How Analytics Are Changing The Game

Gaming analytics is currently one of the most exciting applications of big data, so we sat down with some experts to discuss the future of the industry


Gaming analytics has steadily been rising in popularity over the last two decades, but it has really exploded in the last few years. There are currently over 1.8 billion gamers in the world and over 3 billion hours of gaming occur weekly in the U.S alone. This creates a lot of data to be analyzed, and more and more companies are developing new and exciting ways to utilize it all.

To explore this, we spoke to some of the speakers featuring in our upcoming Gaming Analytics Summit in San Francisco; Thomas Dobbs and Bysshe Easton from KIXEYE and Pallas Horwitz from Skillz. We discussed what they are currently up to in the gaming analytics field and where they see it heading in coming years.

Pallas Horwitz is the Data Science Manager at Skillz, a worldwide leader in mobile eSports, which was recently named the fastest-growing company in America. She focuses on marketing analytics, user segmentation, acquisition models, and economy optimization to build user-friendly data tools.

Bysshe Easton and Thomas Dobbs both work for KIXEYE, the California based video game company. Easton is the Director of Analytics while Dobbs is Data Science Product Manager. Between them, they are responsible for maintaining and building machine learning models from ideation to full implementation, data analysis, economics, intuition and game design sense to solve monetization and content delivery problems in games.

Do you think analytics is driving gaming? Some argue that analytics is taking the joy of out gaming, do you believe there’s a risk of that happening when it’s done badly?


Analytics does drive much of product development for gaming and has helped improved a lot of product elements via quick data feedback loops. There’s certainly an argument that narrow-minded analytics can hinder a player’s experience. However, primarily, analytics serves to improve the user experience by building a more engaging game.


I think well-done analytics enhances gaming. There have certainly been games that have features that are analytically driven with the sole purpose of increasing monetization. Whenever a feature is not designed with the player experience in mind, the joy of gaming will be lost. I agree that poorly done analytics can have a negative impact on the gaming experience, but I think this applies to all aspects of game development. Poorly done game design, shoddy graphics, and a clunky user interface can also decrease player enjoyment.

What are you doing to better understand customer behavior in the gaming industry?


We utilize a few machine learning models to attempt to find specific pain points in the game for users or find in-game features that are highly correlated with churn or increased user revenue.


Luckily, the people that work at Skillz are passionate about gaming. We all play games every week. We share white papers and articles that other companies are publishing about gaming. I want to make sure my team is focused not just on the quantitative, but also the qualitative. The best insights are supported both by analytical insights and hands-on customer research.

And finally, how have advances in machine learning assisted with gaming analytics?


Improvements to open-sourced Machine Learning programming tools have lowered many of the barriers to building effective models in production environments. With these lowered barriers we can quickly build, iterate, and implement models in a streamlined fashion, which beforehand was much more tedious. These advances also help us expand the type of in-game problems we can attempt to address using unique techniques like sentiment analysis and deep learning.


When I started in gaming, we were focused on understanding A/B tests and how to do the most robust A/B test analysis. Today we are simulating the effects of our A/B using bots that behave the same way that our players do. We can get a good read on whether a test is worth further effort, without a single player having to experience a potentially disruptive A/B test. If our simulations pan out, then we move forward with an A/B test because it is always important to validate one’s hypothesis.

For a much more in-depth look at the role of gaming analytics and how it is reshaping the industry with data insight, check out the upcoming Gaming Analytics Summit happening in San Francisco 18-19 April. 


Read next:

Why We Need Data Visualization To Understand Unstructured Data