张桥雨

02 / 10

确定性网格战棋／解谜规则DETERMINISTIC GRID TACTICS / PUZZLE

《衍》

一个确定性网格玩法原型：先观察和预演，再在棋子自身能力内选择行动，改变后续路径。A deterministic grid prototype: observe and predict, then choose an action within a piece’s abilities to change the route that follows.

个人项目：规则设计、教学关设计、Unity/C# 确定性模拟、关卡数据与编辑工具。Solo project: rules, tutorial levels, deterministic Unity/C# simulation, level data, and editor tooling.

观看规则推演动画WATCH RULE ANIMATION↓
↓ 先看项目SHOW ME THE WORK
RULE ANIMATION规则推演动画，非实机录像：记录“前进失败后掉头”的纸面推演。Rule animation, not gameplay footage: a paper simulation of turning after a failed forward move.

默认静音播放，可用播放器控制暂停或开启声音。Plays muted by default. Use the controls to pause or enable sound.无法播放？打开视频文件 ↗Trouble playing? Open the video file ↗

01

语境与问题CONTEXT + PROBLEM

先定义玩家真正需要判断什么。Define what the player actually needs to judge.

这个方向关注“脑内 play”：玩家先观察、预测和预演，再进行有限输入。当前围绕一个基础 Pawn 探索规则与教学，还不是完整战棋。The idea centers on mental play: observe, predict, and rehearse before making a limited input. Current rule and tutorial work focuses on one basic Pawn; this is still an early tactics prototype.

规则的趣味依赖“先看懂，再破局”。如果教学一开始就要求操作，玩家无法建立默认路径；如果解开后还要机械等待，关键转折又会被冲淡。The rule depends on an observe-then-intervene rhythm. Asking for input too early hides the default path; making the player wait after solving dilutes the decisive turn.

02

关键决策KEY DECISIONS

规则不是功能清单，而是一组取舍。Rules are tradeoffs, not a feature list.

  1. 01

    只替换意图，不直接指定结果REPLACE INTENT, NEVER DIRECTLY SET THE RESULT

    玩家只能选择棋子本来拥有的行动，不能让它做出超出自身能力的动作。这条规则用来建立棋子差异。The player chooses only actions the piece already owns, preserving one shared rule set for every intervention.

  2. 02

    先用无操作关卡展示因果链TEACH THE CAUSAL CHAIN WITHOUT INPUT FIRST

    已有教学关草案先用自动行动展示接触水晶的过程，再加入一次右转。玩家可以把干预后的路径与默认结果对照。The tutorial drafts first show automatic movement toward a crystal, then introduce one right turn. The player can compare the altered route with the default outcome.

  3. 03

    删除解开后的机械收尾REMOVE THE MECHANICAL TAIL AFTER THE SOLUTION

    教学关 1 的 V0.2 草案缩短了转向后的自动行走，目的是让玩家更快看到这次选择的结果。The V0.2 draft of Tutorial 1 shortens the automatic walk after the turn, aiming to show the result of that choice sooner.

教学关 0 首次接触水晶 / Tutorial 0 first crystal contact
01教学关 0：先观察默认行动链。Tutorial 0: observe the default action chain first.
教学关 1 右转入口 / Tutorial 1 right-turn setup
02教学关 1 V0.2：一次右转改变后续路线。Tutorial 1 V0.2: one right turn changes the route.
03

结果与反思OUTCOME + REFLECTION

把实现、验证和下一次判断连起来。Connect implementation, validation, and the next judgment.

目前一个基础 Pawn 接近可运行，仍有交互细节未完成。这里保留规则推演动画与教学关截图，用来说明设计过程；尚未进行外部玩家测试，也没有公开试玩入口。One basic Pawn is close to runnable, with interaction details still unfinished. The rule animation and tutorial screenshots here document the design process. There has been no external playtest, and no public playable build is available.

真正需要玩家测试的不是规则能否运行，而是无文字教学能否让首次接触者在正确时刻形成“我看懂了”的判断。The key player-test question is not whether the rules run, but whether wordless onboarding produces the “I understand it” moment at the right time.