Create 4point-Hoops.mdx

[rapha18th]

4point Hoops perplexity showcase.

Description

Brief description of your contribution
A basketball analytics app that showcases the power of perplexity.

Type of Contribution

• Example Tutorial
• Showcase Project
• Article/Integration Guide
• Documentation Update
• Bug Fix
• Other (please describe)

Checklist

• My code follows the cookbook's style guidelines
• I have included comprehensive documentation
• I have tested my code and it works as expected
• I have included all necessary dependencies and setup instructions
• My MDX file includes proper frontmatter (title, description, keywords)
• I have linked to any external repositories or live demos

Project Details

---

title: 4Point Hoops | AI Basketball Analytics & Insights
description: Advanced NBA analytics platform that combines live Basketball-Reference data with Perplexity Sonar to deliver deep-dive player stats, cross-season comparisons and expert-grade AI explanations.
sidebar_position: 2
keywords: [nba, basketball, analytics, ai, perplexity, sonar, sports-data, comparisons]

4Point Hoops – AI-Powered NBA Analytics

4Point Hoops turns raw NBA statistics into actionable, narrative-driven insights. By scraping Basketball-Reference in real time, cleansing the data and routing context-rich prompts to Perplexity’s sonar-pro model, it lets fans, analysts and fantasy players understand the “why” and “what’s next” – not just the numbers.

What 4Point Hoops Does

• Player analytics – season & playoff splits, shot-type breakdowns and performance radar for any NBA player
• Cross-era comparisons – e.g. Michael Jordan ’97 vs. Stephen Curry ’22 side-by-side, stat-for-stat
• Team dashboards – standings, playoff-probability Sankey flows and KPI tiles that auto-refresh daily
• AI Explain & Similar Players – one-click Sonar explanations of a stat line or AI-picked comparable athletes
• Basketball AI Chat – ask an expert LLM anything about NBA history, rosters or projections
• Credit-based SaaS plumbing – Firebase Auth + Realtime DB manage sign-up, Google login, credit wallets, admin tooling and feedback inbox

How It Uses Perplexity Sonar

# Inline “AI Explain” for a selected player/season
payload = {
    "model": "sonar-pro",
    "messages": [
        {
            "role": "user",
            "content": (
                "Explain Giannis Antetokounmpo's 2021 playoff stats in plain English. "
                "Highlight shooting efficiency and defensive impact."
            )
        }
    ],
    "response_format": {"type": "text"}
}
explanation = client.chat.completions.create(**payload)

# Free-form chat endpoint
chat = client.chat.completions.create(
    model="sonar-pro",
    messages=[
        {"role": "system", "content": "You are an NBA analytics expert."},
        {"role": "user", "content": "Who were the top 5 players in PER last season and why?"}
    ]
)

• Context packing – the backend injects cleaned, season-filtered DataFrames into the prompt so Sonar reasons over actual numbers.
• Credit guard – every AI call passes through credit_required() middleware that checks Firebase balance before deducting a point and logging the query.

Tech Stack

Layer	Tech
Frontend	React + shadcn/ui, Recharts (Netlify deploy)
Backend/API	Python Flask (Hugging Face Spaces)
Data Pipeline	BRScraper → cleanse → Firebase cache
Auth & Storage	Firebase Auth + Realtime Database
AI Integration	Perplexity Sonar-Pro, Pollinations TTS
DevOps	Docker, GitHub Actions

Links

• Frontend Repo – https://github.com/rapha18th/hoop-ai-frontend-44
• Backend Repo – https://github.com/rapha18th/4Point-Hoops-Server
• Live Demo – https://4pointhoops.netlify.app/
• Devpost Story – https://devpost.com/software/4point-hoops
• Video Walk-through – https://youtu.be/lThCWq0ij7k?si=TcLA6oP7nkhjHmjq