Announcing the PureSkill.gg Competitive CS:GO Gameplay Data Set

Do you know how many CS:GO matches Coach watches every day? It’s a lot, and ’boy are those photocells tired. This year so far we’ve collected over 60k matches and counting. Today we’re giving it all back to you!

The PureSkill.gg Competitive CS:GO Gameplay Data Set is immediately available for public use, at no cost, under a non-commercial, attribution, share-alike license on the AWS Data Exchange.


Stand back, I'm going to try science!

🍋 Why can’t I hold all this data?

Each match has about 50,000,000 points of data, so that comes to 3,000,000,000,000 data points to build models, visualization, and more!

⚡ It’s alive!

The more you play, the more it grows every day.

🧑‍🔬 Did we mention we have a PhD?

The dataset is battle tested in production. It’s clean, optimized for productivity, and built for data scientists, by data scientists.

🔐 We take privacy seriously.

All player data is thoroughly anonymized: it is impossible to determine a player’s identity, virtual or otherwise.

🧩 But why though?

We want to see the amazing stuff you’ll do with it! Working with AI researchers who are using CS:GO to solve really hard problems inspired us to contribute back and unlock this potential for anyone.

💡 We are here to help.

Join the conversation in the new Data Dojo channels on Discord.

🚀 Let’s go!

We hope this work can fuel everything from one’s first data science experience, to their next hackathon, school project—all the way to groundbreaking research.

Now, you have the power to build incredible things to share with the community!

READY TO BE A NINJA?

It’s Fun!