如何用if语句或决策树处理依赖历史操作的扑克决策逻辑？

核心前提 #

• 固定3个座位玩家，每轮手牌动态分配Button（按钮位）、Small Blind（小盲位）、Big Blind（大盲位）
• 已通过正则提取原始玩家操作序列，需校验操作合法性并判断阶段切换
• 不同位置玩家操作权限：
  • Button：仅可执行 Fold/Call/Raise
  • Small Blind：仅可执行 Fold/Call/Raise
  • Big Blind：可执行 Fold/Check/Call/Raise

阶段终止触发条件 #

需持续判断操作序列，直到满足以下任一条件，进入Flop阶段或Showdown：

• 3人局：累计2名玩家弃牌，剩余1名玩家直接赢取底池（进入Showdown）
• 2人局：1名玩家弃牌，剩余1名玩家赢取底池（进入Showdown）
• 所有活跃玩家完成跟注（无未匹配下注），进入Flop阶段
• 至少2名玩家全下，且其余活跃玩家已完成弃牌/跟注全下操作（进入Showdown或Flop，取决于是否仍有非全下的活跃玩家）

决策逻辑实现（基于If语句） #

初始化变量 #

# 初始化核心状态变量
active_players = ["玩家A", "玩家B", "玩家C"]  # 3人局初始；2人局则移除1人
current_positions = {"Button": "玩家A", "SmallBlind": "玩家B", "BigBlind": "玩家C"}  # 每轮动态更新
bet_amount = 0  # 当前需跟注的金额
pot = 0  # 当前底池金额
player_chips = {"玩家A": 1000, "玩家B": 1000, "玩家C": 1000}  # 各玩家筹码量
action_sequence = []  # 正则提取的原始操作序列，格式如[("玩家A", "Raise", 20), ("玩家B", "Call", 20)...]
stage = "Pre-Flop"

3人局决策逻辑 #

while stage == "Pre-Flop":
    for player, action, amount in action_sequence:
        # 校验玩家是否处于活跃状态
        if player not in active_players:
            continue
        
        # 获取当前玩家位置
        pos = [k for k, v in current_positions.items() if v == player][0]
        
        # 位置权限校验与操作合法性判断
        if pos == "Button":
            if action not in ["Fold", "Call", "Raise"]:
                # 非法操作，可标记或跳过
                continue
            if action == "Fold":
                active_players.remove(player)
                # 检查是否触发终止条件：2人弃牌
                if len(active_players) == 1:
                    stage = "Showdown"
                    break
            elif action == "Call":
                if amount != bet_amount:
                    # 跟注金额不匹配，非法操作
                    continue
                player_chips[player] -= amount
                pot += amount
            elif action == "Raise":
                if amount <= bet_amount:
                    # 加注金额未超过当前下注额，非法操作
                    continue
                player_chips[player] -= amount
                pot += amount
                bet_amount = amount
        
        elif pos == "SmallBlind":
            if action not in ["Fold", "Call", "Raise"]:
                continue
            if action == "Fold":
                active_players.remove(player)
                if len(active_players) == 1:
                    stage = "Showdown"
                    break
            elif action == "Call":
                # 小盲需补足与当前下注额的差额
                required_call = bet_amount - (pot // len(active_players))  # 简化计算：按当前底池均摊盲注金额
                if amount != required_call:
                    continue
                player_chips[player] -= amount
                pot += amount
            elif action == "Raise":
                if amount <= bet_amount:
                    continue
                player_chips[player] -= amount
                pot += amount
                bet_amount = amount
        
        elif pos == "BigBlind":
            if action not in ["Fold", "Check", "Call", "Raise"]:
                continue
            if action == "Fold":
                active_players.remove(player)
                if len(active_players) == 1:
                    stage = "Showdown"
                    break
            elif action == "Check":
                # 仅当当前下注额等于大盲已投金额时可Check
                if bet_amount != (pot // len(active_players)):
                    continue
            elif action == "Call":
                required_call = bet_amount - (pot // len(active_players))
                if amount != required_call:
                    continue
                player_chips[player] -= amount
                pot += amount
            elif action == "Raise":
                if amount <= bet_amount:
                    continue
                player_chips[player] -= amount
                pot += amount
                bet_amount = amount
        
        # 检查是否全员跟注
        all_called = True
        for p in active_players:
            has_called = any(a[0] == p and a[1] == "Call" for a in action_sequence)
            can_afford = player_chips[p] >= bet_amount
            if not (has_called and can_afford):
                all_called = False
                break
        if all_called:
            stage = "Flop"
            break
        
        # 检查全下触发条件
        all_in_players = [p for p in active_players if player_chips[p] == 0]
        if len(all_in_players) >= 2:
            remaining_players = [p for p in active_players if p not in all_in_players]
            # 剩余玩家是否已完成弃牌/跟注
            all_remaining_acted = all(any(a[0] == p and a[1] in ["Fold", "Call"]) for p in remaining_players for a in action_sequence)
            if all_remaining_acted:
                stage = "Showdown" if len(remaining_players) == 0 else "Flop"
                break

2人局决策逻辑 #

2人局中，Button自动兼任Small Blind，仅需调整位置判断与权限校验：

while stage == "Pre-Flop":
    for player, action, amount in action_sequence:
        if player not in active_players:
            continue
        
        pos = "Button_SB" if player == current_positions["Button"] else "BigBlind"
        
        if pos == "Button_SB":
            if action not in ["Fold", "Call", "Raise"]:
                continue
            if action == "Fold":
                active_players.remove(player)
                stage = "Showdown"
                break
            # 后续Call/Raise逻辑同3人局Small Blind
        
        elif pos == "BigBlind":
            if action not in ["Fold", "Check", "Call", "Raise"]:
                continue
            if action == "Fold":
                active_players.remove(player)
                stage = "Showdown"
                break
            # 后续Check/Call/Raise逻辑同3人局BigBlind
        
        # 全员跟注/全下判断逻辑同3人局，仅调整活跃玩家数量判断

关键说明 #

• 实际实现中需补充盲注金额的初始化逻辑（SB/BB的初始下注应提前计入底池与玩家筹码消耗）
• 操作序列的遍历需严格遵循手牌实际行动顺序（Pre-Flop阶段行动顺序为：SB→BB→Button，2人局为Button_SB→BB）
• 非法操作可根据需求标记为无效操作或终止手牌处理

内容的提问来源于stack exchange，提问作者FrakTool