Video game systems can be difficult to understand, and communicating them effectively can have a big impact on quality. Visualizations are an incredibly helpful tool for managing complexity and presenting data in a way that is easily understandable.
Creating a truly effective visualization is often more difficult than simply following a few technical steps. There are many choices with complicated trade-offs between accuracy, conciseness, readability, usability, and maintainability.
To address this challenge, we propose the PAUSE Framework, designed to help designers create effective visualizations that are crafted to best meet the needs of their intended audience.
The PAUSE Framework uses five core questions to assist in choosing the best visualization for a given problem.
Why are you creating this visualization?
To provide or obtain clear answers to questions, it is important to be clear about what those questions are. A visualization that serves one purpose, such as helping to fine-tune an economy, won’t be as effective for another purpose, like explaining that economy to a publisher. Attempting to use the same visualization for both purposes is often unsuccessful.
It’s easiest to think about your purpose in terms of a goal verb. Here are some examples:
Who is the audience for your visualization?
Effective communication is a two-way process. It’s important to consider the expectations, expertise, and goals of the viewer; the content you create should be tailored to the specific audience you are trying to reach. For example, the needs of your development team will differ from those of your players or investors. Creating a visualization that is relevant and valuable to one group won’t be as effective for a different group.
Here are questions to ask about your audience:
How will your visualization be used?
Knowing how your visualization will be used will allow you to tailor it to the context. If it is for a single-use presentation, perfect accuracy is seldom important. If it will go into permanent documentation, it’s important to be as accurate as possible. If it’s in a presentation, it will need to be simpler with fewer details, but if it’s as part of an app that people will be directly using on their computers, it can be more dense.
Here are questions to ask to determine the usage of your visualization:
What is the minimum amount of information that needs to be shared?
Effective communication relies on simplicity. To find the right balance between providing enough information and avoiding overwhelming your audience, focus on what is most important for them to know.
Here are some questions to ask that will help define scope:
Which aspects of your problem are the most important?
Direct your audience’s attention to the most important part of the data you are sharing, establishing a clear hierarchy of importance. If you do not provide the lens for your audience, they will bring their own; what you emphasize will show what you think is important.
Here are questions that determine emphasis:
To see how the PAUSE framework can help focus your visualization, let’s go through an example.
Imagine you are working on a visualization for the crafting system in a survival game, and want to share it with your team.
You can immediately imagine several ways to visualize this design.
To know the right approach, it’s important to first identify your purpose. Do you want to show how the crafting system connects to the rest of the game? Do you want to be able to visually verify that every recipe is reachable? Do you want to balance the economy?
For this one, let’s say you want to see how the crafting system connects to the rest of the game. Spreadsheets and graphs list content instead of showing connections, so we can rule them out. The state diagram is internal to crafting so also less effective, which leaves us the high-level or node diagram. Already, we’ve cut down the possibilities!
Purpose: Communicate how the crafting system connects to the rest of the game
|Additional Insight: It can be tempting to look for a visualization solution that can meet all the needs of all audiences, but some purposes are inherently incompatible . If you find yourself frustrated trying to satisfy multiple purposes, consider whether multiple visualizations would be better.
Now, say that the audience you’re sharing with is fellow designers. Whatever your own visualizations preferences are, you now must consider your team. Have they used an interactive node diagram app before? Do they go straight for whiteboard pens every time?
Let’s imagine you’re a tight-knit tech-savvy team. Your shared context frees you to use shorthand terminology in sketching the shape of major systems, spending your energy on a node diagram app that shows specific details. You trust that your team already knows that inventory and crafting will be connected, but they need to know through which specific subsystems. That context is the razor to let you cut unnecessary detail as you build your diagram.
Audience: The design team, who has a good context for the high level function of crafting and is comfortable with a variety of visualization tools.
|Additional Insight: This approach will likely go over great with the team, but what if you are asked to also share this design with leadership? Tempting as it is to just send what you have, the shared context you used to focus on the most relevant bits isn’t going to land for leadership. They’re expecting a slide deck with a few diagrams, and sending a link to an interactive node diagram isn’t going to work. It isn’t starting from scratch, but revising the assumptions about your audience isn’t just replacing some terminology, it can mean rethinking your whole approach!
Continuing with the crafting system example, it’s worth considering how often you’ve been asked to visualize this system and how often it’s expected to change. Maybe this is your fourth major iteration on crafting, and you suspect it won’t be the last. If you’re using it to communicate changes in the updated system, why start from scratch every time?
An editable and flexible visualization tool is going to take more time to set up than sketching a diagram on a whiteboard, but if your team wants to keep referencing it while making changes, it’s time well spent. If they also want very specific nitty-gritty details, consider how much of your time will go into maintaining that much detail, weighed against the time it would take an engineer to build you an automated update pipeline.
Even if the purpose is just communication, if effective communication relies on up-to-date details over a long period of time, it’s worth investing in finding or building the right visualization tool to meet your needs. An up-to-date view that designers can reference on their own may save you having to answer questions frequently over Slack.
Usage: Consulted throughout the project’s lifetime, used periodically but not daily, expect to be edited as it’s kept up-to-date with planned iteration on crafting.
|Additional Insight: Sometimes a visualization will only be used once, such as for a presentation or pitch. Creating a one-off visualization is still worth the time as long as you consider the usage and choose the right visualization for the job.
In a cohesive survival crafting system, if you exhaustively mapped every crafting connection of any kind, you’d be diagraming the whole game. Obviously that won’t be readable for anyone, so the question you need to ask is: what’s the minimal scope that meets your purpose?
We established earlier that we care about system connections, not the specific content within crafting. We also know our audience has good context for the role of crafting, and just needs to understand the specific details they can work with. That’s already cut down your scope a great deal!
We can narrow it down further with follow-up questions. Do they want specific technical details to inform their implementation? Or do they just want to know what mechanics exist within crafting? Do you already know that some details like UI connections are so crafting-specific they won’t be applicable to this exercise, or is the point to be totally exhaustive and find the creative edge cases?
If you have an engineering team handling implementation, your scope will be more abstract than the technical specifics. However, it does need to be comprehensive in representing all the system pieces at play. Part of why you’re communicating the system to your team is for effective collaboration – there needs to be enough detail to help them spot problems and opportunities you missed. Cutting off the low end of tech specs as well as the high end of simplified big-picture connections leaves you a workable middle – that big-picture is what you’ll refocus on when you prepare that deck for leadership.
Scope: Complete in representing all mechanics within crafting and their existing connections to other mechanics, but not detailed to the point of providing implementation details. Does not include the specifics of content within crafting.
|Additional Insight: Scope is the step in the PAUSE framework that most benefits from an outside perspective. Talk to your audience about what they need from the visualization, how they envision engaging with it, and what decisions they might make with it. Try to get specifics so you can focus on the minimal scope needed to do the job.
At this point, you know your visualization approach, who will use it, how they’ll use it, and what information it will contain. You’ve got everything you need to start building it – but even the most concise visualization is information-dense, and how you organize that information will emphasize different aspects.
You could make every system an identically shaped node, connected by identically drawn lines, but is that an accurate representation? Crafting and inventory are deeply connected, with many interactive mechanics, whereas crafting and lighting may have only a cursory connection via a handful of light-creating items. Do you want to emphasize those more entwined systems so that your team focuses on the most impactful areas, or do you want to draw attention to the sparse areas in the hope of filling them out?
This intentional organization is the last trick that helps your visualization lead the audience towards fulfilling the intended purpose. If node sizes correspond to the amount of effort allocated to a system, your team will be drawn to making conclusions they might have missed before about where you’re relying on risky systems. The same diagram but emphasizing how much content has been created for each system would lead to useful but different observations, so keep in mind which you’re looking for, and never be afraid to provide more than one visualization to support multiple goals!
Emphasis: Representing how much developer effort is allocated to systems via node size, to emphasize how well-supported connections between systems are likely to be.
|Additional Insight: In many ways, Emphasis represents the visual part of the visualization. Tackling Emphasis last is a great way to ensure your visualizations do more than just make a lot of data easier to read. Because you know so much about the purpose, audience, usage, and scope at this point, you can make targeted decisions about how data is represented.
An example end result, built in Kumu. Leaving out specific content allows for exhaustively listing major systems without becoming unwieldy. Emphasizing developer allocation draws attention to observations like a lack of support for Harvesting in contrast to high support for its connections.
Visualization is a core skill for any systems designer, but we rarely discuss our process for creating the tools, diagrams, and charts that do the heavy lifting of communicating our designs. Too often, we grab the first visualization technique we can think of; something we’ve used before, something that seems easy to create, something we can start on right now, only to suffer through maintenance, misunderstandings, and throw away work. But we also know the pride and joy of creating just the right visualization; something simple, something that scales, something really useful.
Players may never see them (although they’ve certainly influenced a fair number of UI elements) but visualizations are an important part of our work. When our team came together to discuss this topic, we got a great deal out of comparing notes, sharing examples, and talking through our own processes. Design is a discipline made up of many, disparate skills and while systems visualization is not one that has gotten a lot of attention, our diverse group of game designers found plenty to sink our teeth into for this topic. The act of committing a unified process down on paper has been clarifying and energizing, and we hope that other designers will find it useful and thought provoking.
Here’s an appendix of specific visualization techniques you can add to your toolbox. This includes discussion of very well known techniques like spreadsheets and flow charts, where we give an especially strong example of the technique and talk about what makes it exemplary. We also include some of the more novel visualization techniques our team shared, not just as inspiration for more people to use the same techniques, but as an argument for the benefits in figuring out a domain-specific visualization technique that works for your project.
The purpose of this visualization is to demonstrate what systems are part of the game, the motivations they are targeted to, and the importance of that audience segment to the game.
It’s very easy to go too complex here or try to be perfectly accurate and mess up the clarity of the communication. For this type of diagram, half of the purpose is to show the intention and thought that went into the design. In many executive pitches, there is not enough demonstration of the ‘why’ of the decisions and instead the focus is on the ‘what’.
The purpose of this visualization is to demonstrate art standards for objects in an easy to understand way to the art team. Often, arbitrary constraints like “Objects must be bigger than 0.5m in size” or “Don’t make things 3.5m tall” are hard to remember; and they’re confusing, since there’s no perceived justification.
Showing relationships between classes of objects next to each other is particularly effective, as it helps the viewer build an intuition for the broader rules that are being demonstrated.
The purpose of this visualization is to illustrate the intended drop ratios of armor types as more are added over time.
This visualization is paired with a step-by-step text explanation to ensure that the reader understands the context, and so ensure the understanding of the visuals.
Adding specific numbers/percentages or values on the timeline would distract from the message and mire the reader in the details.
The purpose of this visualization is to show the economical sources and sinks present in the game, easily distinguishable by category (Gameplay/Paid Content, etc…) as well as the intended ratio of what activity would generate the desired commodity.
Embedding the intended percentages allows the design or quality assurance teams to validate data (once the data is present) with the original intentions, and adjust either the stats/formula within the game, or the guideline document accordingly.
Often, when communicating a system design to others on the development team, what is needed is an understanding of how players will experience the design more than how the design functions or how it is implemented. Simple storyboards that show, step-by-step, how a player will engage with the system can communicate both the design intent of the system as well as direction around UX and interaction design needed to support the system. Creating this type of visualization also has the added benefit of forcing the system designer to consider player experience concretely at a relatively early point in design.
This type of visualization works best when it can be done quickly and with little concern about aesthetics. It is usually created early in the design process to create alignment with others on the development team.
We may wish to communicate the relationships between various systems. Here, showing multiple layers of relationship is a challenge when using a more traditional flow chart, but understanding that two systems affect each other non-directly is sometimes vitaly important. This example uses an expanding ring system to show that some systems affect a player’s power directly (weapon, level, gear) while others affect it indirectly through one or more other systems. Colors are also used to emphasize how one system may influence multiple others.
System designers often use this type of visualization to help themselves gain a holistic view of how systems interact and get a sense for how the whole of the meta-system is balanced. In this example, for instance, weapons have relatively few connected systems, gear has a large number of directly connected systems, and level has the largest number of systems that influence it. Without this type of visualization, it can be easy for unintentional imbalances to emerge and not be discovered until well into play testing.
While spreadsheets sometimes become inscrutable to humans (for which the above systems visualization are potential cures), they remain popular tools for their relative accessibility – they often don’t require a great deal of technical acumen to create or reference.
The below example defines a list of items and various resource costs for each item. Without any bespoke visualization tools, conditional formatting allows us to see which resources might currently be underutilized.
This should be familiar to any designer, but is occasionally overlooked as a tool available when engineering resources aren’t!
Procedural content generation often uses layers of loosely coupled systems, combining many sources of content. The possibility space of possible combinations quickly becomes impossible to mentally map, but it’s hugely important for the team to know what end results to expect.
If a tool can be set up to simulate running the generation rulesets on the current game content, then it can dump the full results into a machine-readable format and surface them with a user-friendly visualization interface. In the below example, data from the game project was converted into JSON, and fed into a simple HTML frontend that presented the combined content in a form that could be filtered based on individual priorities.
Initially created with designers in mind as the audience, it became invaluable for QA who had a daily need to verify the current expectations for content within a procedural world. The time saved in no longer having to chase down answers to “where in the game would I find X?” more than made up for the effort spent creating a bespoke tool!