A hidden-rank strategy game where intellect, deception, and tactical mastery determine who commands the field. Face an AI that reads your intentions.
Position your 21 pieces across three starting rows. Rank and tier are hidden from your opponent — use placement to deceive.
Alternate turns moving one piece per turn to adjacent squares. Move onto an enemy square to initiate combat — ranks are revealed only at the moment of battle.
Privates are the only pieces that can eliminate a Spy. Spies defeat all other ranks. The Flag cannot move — protect it at all costs.
Victory is achieved by capturing the opponent's Flag or leaving them with no legal moves. Outthink, outmanoeuvre, and outlast.
| Symbol | Designation | Tier | Special Rule |
|---|---|---|---|
| 👁 | SPY | T7 / T6 | Defeats all except Privates |
| ★★★★★ | General of the Army | T6 | Highest conventional rank |
| ★★★★ | General (4-star) | T5 | — |
| ★ | Brigadier General | T4 | — |
| 🦅 | Colonel | T4 | — |
| · | Private (×6) | T1–T3 | Only unit that defeats a Spy |
| 🚩 | Flag | T1 | Cannot move. Capture = instant win |
Game of the Generals is a digital reimagining of the classic Filipino board game — a hidden-information strategy experience where every move is a calculated risk and every battle reveals a secret.
A tribute to Sofronio H. Pasola Jr.'s original creation, this adaptation preserves the core mechanics while introducing a dynamic AI opponent. The game is designed to be accessible to newcomers yet deep enough for veterans to master.
At its mathematical core, Game of the Generals is a Zero Sum game — what one player gains, the other loses. There is no shared victory, no negotiated peace. Every decision exists in perfect opposition. But unlike chess or checkers, this Zero Sum battlefield is fought in the dark. You cannot see your opponent's army. You do not know their ranks, their formation, or their intent. This is Imperfect Information Zero Sum — the most demanding arena artificial intelligence can enter. Where classical AI can calculate, here it must also reason under uncertainty, weigh what it cannot see, and commit to action without complete knowledge. The opponent you face is not simply reacting. It is solving the game from the inside out.
We have began to tackle it.
The AI opponent uses a Bayesian belief-state engine, Alpha-Beta Search, Zobrist Hashing, RTS game modelling, Goal oriented action planning to track the probability of every hidden rank on the board. It infers your intentions — flag-rush, blitz, flank and others — and adjusts its strategy in real time. There is no blind play. It is watching.
The game features Stochastic Minimax search, an Opening Book tuned to the tier-based deployment system, and a full piece tracking database for post-game analysis.
There are 2 versions of the game: the Classic version with traditional rules, and an Advance version (soon to come) that evolves the game into an open world strategy experience.
Available on mobile. Command from anywhere.