Hi Streamlit Community! 
I wanted to share a side project I’ve been working on for the past few months: Win Predictor LPF, an interactive Streamlit app that predicts match outcomes for the Argentine Football League (Liga Profesional de Fútbol).
What does the app do?
The app uses machine learning models and statistical distributions (xG / Poisson) combined with real-time data scraping to analyze upcoming matches and provide:
Match Win/Draw Probabilities: Projections based on team form, home/away performance, and historic metrics.
Top 5 Exact Score Rankings: Probabilities for exact match scores.
Octagonal Radar Chart: A custom Plotly radar comparing both teams across 8 key stats (recent form, attack/defense efficiency, local/visitor stats, etc.).
Model Performance Tracker: Transparency on the model’s historical accuracy.
Design & UI
I used custom CSS to style Streamlit with a neon dark-mode aesthetic (#070b14), custom metrics, and Plotly charts that match the dark theme seamlessly.
Tech Stack
- Frontend: Streamlit (
st.markdowncustom styling, Plotly charts) - Data & ML: Pandas, NumPy, Scikit-learn / Joblib, BeautifulSoup
Try the App
- Live App: https://win-predictor-lpf.streamlit.app/
I’d love to hear your feedback, suggestions for the UI, or ideas on how to further optimize performance and model visualizations in Streamlit!
Link:
Thanks for checking it out! 
