ENGINE V1.0 UNITY C#

BALOOT ARCHITECT

A technical deep-dive into the state machine, heuristic AI, and card interaction logic powering the game.

01. Core Classification

GameManager

SINGLETON

The central nervous system. Orchestrates the Game Loop (Mukabara -> Dealing -> Playing) using Unity Coroutines.

Instance (Static)

State Machine

BotBrain

STATIC LOGIC

Pure logic container. No state storage. Evaluates hands for Buying and determines optimal Card Play.

EvaluateBuy()

ChooseBestCard()

Card Entity

MONOBEHAVIOUR

Hybrid data-visual object. Handles Vector3.Lerp movement and stores Suit/Rank enumerations.

Interactable

Rank Value

Execution Lifecycle

COROUTINE SEQUENCE

Mukabara Dealer Selection

Visual Cut Deck Split

BUY

Buying Phase Sun / Hokom

Playing 8 Tricks Loop

Result Score Calculation

Initialization Decision Making Runtime

03. AI Decision Structure

EvaluateBuy() Logic Tree

Input: Hand + Ground Card

Has Winning Project?

400 (Aces) OR 100

SUN Logic

Points >= 20

BUY SUN

HOKOM Logic

Jack + 9 + Ace

BUY HOKOM

BotBrain.cs

// Evaluating whether to buy or pass
public static int EvaluateBuy(List<Card> hand, Card groundCard) 
{
    // 1. Check for '400' or '100' projects
    if (HasWinningProject(hand)) return 1; // Buy Sun

    // 2. Calculate Sun strength (A=11, 10=10...)
    int strength = CalculateSunHandStrength(hand);
    if (strength >= 20) return 1;

    // 3. Check for specific Hokom combos
    if (IsGoodHokomHand(hand, groundCard.suit)) 
    {
        return 2; // Buy Hokom
    }

    return 0; // Pass
}

04. Data Models & Values

Card Rank	Sun Points (San)	Project Capability	Hokom Role
ACE (7)	11 Points	400 (4 Aces)	Strong (3rd)
TEN (3)	10 Points	100 (4 Tens)	Moderate
KING (6)	4 Points	100 (4 Kings)	BALOOT (K+Q)
JACK (4)	2 Points	100 (4 Jacks)	HIGHEST (Akbar)
NINE (2)	0 Points	-	2nd Highest