Do you have what it takes to be in the game analytics industry?
Games are different beasts — they are not websites, they are not productivity apps, they are not supply chains. None of the challenges are unique, but the combination of challenges make games different when it comes to analytics.
— Anders Drachen
In our last lecture with Professor Lennart Nacke, we have learned about analytics, the field of discovering and communicating patterns in data to solve problems in business, develop predictions to support future business decisions, and improve product performance (Drachen et al, 14). Analytics is about datamining, mathematics, programming, and research, and it is an umbrella term for the process of collecting and analyzing raw data, and communicating meaningful insights based on the observed trends. Professor Nacke illustrated the applications of analytics in gaming, which is an interesting area for an analytics case study.
Since games are not concerned with productivity, game analytics seeks to measure the extent of user satisfaction and the quality of the user experience while also — oftentimes — trying to balance it with strategies to drive revenue. Depending on the complexity of the game, game analytics are most often characterized by high dimensionality, meaning that analytical data can return a high number of attributes about the users’ interactions with the numerous elements and items in the game. Moreover, with the changing business models that preference online subscription-based games and software, the realm of big data is expanding to allow us to gather analytics on user actions that were out of reach just under a decade ago. What’s particularly exciting is that when it comes to games, there are massive quantities of telemetric data waiting to be analyzed, but a shortage of individuals who can do the job. According to an article published on Forbes.com, “the global market for video games is expected to swell from $67 billion in 2012 to $82 billion in 2017. However, most video game companies appear unprepared to generate revenue from this onslaught of new customers.” What sort of information needs to be analyzed and what kind of a skillset is required to tackle the challenges that come with game analytics? I have set out on a quest to find out.
Finding the perfect fit
In a web podcast featured on DataSkeptic.com, Aalborg University Associate Professor and data scientist, Anders Drachen, discusses the realities and challenges of video game analytics. Unsurprisingly, he starts off by saying that loving games and having a passion for observing human behaviour in games are critical to anyone who wishes to pursue this particular niche of analytics. Unlike website or productivity-focused analytics, video game analytics deal with massive and extremely complex sets of data that can benefit from an interdisciplinary approach. Game analytics has been around for about 5–10 years, so there are many opportunities, but it’s still surprisingly challenging to find good data scientists to work with the analytics. It’s still not clear what qualifications are required to fit the positions, either. The good news is that many industries and areas of academia are involved in studying human behaviour in games from their unique perspectives. Human psychology, business and finance, data science, architecture, and bio-statistics are some of the specializations that investigate different aspects of player behaviour in games.
The unique challenges of video game analytics
At its core, game analytics is still analytics. Yet, what sets game analytics apart and gives it its unique challenges is the enormous variety of features that can be monitored per player, multiplied by years’ worth of collected player data. When it comes to big data, this is as big as it gets.
Game analytics are also not about achieving productivity, but rather about measuring user experience and engagement. According to Drachen, for anyone working with game analytics, it’s necessary to go into the dimensions of the game to understand user actions in the context of the game environment. For example, how do you determine what the standard should be for any given action or game outcome? How many deaths are too many, and how many are too few to make the game engaging and worthwhile over a period of time? This is where previous experience helps, but it is still incredibly subjective. There are no hard and fast rules that you can apply. With games that involve a monetary aspect, there’s also a balancing act between achieving player satisfaction while maintaining revenue. To complicate matters further, player audiences are dynamic and change over time. The best practices that were working yesterday might not work tomorrow.
What to track and why
The short answer is: It depends. The amount of information that can be tracked is immense, so deciding what to track can be a very complex question. The game analytics industry does not have clear standards because games are very different from each other, so the needs are different depending on the platform and the type of game you’re analyzing. Some games might require data about purchase metrics, average revenue, or when users are making purchases. For example, after how many hours of play do Candy Crush Saga players decide to pay for game items? What events in the game push the users to leave the game? These questions are examples of affinity mining, where we pose questions to establish connections between behaviours, entities, or objects. We look for an association between events. For example, we might learn that players who invest more time in the game are more likely to pay for items that help them advance in the game. Of course, it’s not that simple, since we also have to look at the frequency in which the two events correlate, which then helps establish the confidence level that allows us to assert the rate at which we can expect a certain correlation to occur.
Game analytics also needs to track if players are encountering any problems. Do players die too much or too soon? Is the game too hard or too easy based on how long it takes players to reach certain milestones?
For those interested in looking at games from a human psychology perspective, game analytics provide an opportunity to examine human behaviour. Drachen explains that we can gain valuable insight into how game design can shape behaviour. Analytics can show us how humans interact with other humans in a virtual environment, how communities form, and how people collaborate when they can’t communicate in a traditional sense. These insights can further help designers create meaningful game experiences.
Relationship between analytics and design
Are game designers losing creativity to analytics, or does it strengthen the user experience? On the one hand, analytics can be a wonderful source of information, and a feedback loop to help game designers fix issues in the user experience. On the other hand, analytics can be abused by taking the power away from those who ultimately use their creativity to design the game. Drachen warns that relying on analytical data too much can be dangerous. Human experience is incredibly subjective and individual, and we can’t predict how people will play the game.
Can constraining creativity based on analytical data take the playful element out of designing a game?
Works Cited
Data Skeptic. ‘Game Analytics with Andres Drachen’. Data Skeptic. N.p., 2015. Web. 7 Dec. 2015.
El-Nasr, Magy Seif, Anders Drachen, and Alessandro Canossa. Game Analytics. London: Springer, 2013. Print.
Nacke, Lennart. ‘Analytics’. 2015. Lecture.
Siranosian, Kathryn. ‘Video Game Analytics: Not Child’s Play’. Forbes. N.p., 2015. Web. 7 Dec. 2015.
Cover image: http://www.reasonforgaming.com/wp-content/uploads/2013/10/Gamer4Life.png
