the suicide pattern in the middle of an encounter.

     

    Examples:

    Halo: Combat Evolved - The Grunt is a small unit that appears in every game within the Halo franchise. It has a low amount of Armor and is usually to be equipped with an assault weapon that does a low (Plasma Pistol) or medium (Needler) amount of damage. They exhibit the special relationship with the Suicide pattern in that they will self-destruct in times of desperation. The range it keeps is either short or medium but tries to pester the player by implementing the Cautious movement type.

     

    During the campaign they primarily occur within encounters to create a lower challenge but increase the pace of the encounter. As a consequence, the player feels more empowered and will pursue a route that contains a higher ratio of grunts compared to any other path. This occurs in the level The Pillar of Autumn; often the designers put grunts down a particular corridor to encourage the player to move in that direction. This signals to the player that it is the correct route to follow while lowering challenge, increasing the pace, and lowering player tension.

     

    4.jpg.01f14fd2f1beb9bfca47a635200408d2.jpg

    Figure 4: A Group of Grunts in Halo: Reach

     

    Half-Life 2 - The Metro Police Officer utilizes a Slow Push or Cautious Movement Type and primarily is equipped with an assault weapon, typically a sidearm. They will shift between the movement types in an effort to move a player forward. Typically this means that they will begin in a cautious movement type and, if they player doesn’t pursue them, will move toward the player in order to get the player to move. This doesn’t occur in any particular instance but can be seen where there are Metro Police Officers in levels such as Route Kanal or Water Hazard. In the game, they basically act as bait to simply pull the player forward. They are primarily seen as the main enemy in the early game and are increasingly used as bait in the latter half of the game.

     

    5.png.a28b7a5ad28948b1a788c1e8dcb88e0d.png

    Figure 5: Two Metro police officers in Half-Life 2

     

     

    EXAMPLE ANALYSIS

    To show the usefulness of NPC design patterns we will use them to analyze a short encounter and generate a new enemy type. The level Winter Contingency in the game Halo: Reach contains an encounter in which the group is tasked with bringing a communications outpost back online. This sequence starts with the team landing in front of the communications outpost in order to secure the location.

     

    After starting the level, the player encounters their first group of enemy NPCs in an Arena with Flanking Routes to the left and right. The NPCs that populate the arena are a small force of Grunts and Jackals. This encounter has a low amount of challenge and allows the player to gain a foothold without much effort. It is fairly easy for the player to move forward and incapacitate the Grunts, which fall under the Grunt NPC pattern. However, it is much harder kill the Jackals in a head on attack since they are a part of the Shield NPC pattern. The interplay between the Grunt and Shield patterns help to create a much easier encounter for the player by driving them to explore the area and flank the Jackals.

     

    The player goes into the encounter and immediately recognizes that most of the Jackals were located in the Arena, where the player is at a disadvantage. Since that place is the hardest to break through, the player is drawn to the left because the Grunts offer a lower level of resistance. The Grunts signal to the player this path is safer and encourages them to move through the Flanking Route. The player can now flank the exposed back of the Jackals, which has a pattern specific weakness of only being able to withstand a large amount of damage from one direction.

     

    We can analyze this encounter and explain it through the enemy NPC patterns that we have created. The designers used Shield NPCs in order to bar the player’s way from one direction and give the illusion of a higher degree of challenge. However, by adding in the Grunt NPCs it allowed them to encourage the player to move into an advantageous position. The interplay between these two types helped to create an encounter with a low amount of challenge but high amount of tension.

     

     

    WEAPONS IN FIRST-PERSON SHOOTER GAMES

    *Note: The work presented in this section is based on material originally developed in collaboration with Rob Giusti.

     

    To define and discuss weapons, game and level designers have re-purposed an existing classification system: the terminology used to refer to real-life weapons, terms such as “Sub-machine Gun” and “Sniper Rifle.” Though these classifications do easily explain the mechanics of the weapon, the use of such terminology fails to accurately describe gameplay behaviors and to encompass the fictional aspects of digital games.

     

    Knowing how a particular weapon functions in real life does not actually give an accurate depiction of how the weapon functions within a game. For example, the shotgun in Halo has a much shorter effective range than its real-life counterpart. Many similar weapons fall into different weapon patterns depending on how designers implement them.

     

    Though many action and adventure games use weapons, shooter games are affected by this lack of terminology more than others due to the fact that weapons are at the core of gameplay. In the vast majority of first-person shooters, the player's weapon never even leaves their view. In addition, weapons are the central method through which players interact with the world in these games.

     

    With this pattern collection we hope to create a language that can be used to describe weapons in a way that encapsulates the gameplay behaviors that each pattern elicits. Each pattern is named in a way that aims to be inclusive of all weapons, fictional or nonfictional, that elicits similar player behavior.

     

    We accumulated these patterns through analyzing weapons in popular and historically significant first- and third-person shooter games.

     

    ASPECTS OF WEAPON PATTERNS

    To provide a basis for defining patterns in weapon design, the following template will be used:

     

    Name – A descriptive identifier used to refer to the pattern that is recognizable and imparts the core functionality of the pattern.

    Description – A brief explanation of the typical features of a weapon derived from this pattern.

    Affordances – Aspects of the pattern that can be varied between different weapons within the pattern.

    Consequences – How use of the weapon pattern affects gameplay.

    Level Patterns – Relationships between the weapon pattern and patterns in level design.

    NPC's – Relationships between the weapon pattern and patterns in non-player character design.

    Examples – Uses of the weapon design pattern from popular commercial shooter games.

     

    Patterns contained within another are considered to be super- or sub-patterns of each other. Patterns are not mutually exclusive from each other; a weapon can fit multiple weapon patterns.

     

    A large number of affordances can be considered universal among weapon patterns, including:

    • How much damage the weapon deals
    • The range of the weapon
    • The area of effect of the weapon
    • How often the weapon can be used ("Cooldown")
    • How many times the weapon can be used before needing to be reloaded (“Capacity”)
    • How much ammunition a player can carry
    • How carrying the weapon affects the player’s movement
    • How the weapon imparts damage to the enemy (On hit, delayed, continuous, etc.)
    • Any special effects that the weapon has on the enemy 
    • Any special abilities that the weapon bestows

     

    Repetition of a Universal Affordance within a particular pattern description signifies that pattern differs significantly within the pattern in that aspect.

     

    PATTERN COLLECTION

    PROJECTILE

    Description:

    Objects thrown or fired in a physics-defined arch. Most often, Projectiles are explosives that deal damage in a large area of effect. Projectiles are also associated with long reload times and small capacities. Projectiles also often have a low amount of maximum ammunition.

     

    Affordances:

    • The range of the weapon
    • If the effect is immediate or delayed
    • The area of effect of the weapon
    • Any special effects of the weapon

     

    Consequences:

    Projectile weapons are useful for circumventing cover. Also, they heighten the challenge through being more difficult to aim than other weapons.

     

    Level Patterns:

    Projectiles can be used to harm enemies in Sniper Locations or guarding Choke Points without directly engaging them. Players using Projectiles are often vulnerable to Split Levels and Galleries, due to ammunition limitations and a lack of sufficient cover.

     

    NPCs:

    Grenadiers, Elites, and sometimes Tanks use Projectiles to force the player out of cover and impose a greater threat.

     

    Projectiles allow players to take on large groups of enemies, such as Swarms and Carriers, and fight against heavy enemies, such as Tanks and Snipers, without engaging them directly. The long recharge times and tendency for Projectiles to have large areas of effect make them less effective against Berserkers and Suicidals.

     

    Examples:

    The Demoman class from Team Fortress 2 [54] has a Grenade Launcher that allows the player to fire pipe bombs at enemies. These pipe bombs explode on impact with an enemy; otherwise the bombs roll for a few seconds before exploding.

     

    In the Halo series, the rocket launcher is a weapon that is both a Launched Projectile and Power Weapon. The weapon launches a rocket at high velocity, creating a large explosion that can instantly kill targets, both those on foot and those in vehicles. However, the weapon carries very limited ammunition and takes up space in the player’s limited arsenal.

     

    6.png.e7bdeccfb6de14d07d7cff87ccaca924.png

    A player firing Projectiles in Team Fortress 2

     

    Thrown Projectile

    Description:

    A non-bullet object thrown by the hand of the player's character and categorized by short range and highly affected by gravity. Thrown Projectiles often have high damage or severe special effects, balanced by scarce ammunition.

     

    Affordances:

    • Special effects associated with the physical object of the projectile

     

    Consequences:

    The player is able to attack opponents who are behind cover, however they are forced to keep in mind their ammunition and range limitations.

     

    Level Patterns:

    Thrown Projectiles allow players to defeat an enemy guarding a Choke Point, or players on another level of a Split Level. In areas with long distances, such as Sniper Positions, or with enemies at multiple angles, such as Arenas and Flanking Routes, Thrown Projectiles are not very effective.

     

    NPCs:

    Elites utilize Thrown Projectiles in order to pressure players who are taking cover. Some Summoners use their spawned units as a sort of Thrown Projectile as a way of deploying them.

     

    A player can use Thrown Projectiles much like normal Projectiles to attack heavy Tanks from behind cover. Thrown Projectiles are often more effective against solitary, close-range targets and less effective against loosely grouped Swarm and Grunt enemies.

     

    Examples:

    In Call of Duty 4: Modern Warfare 2 [52], the throwing knife is a powerful Thrown Projectile with harsh limitations. The weapon has a short range, however a hit with the knife immediately kills the enemy. A player also may only carry one knife at a time.

     

    Halo 3 offers players a handful of varied thrown projectiles. Fragmentation grenades can be thrown a good distance and rebound off any obstacles until they detonate after a set amount of time. Players also have the option of using plasma grenades instead, which attach themselves to level geometry and players on contact, but have a shorter range and smaller blast radius

     

     

    EFFECTS OF WEAPON PATTERNS ON LEVEL DESIGN

    By forcing the player to use particular weapons in certain parts of a level, the level designer utilizes the relationships between the weapon and level to best control the experience and gameplay.

     

    For example, in the Ravenholm section of Half-Life 2, the player begins the level with a weak Melee Weapon, Sidearm, and Assault Weapon. The player progresses through Arenas and Chokepoints with a numerous number of Grunt and Swarm enemies, resulting in high tension and challenge. Later, the player fights Berserker and Carrier enemies, but acquires a Close Blast weapon and moves into Choke Points where the player has the advantage. The tension and challenge drop to give the player a respite and allow them to learn how to utilize the weapon. As the player proceeds, the level patterns become more Arenas and Split Levels, forcing the player to use weapons accordingly, bringing the challenge and tension back up for the climax of the level.

     

    In multiplayer levels, weapon placement allows the level designer to direct players. The designer can hint at what weapons are best suited for a certain area, force players to carry an unsuitable weapon across an area to get somewhere where that weapon is more useful, or even make it more difficult to use a particular weapon from a particular location.

     

    The multiplayer level Blood Gulch in Halo has Sniping Weapons atop each base at the ends of the map, overlooking large amount of the level and subtly hinting at the advantageous Sniper Position. A Power Weapon, the rocket launcher, is placed in the center of the map, forcing players to travel a long distance and expose themselves in order to procure the weapon.

     

    7.png.76ed76d83356fe5c55ea972c542d91b6.png

    The multiplayer level Blood Gulch in Halo

     

     

    APPENDIX A - ADDITIONAL DESIGN PATTERN COLLECTIONS

    MULTIPLAYER FIRST-PERSON SHOOTER LEVELS

    *Note: The work presented in this section is based on material originally developed in collaboration with Chris Ueda.

     

    In our examination of multiplayer levels, we will be paying particular attention to their relationship to single-player levels and their associated patterns. Certain elements of multiplayer design patterns have parallels to their single-player counterparts. While these parallels suggest a large overlap in design principles for the design of levels in a (FPS) game, there is a difference in design goals between single and multiplayer levels.

     

    The goal of the level designer is to provide a specific gameplay experience to the player. Experiences such as a distinct gameplay experience or narrative diegetic effect can be produced by designers through the use of level geometry, item placement, scripted events, and other level design elements. A single-player level is designed as a linear space, segmented into rooms separated by corridors. This allows the designer to create highs and lows in player tension, pacing the gameplay and giving the player opportunities to experience moments of intensity without tiring themselves out. For example, Half-Life 2, a single-player FPS, often makes extensive use of open spaces in which the player is guided through the level while being given visual cues tying narrative and world space together. The level tells the story rather than large blocks of text or cutscenes, adding to a sense of immersion.

     

    The difference in player count between single-player and multiplayer affects the way in which the designer needs to approach level design. When crafting a single-player level, the designer aims to tailor an experience to one player, but in designing multiplayer levels, the game state is now based on the inputs of other players, whose game-playing experiences the designer must all consider.

     

    An example of the differences between single-player and multiplayer levels is apparent in spawning points for players versus spawning points for NPCs (non-player character). While they have similar purposes (introducing new entities into the level), in multiplayer levels additional players are spawned in place of NPCs. In a singleplayer level a NPC can be created whenever the designer chooses, but in a multiplayer level, the designer must equally consider all players when designing spawning points in a level. As the spawn points of each player affects the encounter rate, and therefore the pacing of the game. If too high, a player may get exhausted by constant action, or get bored between respawns if it's too low.

     

    Level design patterns are employed by designers to explore design choices and craft the desired gameplay for a level. These patterns vary based on the requirements of the game. For example, FPS gameplay involves the use of space and resources in real-time in a way that makes cover or item pickups useful. Therefore, patterns emerge that relate to the placement and frequency of these objects, and these patterns differ according to the unique features that distinguish multiplayer from single-player gameplay.

     

    KEY CONCEPTS

    CONFLICT POINTS

    A conflict point is a location in a level which is designed to bring opposing forces into an encounter. These locations are key in managing rhythm and flow in multiplayer levels. By designing a level with conflict points in mind, the intensity and pacing that a given player experiences can be adjusted. To do this, designers can utilize elements of a conflict point such as chokepoints, strongholds, pickups, and objectives. Chokepoints and strongholds change the movement of players in and about a conflict point, while pickups and objectives provide players a focal point for encounters. A powerful weapon or a bunker may motivate players to prioritize combat in that area, increasing the overall intensity of the location over others. Examples include the flag's location in a CTF game of Halo: Combat Evolved, Control Points in Team Fortress 2, or the Farsight XR-20 (an extremely powerful weapon) in Perfect Dark. These are objectives that players can obtain to get an advantage, and naturally conflict will occur in their vicinity.

     

    Use of conflict points is critical to many design patterns, as multiplayer FPS levels depend on them for creating player encounters. For example, bomb sites in CounterStrike serve as the objective destination for the Terrorist forces. The objective of the Counter-Terrorist forces is to prevent the Terrorist demolition mission, and both teams are aware of the state of the bomb sites through in-game HUD cues. These areas are often camped, with one team lying in wait to ambush the other team. The expected combat in the conflict point reinforces player planning and coordination followed by a burst of high-intensity combat. To support this style of gameplay, these bomb sites often contain various types of cover and are connected to the rest of the level via small, easily ambushed entryways serving as chokepoints.

     

    PATTERNS IN MULTIPLAYER GAME TYPES

    Multiplayer FPS games require a different set of game rules and objectives from single-player. Sets of rules collectively known as game types are defined in order to provide specific gameplay experiences. These may include rules such as a priority object or location, or a score objective. Level designers apply key concepts of multiplayer level design in the context of a specific game type in order to create a playable level.

     

    CAPTURE THE FLAG (CTF)

    This game type has both teams simultaneously on offense and defense, trying to claim the other team's flag and bringing it back to their own base while protecting their own flag. The game type is similar to Control Point, especially when the flag is located at a team's base. The flag's starting location serves as a point of conflict, and is often a strongly fortified location, making defense easy and requiring coordinated offense to capture.

     

    After claiming the flag, a player must bring the flag to their team's own base. The enemy team must prevent the flag from being delivered by attacking the carrier. Flag carriers are encouraged to use alternate paths and shortcuts in order to evade the opposing team. Levels are often symmetric to ensure balance. Respawn times are long, allowing a team to press their advantage after defeating opposing forces.

     

    Examples include Unreal Tournament - Facing Worlds (symmetrical) and CTF4 in Quake 3 Arena. Blood Gulch in Halo: CE is a classic example, set in a wide, open canyon with rolling hills. On the two far ends, a single bunker houses each team's flag. Teleporters quickly move players from a base to the middle of the stage, but not the other way, allowing respawned players to return to the action.

     

    Team Fortress 2's Payload maps are a variation of the CTF format. In this game type the offensive team moves a cart forward by standing besides it, while the defense sets up fortifications to prevent progress. The linear path of the cart and the respawn system of TF2 distinguishes this game type as being closer to CTF rather than Delivery, described further below.

     

    8.jpg.8c8b43c67a6e319101f1b57a23cac819.jpg

    Team Fortress 2's Goldrush, a Payload map where the blue team moves the cart along to its destination

     

    PATTERNS IN MULTIPLAYER LEVELS

    Multiplayer level design strives to create a level playing field. To provide gameplay options while maintaining this balance, beneficial structures such as sniper locations and alternate routes need to be viable, while the opposing players are provided with a valid counterstrategy. In Halo: Combat Evolved single-player, a sniper location provided a significant advantage to the player. In the multiplayer game, players in sniper locations must also be wary of counter attack from the complementary sniper location on the other side of the level, or rely on their teammates to protect a poorly defensible position. Team strategy may be required to make the most of a given pattern's potential, often reflected in a strong offensive or defensive feature of a location.

     

    ARENA

    Description: 

    Open areas with good sight ranges. Promotes encounters as a result of visibility or traffic – arenas are often conflict points

     

    Affordances:

    • Can contain a Control Point.
    • Pickups will increase traffic and conflict in the area.
    • Can include features such as battlements and alternate paths to prevent overcongestion.

     

    Consequences:

    If surrounding area is confusing or congested, adding arena features may improve traffic flow.

     

    Has sporadic cover, providing good defense but not concealment.

     

    Examples:

    de_aztec (Counter-strike) - The terrorist force cross an open, unprotected area and take cover behind the crates located at demolition point A. A ramp up from a lower floor and a hallway with clear view of the bomb point threaten the terrorist force's objective.

     

    Hang em' High (Halo: Combat Evolved) - An extremely open map, with small blocks for cover, and ramps leading up to a second level which surrounds the map. Catwalks crisscrossing the level can be accessed from the second level. These lead to powerful weapons, but players are vulnerable to attacks from below.

     

    9.png.be7f0ff740d163c8a3b4ea68b757d7fc.png

     Halo: Combat Evolved, Hang em' High: Many catwalks cross the length of the map

     

    Follow this link for the full writing: https://users.soe.ucsc.edu/~ejw/dissertations/Ken-Hullett-dissertation.pdf

     

     

    Follow Ken

    Linkedin: https://ca.linkedin.com/in/khullett

     

     

    Follow Next Level Design

    Join the Forum: http://www.nextleveldesign.org/index.php?/register/

    Follow us on Twitter: https://twitter.com/NextLevelDesig2

    Discuss on Discord: https://t.co/hkxwVml0Dp


    Article Preview: This article is a portion of a dissertation by Kenneth Hullet. It aims to scientifically present a set of level design patterns for FPS games, and be "a resource for designers to help them understand how they are creating gameplay through their art."


    User Feedback

    Recommended Comments

    There are no comments to display.



    Join the conversation

    You can post now and register later. If you have an account, sign in now to post with your account.

    Guest
    Add a comment...

    ×   Pasted as rich text.   Paste as plain text instead

      Only 75 emoji are allowed.

    ×   Your link has been automatically embedded.   Display as a link instead

    ×   Your previous content has been restored.   Clear editor

    ×   You cannot paste images directly. Upload or insert images from URL.


  • Similar Content

  • Categories