Game Analytics: Getting started with analytics

Analytics is a powerful tool necessary for informed decision–making at the strategic and operational levels in the process of game development. It will help answer key questions about design, fascination, engagement, gameplay, monetization, etc. However, analytics is a relatively new process in game development, and it can be difficult for non-specialists to find useful information.

Analytics has become a widely discussed topic in the field of game development in recent years, and not only for free-to-play projects. However, the relative novelty of this process narrows the range of knowledge available to a layman, and the available information is very unstructured. Just by googling various terms, you can find tons of information about analytics, since it is used by a number of industries, which is not very useful.

Excluding information related, for example, to healthcare, stock trading or customer analysis of an online store, and focusing on games, you can find a number of blog posts, several good articles in Gamasutra and Game Developer Magazine, presentations from industry events such as Game Developers Conference and Casual Connect, scientific articles, official documents and information from various third-party analytics sources, such as Game Analytics.

These resources, in turn, are full of a huge number of abbreviations (DAU, MAU, ARRPU, LTV, …) that can change their meaning depending on the context. In addition, a lot of attention is paid to analytics related to monetization, that is, the process of analyzing the behavior of players in order to maximize income, as well as related to free-to-play / online games. At the same time, there is much less information on the use of analytics on the Internet, which would help to improve the design in order to increase the user’s pleasure and involvement. In other words, knowledge is fragmented and does not cover many areas that are important for those who have just started using analytics in their games.

The purpose of this post: try to compare some of the most relevant knowledge for a layman about working with analytics, regardless of the type of game or platform. This material does not cover everything a game developer needs to know to create a mature business intelligence methodology, but a possible path is indicated here.

Getting Started: 12 steps

There are many good ways to get started with analytics. The model that we will discuss here is based on the standard information retrieval model used in business analytics almost everywhere, while it is adapted and presented as a guide for a layman in game analytics on the implementation of analytical practices and obtaining results. Accordingly, the article assumes that the reader has no prior knowledge in the field of analytics.

Like everything else that is worth doing, you need to devote some time to learning how to use game analytics correctly. So, everything in order:

1. Basics: first you need to understand what analytics is, what role it plays in game development and management. Analytics is a larger tool than, for example, monetization. Analytics is a tool for the whole company, not just for a specific game.
2. Key terms: at the second stage, you need to get acquainted with the key terms of game analytics and their meaning. Due to the novelty of using analytics in games, there are some differences in how the basic terms are used. Without understanding the key terms, it is much more difficult to find the necessary information.
3. Reading: At this stage, you should find the key materials available on the Internet. This may seem a bit strange, however, it is very important to research the experiences gained by other people before thinking about what to do in your own game.
4. Process: It is necessary to study the process of analytics from the moment when questions arise to getting answers to them and changing the design of the game or the company’s working methods. The analytics process is influenced by a number of key factors, in particular, the specific requirements of stakeholders.
5. Goals: After learning what game analytics is and getting an idea of how it is used within the gaming industry, you need to determine why you need it. Analytics is such a broad field that it can be applied at the strategic and tactical levels. Awareness of the full range of possibilities and setting goals is already an achievement, but at the initial stage of implementation and testing they must be strictly coordinated in accordance with a narrow range of tasks. This process provides for an evolutionary prototyping strategy, in which analytics is improved in an iterative mode, and the need for initial investments is minimized. Setting goals can be an incredibly difficult task, especially at the earliest stages of design, so iterative development helps to gain the necessary experience in the learning process, which is very useful for those who are just starting to work with analytics.
6. Planning: having set goals, it is necessary to develop a strategy to achieve them. At the practical level, this process involves figuring out exactly what data needs to be obtained in order to achieve goals, then how to obtain and analyze them in order to achieve concrete results, how to visualize them and present them to interested parties. This stage includes several steps, each of which covers consideration, selection, planning of the message hierarchy, stakeholder requirements, analysis and visualization. It may seem that you will have to spend a lot of time planning before you start collecting data, but this is not the case: the complexity of planning is directly proportional to the complexity of the goal. If all you want to know is how many players are competing in your game, planning will be very simple. If you want to be able to engage in 3D activity maps, profiling players through cluster analysis, or tracking hundreds or thousands of behaviors, the information retrieval process will be more complex, and its planning will take some time. As mentioned above, the use of an evolutionary prototyping strategy – starting small and gradually increasing momentum – is recommended for all non-specialists or when working with unfamiliar types of games.
7. Gathering information: After setting clear analytics goals, it’s time to collect data. There are many ways to do this – the development of an embedded system, the use of third-party tools, etc. Let’s assume that a certain data collection system has been implemented.
8. Analysis: it’s time to analyze the collected data. Again, there are many ways to cope with this task. Often simple statistical methods, such as calculating averages or sums, provide all the necessary information, in other cases you will have to turn to multidimensional statistics or methods of machine understanding of data.
9. Visualization: After the analysis, the data should be visualized so that they are understandable to the audience. If the analyst himself is this audience, this process will be relatively simple, since all the numbers are familiar to him. However, people of different professions, as a rule, think differently. For example, while an analyst will be happy to get a data table, a designer may need a visual representation of it. A well-known example of visualization in the industry is the intensity map, which clearly reflects, for example, the death of heroes inside a virtual environment.
10. Reporting: Now it’s time to show the results of the work done to the team, managers, marketing colleagues, user researchers, testers, etc.
11. Implementation: having received any satisfactory results, it is possible to formulate recommendations on how to change the design of the game, the process of work or anything else. This work is rarely assigned to analysts – they report on the results, but decision-making is the prerogative of experts in various fields. For example, if it was found that players had to work hard to pass a certain boss because they did not find a hidden rocket launcher, the analyst usually presents the result, but leaves it to the designers to decide how best to make this grenade launcher more visible. It is known that Microsoft Studios research is based on the creation of successful and close interaction of teams engaged in user research, testing, analytics and design.
12. Maturity: after going through the first cycle of planning, data collection, analysis, you can take a step forward – maybe it’s worth tracking more data, using more sophisticated analysis methods, conducting an experiment with interactive data visualization, reading about analytics in other industry sectors and seeing if there are ideas there that are useful for your specific situation, etc. Opportunities of unlimited scale are open to you.

Anders discusses each of these stages in his posts on the Game Analytics blog. We will publish them in the future. 

Editor’s note: The original was written by Anders Drachen, chief game analyst at Game Analytics. He is one of the most published experts on the issues of game analytics, game data collection, research, the object of which are users, etc.

Source: blog.gameanalytics.com