HarnessTube System Architecture

System Purpose

HarnessTube uses an automated pipeline to import harness racing race cards and generate broadcast graphics for live streams.

Data Pipeline

USTA Race Card ↓ CTRL + U HTML copy ↓ USTA Import Tool ↓ Race data parsed ↓ Google Sheets database ↓ Published CSV ↓ import_sheet.php parser ↓ Overlay Dashboard ↓ JSON files written ↓ Overlay graphics update

Core Components

USTA Import Tool Parses race card HTML and extracts race data.
Google Sheets Database Stores race cards by track tab.
Overlay Dashboard Loads race data and controls broadcast graphics.
JSON Data Files Provide live data to overlay graphics.
Overlay Graphics Display race information during broadcast.
Results System
Publishes official race result JSON files and powers public race result pages.
Statistics Engine
Scans all result JSON files to generate driver, trainer, and future analyst statistics.
Public Website Pages
Public-facing Results and Stats pages hosted on the main HarnessTube website.
Overlay Data Subdomain
Overlay JSON files and APIs are hosted on overlays.harnesstube.com.

Google Sheets Structure

Each track is stored in its own sheet tab. Race blocks contain metadata, horses, and analyst selections.

Operational Workflow

Entries posted ~3 days before race
Race cards imported via USTA Import Tool
Programs verified against printed program
Track tab published and mapped in PHP
Broadcast operator loads race in dashboard

Phase D – Results & Statistics System

Phase D expands the system from live broadcast graphics into a full results and statistics platform.

Race Results (Dashboard Publish) ↓ JSON files written to /data/results/ ↓ System scans all result files ↓ Aggregate statistics generated ↓ Results pages + standings + records

Data Source

All race results are stored as individual JSON files:

/data/results/{location}_{date}_race_{race}.json

These files contain race outcomes, drivers, trainers, and finish positions.

Key Principle

JSON result files are the single source of truth. Google Sheets is not used for statistics processing.

Planned Features

Results Web Pages
View results by track and date.
Driver Statistics
Wins, starts, win percentage, leaderboards.
Horse Records
Performance tracking and historical results.
Analyst Scoring
Compare selections vs results and track standings.

Data Processing Approach

Scan all JSON result files
Aggregate data dynamically
No manual data entry required

Future Automation Goals

Auto-update driver championship Google Sheet
Auto-update analyst scoring sheet
Generate leaderboard pages dynamically

Constraints

Do not modify results JSON structure
System must tolerate incomplete race data
No dependency on Google Sheets for stats

HarnessTube Broadcast System Architecture