Notorious Studios Builds a Scalable Game Analytics Pipeline for Player Research on AWS

Notorious Studios was formed by experienced game developers who have a passion for action-based fantasy games. Some of these developers came from Blizzard and worked on popular games such as Overwatch and World of Warcraft. Legacy: Steel & Sorcery, the company’s first title, is an online fantasy-action role-playing game. Players can select characters and classes to play as, with the option to go on solo adventures or play with their friends.

When Notorious Studios began playtests for Legacy: Steel & Sorcery, it didn’t have a structured solution for tracking player data. To get gameplay insights, its design team would consolidate data manually, which could take up to 3 days. Notorious Studios wanted to combine both qualitative and quantitative data in near real time. “Since we are a small team, we needed a solution that was simple to deploy, integrate, and maintain,” says Perez.

Scalability was also a factor, because the team expects that its playtests will grow as the game gets closer to release. So, Notorious Studios used Guidance for Game Analytics Pipeline on AWS, which provides a reference architecture and sample code that helps game developers implement a codified, modular, and serverless analytics pipeline that ingests telemetry events from game clients and backend services.

The Guidance for Game Analytics Pipeline on AWS helped Notorious Studios align with AWS best practices and take a modular approach. The team used AWS Cloud Development Kit (AWS CDK) to define cloud application resources using familiar programming languages (see figure 1. Notorious Studios’ game analytics pipeline).

For cleaning and processing data, Notorious Studios uses AWS Glue, helping the company discover, prepare, and integrate all its data at any scale. To translate its data into Apache Parquet, Notorious Studios adopted Amazon Kinesis Firehose, which reliably loads near real-time streams into data lakes, warehouses, and analytics services. The team can then extract analytics using Amazon Kinesis Data Analytics to process and analyze streaming data using standard SQL.

Notorious Studios also created data overlays using Amazon QuickSight, which powers data-driven organizations with unified business intelligence. Using these solutions, Notorious Studios customized a game analytics pipeline in 2 days and deployed it in under 10 minutes. “The top benefit is how simple it was to get our pipeline running,” says Perez. “We can add and visualize data with ease by using Amazon QuickSight.”

Using analytics from its pipeline, Notorious Studios has found trends in player engagement, such as popular classes and map regions. The company uses these insights to balance the game design and deliver features that players will enjoy. “When our designers have a new question, we can query past playtests and answer them with real data,” says Perez.