G-Loot leads esports with real-time data and analytics

Ernesto Ongaro, Analytics Engineer, Looker at Google Cloud

May 27, 2021

G-Loot, a Stockholm-based esports startup, is shaking up the world of electronic sports by challenging the status quo idea that only the top players can win prize money in a video game competition. They’re making it possible for gamers to improve their skills and get rewarded playing games they already enjoy. As an overlay on top of a player’s game, G-Loot receives an enormous amount of data from multiple sources.

To make data-driven decisions about the way they analyze, monetize, and enhance the player experience, G-Loot turned to Google BigQuery for its cloud data warehouse and Looker for business analytics. During a recent webinar, Chief Growth Officer at G-Loot, Jamie Dunbar Smyth, and VP of Insights George Baker joined Google and Looker experts to explore real life examples of what it means to put gaming data and analytics into practice. Participants also included Google Head of Cloud for Gaming Jonnie Ronnqvist, Looker Nordics Account Executive Robbie McKiernan, and Looker Enterprise Customer Engineer Mike Burke.

G-Loot sought a solution to manage querying large data sets

To determine the outcome of a competition, G-Loot collects in-game events like a player’s knockdowns, assists, and wins. They do this across the 15 game titles on their platform for three types of competitions: head-to-head, versus others, and solo (to reach a target). All this data adds up quickly, as Ronnqvist notes, and handling such huge amounts of data is a shared challenge among gaming companies.

“There’s a major problem here, and it’s around data volume growing significantly. It’s also about data fragmentation, which makes it difficult — and sometimes even impossible — to uncover those actionable insights,” he says.

On top of the fragmentation and sheer volume of the data, analyzing that data brings its own challenges and costs. “While BigQuery is really good at showing the amount of gigabytes in a per-query fashion on the console, it’s harder to manage when you have a number of users, a growing number of tables, automated reports, dashboards, and services all running at once,” adds Baker. Conventional partitioning and clustering allowed them to reduce query volume and, therefore, spend, but it became difficult to manage those queries.

They needed more visibility. Prior to using Looker in combination with BigQuery, G-Loot was using a third-party visualization tool. They found that because their data sets were so large, query costs were racking up.

Esports dashboards lead to better business decisions

Shortly after implementing Looker, G-Loot analysts realized substantial benefits from the Gaming Analytics Block. From there, the G-Loot team quickly built out a number of models and dashboards that other teams across the organization had been requesting. For example, one set of dashboards includes metrics on paying versus free players, entry win rations, and whether funds have locked; all split by game, and the time zone. Another dashboard for the C-suite shows monthly cohort-level data around revenue, retention, and churn (which is also split per game, region, timeframe, and time zone).

“Looker is live querying that database, and bringing back those most recent results all the time. This is very powerful when you want to give your end users access to a live analytics platform,” says Burke. The first thing a user does after looking at a dashboard is determine how the metrics change when focused on a particular subset of users. However, it’s not always possible to show the metrics and graph equivalence of every subset they have in mind. Baker found that “being able to add dashboard-wide filters that can automatically update each report based on user inputs is super useful — and really easy to implement, almost as an afterthought after dashboard creation.”

In addition to its daily key performance indicators (KPIs), G-Loot leverages these dashboards to make real-time decisions in order to improve the player’s experience at the regional level. One such use case is strategically targeting specific time zones to optimize the number of competitions to the number of active players, based on their peak play times. The outcome? The platform feels more alive and ensures that the competitions are as profitable as possible for G-Loot.

Data investigation and abuse detection made easy

Detecting and stopping cheaters requires a data-intensive investigative approach. It’s worth the effort to ensure a positive player experience and to attract and retain honest and highly skilled gamers. As an overlay on top of a player's game, G-Loot is in a unique position to guide their abuse identification strategy. This means they can see a player’s activity across multiple games.

Abusers use many tactics to gain an advantage in the game. These include in-game hacks like smurfing, where cheaters intentionally play badly or start a fresh account so they’re matched with less skilled players in a tiered competition where the cheaters can dominate. There’s also server manipulation, in which players can take advantage of a period of time where a server has no or very few players, allowing them, along with the help of a friend, to easily rack up kills, points, and wins.

Pulling real-time esports data and drilling down into specific views allow G-Loot to analyze data across multiple metrics. “We take a look at the game round duration for all games on the platform. I think it's pretty special, in that no other company would have this data, at least not for this amount of games,” Baker notes.

G-Loot found that comparing the average round duration of a player under investigation to the general population of players goes a long way toward understanding whether a player is abusing or not. While there’s no simple formula, quickly cross-referencing these various data points gives G-Loot an advantage in their anti-abuse policy.

G-Loot and real-time data for the win

The combination of BigQuery and Looker has helped G-Loot make data-driven decisions that create great player experiences. Dunbar Smyth reflects on the integration, “These tools enable our teams to be data-driven in the way they advertise, analyze, communicate, and A/B test our user experience — and optimize towards drawing insights for the team so they can better understand, and in turn, hyper-personalize that player experience.”

This has enabled G-Loot to reduce costs while providing actionable data to teams across their organization. Now they can make strategic decisions and increase profitability by tracking daily KPIs with key dashboards. And, they’re ensuring players have a superior in-game experience by identifying and catching abusers with real-time esports analytics data. As the volume of data continues to grow, G-Loot is well positioned to take advantage of these actionable insights and to continue disrupting the world of esports.