Marjon Cajocon

HomeApps › Brazilian Checkers

Brazilian draughts

Brazilian Checkers

8×8 Brazilian rules with international capture logic, on a 256-wide NNUE.

NNUE neural network MMC Solo Dev

The engine

The evaluation is an NNUE neural network — 256 hidden units over 128 input features, 4 phase buckets on piece count — written in C and compiled straight into the app. It is trained from self-play by a Go trainer, quantised to 16-bit integers, and evaluated incrementally so the search can afford to call it at every node.

The rules of this variant

Brazilian draughts is played on an 8×8 board with international rules: men capture both forwards and backwards, kings fly along a diagonal, and you must always play the capture that takes the most pieces.

What you get

  • Play offline against a neural-network engine — no account, no connection
  • A real 3D board rendered with three.js, alongside the classic 2D board
  • Analysis board with an evaluation bar and best-move hints
  • Game review: step back through a finished game move by move
  • Custom position setup, pass & play on one device, and AI vs AI
  • Board and piece themes, and a rules panel for the variant

Screens

Click any image to view it full size. The 3D shots are the real board renderer, captured from the same page the app ships.

Marjon Cajocon

Software engineer · Talibon, Bohol, Philippines

I build game engines in C and the apps that ship them. Seven board-game engines, six of them running NNUE neural networks I trained myself, plus a from-scratch LLM engine written in dependency-free C.

Developer
MMC Solo Dev — the name my apps ship under on Google Play and the Microsoft Store
Primary
C — engines, evaluation, search, the LLM stack
Then
Go · Flutter (Dart) · Python · TypeScript / JavaScript
Certification
EDP Specialist — Civil Service, rated 94.65% (80% to pass)
Published
Six apps on the Microsoft Store, five on Google Play
Contact
marjoncajocon08@gmail.com
Code
github.com/marjoncajocon