Whenever I watched others play Backgammon, I wanted to remember certain positions to analyze them later with GNU BG (and later BGBlitz). Especially when I would have played a different move or chosen take instead of pass.
I used to carry a paper notebook with a pre-drawn board, but quickly noting a position was stressful since the game wouldn't wait for me.
On top of that, manually entering these positions into GNU BG was tedious before I could finally analyze them and understand why the player didn't choose my move.
Many years later, technology now offers ways to solve this problem in a much more efficient way.
- I simply take a photo of the board (or a screenshot).
- A tool analyzes the image and recognizes the Backgammon position.
- It asks for missing or unclear information, such as:
- Money game or match play?
- Current match score?
- Where is the cube? What level?
- Is the home board on the left or right?
- If checkers are missing, are they played out or on the bar?
- Once everything is clear, the tool converts the detected position into a XGID/GNU ID.
This ID can then be used in BGBlitz, GNU BG, or XG Gammon for further analysis.
- The tool should run locally on all major platforms (Mac, Windows, Linux, iOS, Android).
- I don’t want to reinvent the wheel from the start—so I’ll build upon existing tools where possible.
- Image recognition will be done using OpenCV, later replaced by a custom-trained AI model.
- The frontend is still undecided.
- Since I currently have zero experience in these areas, I’ll start by taking a Python course.
- All documents (including temporary drafts) will be stored on GitHub.
- I will document progress, challenges, and setbacks in the form of a project journal.