BGG Dash Viewer

A Dash + Flask application for exploring BoardGameGeek data and model predictions from a BigQuery data warehouse.

Overview

BGG Dash Viewer pairs a Flask-hosted landing page and auth layer with a Dash app that surfaces game search, predictions, similarity, and monitoring views backed by BigQuery.

Features

• Game Search — card + AG Grid table views with player-count, complexity, and year filters, inline expandable details, and search-summary chips
• New Games — recently released games with designers, publishers, and categories
• Upcoming Predictions — model predictions for unreleased titles
• Game Similarity — find similar games via embeddings, with an Explore Embeddings tab
• Game Ratings — rating distributions and trends
• Experiments — combined experiment + version selector with synced tabs for model comparison
• Monitoring — BigQuery pipeline status and deployed model display
• Auth — Flask-Login with BigQuery-backed users, bcrypt hashing, and a registration code gate

Installation

Prerequisites

• Python 3.12+
• Google Cloud project with BigQuery access
• Service account credentials

Setup

git clone https://github.com/phenrickson/bgg-dash-viewer.git
cd bgg-dash-viewer
uv sync
cp .env.example .env  # fill in BigQuery + SECRET_KEY + registration code
uv run python dash_app.py

Development

Project Structure

bgg-dash-viewer/
├── dash_app.py            # Flask + Dash entry point
├── assets/                # Dash CSS
├── static/, templates/    # Flask landing page + auth pages
├── config/                # BigQuery config
├── src/
│   ├── auth/              # Flask-Login, user storage, registration
│   ├── callbacks/         # Dash callbacks (one per feature)
│   ├── components/        # Reusable Dash components (GameInfo, pills, etc.)
│   ├── data/              # BigQuery data access
│   ├── layouts/           # Page layouts (search, predictions, similarity, ...)
│   ├── theme/             # Vizro Bootstrap theming
│   ├── landing.py         # Landing page
│   └── config.py
├── docs/plans/            # Active design docs
├── tests/
├── Makefile
└── Dockerfile

Common Commands

make app          # run the app locally
make test         # pytest
make format       # black
make lint         # ruff
make type-check   # mypy
make all          # format + lint + type-check + test
make build && make up   # docker build + run

Always use uv run python ... rather than system Python.

Deployment

Containerized via Dockerfile and deployed to Cloud Run. Required env vars include BigQuery credentials, SECRET_KEY, and the registration code.

gunicorn dash_app:server

License

MIT