Inspiration

Valorant players lack access to high level analytical tools that high level e-sports teams have access to. Our goal with ValorAnalysis is to bridge that gap, offering insights and match analysis to all players. Analyzing gameplay manually and doing manual video reviews take hours, time in which many aspiring competitive players do not have.

What it does

ValorAnalysis is a real-time dashboard providing win rate, scatter plot of teammate and defender positioning during important moments (such as bomb plant), and AI-generated match analysis to improve gameplay. These metrics are helpful for aspiring competitive players to improve team play and positioning.

How we built it

We used Gemini AI to generate a match analysis taking in positioning data. All code was written in Python and published to Github, and Streamlit was used for web-app development.

Challenges we ran into

Tweaking AI prompts turned out to be a challenge because of the amount of hallucinations that we encountered. A lot of work had to be put in to put together a prompt that gives a correct analysis of the scatterplot. Another challenge that we ran into was the limited amount of data that we could get from the Riot API. The coordinate data that was given was non-intuitive, and we had enter in game to gather in-game coordinate data to correctly scale the API coordinate data to map images that we used.

Accomplishments that we're proud of

We are proud of our web-app design and out use of Gemini to accurately analyze gameplay. We think that this app could be very useful for any player looking to improve their gameplay, professional and non-professional alike.

What we learned

We learned how to utilize the Gemini and Riot API to efficiently analyze gameplay and provide a nice and concise summary. We also learned how to use Streamlit and their library to create an interactive web app.

What's next for ValorAnalytics

Adding AI analysis functionality for all Valorant maps including functionality for the entire match, and analytics that also analyze economy usage. In the future we would love to expand on our scatterplot generator and analyze the attacker positions and attacking strategies as well.

Built With

Share this project:

Updates