Skip to content

krazyeone/loto-mindhunter

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🎲 Loto Mindhunter

Statistical analysis engine and prediction system for Romanian Lottery (6/49)

Python Flask SQLite License

A Python/Flask web application that analyzes 60+ years of Romanian Lottery draw history to generate statistically-informed number combinations. Combines classical statistics, frequency analysis, and multiple prediction models with a schema-48 wheeling system.


Features

  • 60+ years of historical data — 1958 to present, auto-updated twice weekly via scraper
  • Multiple prediction models — frequency analysis, gap/overdue detection, hot/cold number tracking, co-occurrence chi-square
  • Schema 48 wheeling — 12 guaranteed combinations from 9 chosen numbers (official Romanian coverage system)
  • Walk-forward backtest — validate model performance against out-of-sample draws
  • Noise filter — eliminates statistically improbable combinations
  • Web dashboard — live stats, predictions, and combination generator via Flask UI
  • Auto-update — cron job scrapes noroc-chior.ro every Thursday and Sunday at 21:00

Tech Stack

Component Technology
Backend Python 3.10+, Flask
Database SQLite (60k+ rows)
Data scraping requests, BeautifulSoup
Statistics scipy (chi-square), collections, numpy
Frontend Server-side rendered HTML templates
Scheduling cron / systemd timer

Architecture

app.py                  Flask entry point — routes, API endpoints
modules/
  ai_predictor.py       Hot/cold/gap frequency analysis
  noroc_predictor.py    Pattern recognition for Noroc Chior variant
  combo_generator.py    Filtered combination generator
  wheeling_system.py    Schema 48 — 12 tickets from 9 numbers
  stats_analyzer.py     Walk-forward backtest, chi-square co-occurrence
  noise_filter.py       Statistical filtering of weak combinations
  dashboard_v2.py       HTML dashboard renderer
scripts/
  fetch_noroc_chior.py  Web scraper (twice-weekly auto-update)
  rebuild_db_v2.py      Database rebuild + deduplication
data/
  loto_mindhunter.db    SQLite — main history table
  noroc_chior_history.csv  CSV backup

Quick Start

git clone https://github.com/yourusername/loto-mindhunter.git
cd loto-mindhunter

pip install -r requirements.txt

# Seed the database (requires historical CSV)
python3 scripts/rebuild_db_v2.py

# Start the web server
python3 app.py
# → http://localhost:5001

API Endpoints

Endpoint Description
GET / Main dashboard with stats
GET /api/predict Run all prediction models
GET /api/schema48?numbers=1,2,... Generate Schema 48 tickets from 9 numbers
GET /api/stats Frequency, gap, hot/cold data
GET /api/backtest Walk-forward backtest results

Schema 48 Explained

The official Romanian Schema 48 guarantees a 3-number match if all 6 winning numbers fall within your chosen 9. It produces exactly 12 tickets covering all required combinations.

pool = [5, 12, 18, 23, 31, 37, 42, 44, 48]  # your 9 chosen numbers
tickets = schema_48_variants(pool)            # returns 12 tickets

Data Sources

  • Primary: noroc-chior.ro — scraped automatically
  • Historical: CSV import from official Loteria Română archives (1958–2024)

Requirements

flask>=3.0
requests
beautifulsoup4
scipy
numpy

Built with Python · Statistics-first approach · No magic, just math

About

Statistical analysis engine and AI prediction system for Romanian Lottery 6/49 — Flask, SQLite, 60+ years of history

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages