Inspiration
In competitive esports, raw match data is abundant, but turning that data into clear, actionable strategy is still difficult and time-consuming. Coaches and analysts often have limited time to manually review games, identify patterns, and prepare scouting notes.
This project was inspired by the idea of automating that process — transforming historical match data into a concise scouting report that highlights tendencies, weaknesses, and strategic opportunities against an upcoming opponent.
What it does
The Automated Scouting Report Generator analyzes historical match data for a given opponent and produces a structured scouting dashboard.
It highlights map-based performance trends, player-level impact metrics, draft pick and ban tendencies, win conversion rates, and loss collapse patterns.
Most importantly, it converts these insights into actionable recommendations that teams can directly use for preparation, drafting, and game planning.
How we built it
The project was built using Python as the core language. Pandas is used for data processing and aggregation, while Matplotlib is used for visualizing trends and comparisons.
The dashboard itself is implemented using Streamlit, allowing the insights to be presented in an interactive and easy-to-read format.
Historical esports match data provided through GRID is processed to generate map tendencies, player statistics, draft behavior, and win/loss conversion metrics.
Challenges we ran into
One major challenge was deciding how to present information without overwhelming the viewer. Raw data alone is not useful, so careful decisions had to be made about which metrics to visualize, which to summarize, and which to turn into key performance indicators.
Another challenge was structuring the dashboard so that it tells a clear story — moving from high-level trends down to specific, actionable strategies.
Accomplishments that we're proud of
We successfully built a fully functional automated scouting tool that goes beyond simple charts and tables.
The dashboard not only analyzes data, but also surfaces meaningful insights such as draft tendencies, star player identification, and win conversion vulnerabilities.
The final result feels like a real analyst tool rather than a static data visualization.
What we learned
Through this project, we learned how to design analytics tools with clarity and intent.
We gained experience in transforming raw esports data into structured insights, choosing appropriate visualizations, and building dashboards that support decision-making rather than just displaying numbers.
What's next for Automated Scouting Report Generator
Future improvements include extending the system to support multiple teams and matches automatically, adding real-time or near-real-time updates, and incorporating predictive modeling for draft and strategy recommendations.
With additional data, the platform could evolve into a comprehensive assistant tool for coaches and analysts across multiple esports titles.
Log in or sign up for Devpost to join the conversation.