AI-Powered Horse Racing Prediction Platform

Project Overview

The NagMee platform combines artificial intelligence, and comprehensive data analysis features to provide horse racing enthusiasts with race predictions and personalised watch list management.

The platform analyses thousands of historical race results to identify winning patterns and generates predictions for upcoming fixtures, whilst allowing users to track their favourite horses and receive notifications.

Key Features

AI-Powered Predictions

• Multi-factor analysis system evaluating 10+ performance indicators
• Real-time prediction scoring with confidence ratings
• Automated prediction storage and accuracy verification
• Historical performance tracking and pattern recognition

Smart Watch List Management

• User-customisable horse tracking
• Automated email notifications for upcoming races
• Configurable notification timing and preferences
• Race fixture calendar integration

Comprehensive Data Analysis

• Form analysis over recent performances
• Course-specific historical performance
• Distance suitability assessment
• Jockey and trainer partnership success rates
• Going condition preferences
• Rating progression trends
• Rest period optimisation
• Margin and time performance metrics

Performance Tracking

• Prediction accuracy statistics
• Venue-specific accuracy analysis
• Win/place prediction scoring system
• Historical accuracy trends

Technical Deep Dive

Architecture & Data Pipeline

The platform is built on WordPress with a custom plugin architecture that handles data ingestion, processing, storage, and frontend delivery. The system processes racing data through multiple stages:

1. Data Acquisition: Automated CSV imports from racing data providers, supplemented by fixture API integration for real-time race information
2. Data Normalisation: Incoming race data is parsed, validated, and stored in a relational database structure with dedicated tables for horses, fixtures, races, runners, results, ratings, and predictions
3. Machine Learning Pipeline: Historical race data is processed through a scoring algorithm that evaluates 10 weighted performance factors per horse
4. Prediction Generation: The system generates predictions for upcoming races using the Traditional Analysis method (52.1%+ win accuracy), which combines:
   • Recent Form – Performance trends over last 90 days
   • Course Performance – Historical venue success rates
   • Distance Suitability – Performance at similar race distances
   • Rating Trends – Official ratings progression
   • Margin Progression – “Buttons beaten” analysis from racing data providers
   • Rest Period – Optimal days since last race
   • Jockey/Trainer Partnership – Combination success rates
   • Going Performance – Ground condition preferences
   • Time Performance – Race time analysis
5. Prediction Verification: Post-race results are automatically verified against stored predictions, calculating accuracy metrics and updating performance statistics

Note: All predictions are for informational purposes only. The system continuously learns from results to improve future accuracy, but past performance does not guarantee future results. Users should always conduct their own research and bet responsibly.

Visit the website: https://nagmee.co.uk