Bringing Value to Game Dev with Big Data Analytics
While playing Hades today, I was wondering: What data do I produce while playing and will it be analyzed? Kinda strange to focus on that while battling the King of the Underworld. Nonetheless, the thought is interesting and totally not a cheap excuse to write about two passions of mine: Games and data.
If you're not familiar with the issue: Big Data describes the process of collecting, processing, unifying, analyzing as well as visualizing data from a wide variety of sources. It is often used to practice business intelligence to drive business decisions and process optimization. The ultimate goal is to create a digital twin of the business.
Why you should think about data!
I’ll start by asking a few questions:
What are your monthly active users (MAU)?
What is your churn rate and did the latest update lead to improvement?
On a scale from 1-10: How happy is your community with your game?
If you can answer the questions: Great! Let's get in touch and discuss games and data! If not: No worries - Let’s get in touch!
Don’t get me wrong you should primarily focus on the game itself and transform it into the best product possible. Still, data analytics is a powerful tool to help you along this journey.
For example, you could use data to detect areas with technical problems or help with the game design. Imagine you could highlight areas with frame time issues or detect balancing problems based on player data and not only verbal feedback. You could go even deeper and run an accurate simulation based on historical player data to test a new feature or server deployment.
I believe data can also aid a developer in their community management. Based on historical data, we could build personas to identify malicious players (toxic behavior, cheating, etc…) and quickly remove them from our community. A must watch is the amazing talk about deep learning at Valve by John McDonald at GCD 2018. We could also use Discord messages to run a sediment analysis in order to prioritize development of features that seem important to players.
Additionally, I think data can also provide new opportunities to develop new products around a game. Leetify, for example, uses Counter Strike match recordings in order to analyze player performances and offer free as well as paid services. In my humble opinion, building a new service around a game feature which is already existent is quite phenomenal.
Service games with a large store could also use data from the game and combine it with other sources like Instagram, Discord, Reddit, etc. in order to personalize in-store recommendations .
Quick disclaimer: It’s a fine balancing act and can go south very fast. See Activision’s questionable patent of an algorithm to drive microtransaction with players envy.Lastly, business intelligence and data analytics can give you a general advantage over your competitors. If you have a 360° view of your game, your company and the industry as a whole you can make the best strategic decisions possible.
It is easier than you think!
The great thing about computer games is that digital data is collected and often stored by default in data structures. Games usually have a database attached to it and if you publish over Steam, you will also have access to a suite of tools that can get you started.
In addition, there are already many services like GameAnalytics or Azure PlayFab Analytics that do the heavy lifting for you and have plugins for common game engines readily available.
I hope this gives you some inspiration to think more about data and its many use cases in game development. If you want to start working with data but don’t know where to start:
Don‘t be afraid to hire a professional or discuss data with fellow developers … or shoot me an email!
Are you or your company already working with data in your game?