game-brainstorm

AlterLab GameForge——结构化创意构思工作流

大多数游戏创意夭折，并非因为缺乏创造力，而是因为缺乏结构。一堆写着“如果……会怎样”的随手涂鸦，若不经过有意识的打磨，永远无法成为一款正式发布的游戏。这个工作流将原始的创意能量引导经过七个逐步聚焦的阶段，把模糊的兴奋感转化为具体、经市场验证且考虑到项目范围的游戏概念，供团队付诸实践。

这不是一份让你默默填写的模板，而是一个互动过程——每个阶段都会直接向你，也就是设计师，提出问题。你的回答将引导方向。早期阶段没有错误答案。到最后，每个答案都必须经得起市场现实、玩家心理和诚实的项目范围评估。

这七个阶段在发散思维（拓展可能性）和收敛思维（缩小范围并作出决策）之间交替进行。跳过阶段会导致可预见的问题：跳过探索阶段，你会做出自己并不在乎的东西。跳过市场验证，你会做出没人愿意买的东西——Baba Is You 能够成功，是因为 Arvi Teikari 验证了益智游戏爱好者迫切需要基于约束机制的创新，而不是因为他运气好。跳过核心循环设计，你得到的概念听起来很酷，玩起来却毫无特色。跳过项目范围评估，你就会宣布开发一款自己无法完成的游戏。

目的与触发条件

在以下情况使用此工作流：

• 开发者说：“我想做一款游戏，但不知道做什么类型”
• 有人有一个模糊的概念（“做一个和时间旅行有关的东西”），但没有围绕它进行结构化构思
• 团队需要探索多个方向再作出决定
• 用户完成了 game-start，并被引导至此进行概念开发
• 有人陷入创意瘫痪，不断在各种概念之间徘徊，却始终无法作出决定
• 开发者想要调整现有概念的方向，需要重新构思

此工作流能解决的问题：

• 从零开始时面对“无限可能”而陷入瘫痪
• 听起来令人兴奋，却没有可玩的核心玩法的创意
• 忽略市场现实的概念——在没人购买的类型中，开发没人想要的游戏
• 没有明确目标玩家类型、仿佛没有为任何人设计的游戏
• 对项目范围缺乏判断——用 5 个月的时间表设计一个需要 5 年的项目
• 在不知情的情况下意外照搬现有游戏的同质化设计

核心规则

1. 互动，而非生成。 不要替用户生成游戏概念。引导他们发掘自己的想法。提出问题。提供框架。让他们填充实质内容。你的工作是提供结构并激发思考，而不是代替创作。

2. 情感先于功能。 第 1 阶段从感受开始，而不是从机制开始。“你希望玩家获得什么体验？”应先于“他们要按哪些按钮？”功能应服务于体验，而不是反过来。

3. 限制是创意燃料。 一位只有 3 个月时间、使用免费引擎的独立开发者并非受限——而是更专注。ConcernedApe 独自制作了 Stardew Valley。Toby Fox 几乎从零开始创作了 Undertale。游戏史上最出色的作品，都是在严苛限制下诞生的。

4. 30 秒循环神圣不可侵犯。 如果核心 30 秒游戏循环脱离所有进度、解锁内容、剧情或奖励后，本身并不好玩，那么这个概念在结构上就有缺陷。Vampire Survivors 便证明了这一点——在加入任何升级之前，躲避和自动攻击的 30 秒循环就已令人着迷。

5. 不要珍视任何点子。 第 2 阶段产生的每个概念都必须经受第 3 至第 7 阶段的审视。如果经受不住，就毫不留恋地舍弃。在头脑风暴时扼杀概念，而不是等开发了 6 个月之后再扼杀。

6. 关注市场不等于向市场妥协。 个人热衷的项目也需要玩家。Outer Wilds 具有强烈的个人色彩，同时也找到了自己的受众，因为 Mobius 了解“好奇的探索者”这一市场确实存在。验证真实的人是否想玩你正在制作的东西。

7. 从一开始就诚实评估项目规模。 如果这个概念需要 20 种独特的敌人类型，而团队只有一位兼职美术，就直说。诚实评估项目规模，是对未来的自己负责。

8. 引用理论。 此处引用的所有框架（MDA、SDT、Bartle、Flow）都在 docs/game-design-theory.md 中定义。不要重新编造定义，请指向该文件。

工作流

---

第 1 阶段：创意探索

目标：在接触机制之前，探索情感与体验空间。

本阶段完全围绕挖掘真正能让设计师兴奋的事物展开。不是看起来有市场的东西，也不是 Steam 上的流行趋势，而是什么能让他们在椅子上身体前倾。

练习 1.1——兴奋点清单

提出以下问题，并仔细留意哪些问题让他们眼前一亮：

DISCOVERY QUESTIONS (ask all, discuss responses)
-------------------------------------------------
1. What is the last game that made you lose track of time? Why?
2. What is a moment in any game that you still think about years later?
3. What emotion do you most want your player to feel?
   (wonder, tension, triumph, melancholy, mischief, mastery, connection, dread)
4. What non-game experience would you love to translate into gameplay?
   (a sport, a job, a relationship dynamic, a natural phenomenon)
5. If you could only keep ONE system from your favorite game,
   which system and why?
6. What is a game that almost got it right but fumbled? What would you fix?

不要匆忙完成这些问题。每个回答都蕴含着种子。找出回答中反复出现的主题。这些主题构成了概念的情感基础。

练习 1.2——参考素材挖掘

不要局限于游戏。最具原创性的游戏往往借鉴了意想不到的来源。

REFERENCE MAPPING
-------------------------------------------------
Category       | Title              | What it contributed
-------------------------------------------------
Games          | [user fills in]    | [specific mechanic or feeling]
Films          | [user fills in]    | [visual style, pacing, theme]
Books          | [user fills in]    | [world logic, character arc]
Music          | [user fills in]    | [mood, rhythm, atmosphere]
Real life      | [user fills in]    | [system, interaction, sensation]
-------------------------------------------------

寻找不同参考素材之间意想不到的联系。一款融合了某部电影节奏与现实职业系统逻辑的游戏，已经比“像 Dark Souls，但有枪”有趣得多。

练习 1.3——限制条件梳理

限制条件划定了创作空间。诚实地梳理这些条件。

CONSTRAINT INVENTORY
-------------------------------------------------
Dimension       | Reality
-------------------------------------------------
Team size       | [1 person / 2-5 / 6-15 / 15+]
Timeline        | [jam / 1-3 months / 3-12 months / 12+ months]
Budget          | [zero / small / moderate / funded]
Platform        | [PC / console / mobile / web / VR / multi]
Engine          | [chosen or undecided]
Art capacity    | [programmer art / asset store / dedicated artist / art team]
Audio capacity  | [stock / freelance / dedicated / full audio team]
Tech skill      | [beginner / intermediate / advanced / expert]
Genre comfort   | [genres the team has shipped or prototyped before]
-------------------------------------------------

练习 1.4——反向灵感

同样有价值的是：你不想制作什么。

ANTI-INSPIRATION (ask the user)
-------------------------------------------------
- What genre are you tired of?
- What game mechanic do you find tedious even when well-executed?
- What trend in current games makes you roll your eyes?
- What type of player experience do you want to avoid creating?
- What game do people keep suggesting you should clone, and why do you resist?
-------------------------------------------------

通过排除来明确不想要的方向，能让概念更加鲜明。知道自己拒绝制作什么，与知道自己想制作什么同样有助于厘清思路。

---

第 2 阶段：概念生成

目标：通过结构化的发散技巧，提出 3-5 个不同的游戏概念。

不要满足于第一个想法。使用三种不同的方法生成多个候选概念，然后相互评估。每种方法都从截然不同的角度切入构思，这是刻意为之——可以避免视野狭窄。

技巧 2.1——动词优先设计

核心动词就是游戏本身。去掉主题、美术、剧情和进度系统。玩家在每个时刻都在做什么？

VERB-FIRST GENERATION
-------------------------------------------------
Step 1: Choose a core verb
  build, explore, survive, negotiate, compose, rescue, deceive, cultivate,
  investigate, perform, dismantle, connect, translate, terraform, orchestrate

Step 2: Add a constraint world
  "What if the player could [VERB] in a world where [CONSTRAINT]?"

  Example constraints:
  - gravity reverses every 30 seconds
  - every action has a permanent consequence
  - you can only perceive one sense at a time
  - resources are emotions, not materials
  - the world is a single continuous loop
  - you share a body with an AI that has different goals

Step 3: Test immediacy
  "Is the verb itself fun to execute, regardless of WHY you are doing it?"
  If yes: strong candidate. If no: the verb needs a more interesting expression.
-------------------------------------------------

生成 2-3 个动词优先的概念。每个概念用一个句子来描述。

技巧 2.2——混搭法 将两个本不该组合在一起的系统、类型或主题融合起来。毫不相关的创意碰撞会带来新意。

MASHUP GENERATION
-------------------------------------------------
Step 1: Pick System A (a gameplay system from any genre)
  tower defense, dating sim, courtroom drama, factory management,
  rhythm game, deck building, flight simulator, cooking show

Step 2: Pick System B (a completely unrelated system or theme)
  underwater archaeology, corporate espionage, ant colony behavior,
  musical composition, geological survey, political negotiation

Step 3: Force the connection
  "What if [System A] mechanics existed in a [System B] context?"
  "What if the core tension of [System B] was expressed through [System A] gameplay?"

  Example: "Tetris meets courtroom drama"
  --> Physical evidence blocks fall; you must arrange them to build a legal argument
      before time runs out. Misplaced evidence weakens your case. The jury reacts
      in real time.

Step 4: Evaluate surprise
  "Does this combination produce an experience that neither source could alone?"
  If yes: strong candidate. If no: the mashup is cosmetic, not structural.
-------------------------------------------------

生成 2-3 个混搭方案。组合越奇特，就越有可能催生真正新颖的想法。但奇特本身还不够——混搭必须带来涌现式玩法，而不能只是换一层新奇的外观。

技巧 2.3——MDA 反向设计

大多数设计师从机制入手，寄望于美学自然浮现。这项技巧反其道而行：先选择目标玩家体验，再推导出能够实现这种体验的机制。完整的 MDA 框架解析请参阅 docs/game-design-theory.md。

MDA-BACKWARD GENERATION
-------------------------------------------------
Step 1: Choose 2-3 target aesthetics from the MDA taxonomy
  Sensation, Fantasy, Narrative, Challenge, Fellowship,
  Discovery, Expression, Submission (zen/comfort)

Step 2: For each aesthetic, identify what dynamics deliver it
  Discovery --> exploration of unknown spaces, revealed information, "aha" moments
  Challenge --> escalating difficulty, mastery curves, meaningful failure
  Expression --> customization depth, creative tools, shareable output

Step 3: For each dynamic, design mechanics that produce it
  "Aha" moments --> fog of war, hidden connections, environmental storytelling
  Escalating difficulty --> adaptive enemy scaling, resource scarcity curves
  Creative tools --> modular building system, color palette, user-authored levels

Step 4: Assemble into a coherent system
  "A game where the player [mechanic cluster] to experience [aesthetic pair]"
-------------------------------------------------

生成 1-2 个 MDA 反向设计概念。这类设计往往最能立足于心理需求，因为它们从玩家的情感需求出发。

概念比较

运用三种技巧生成 3-5 个概念后，对它们进行比较：

CONCEPT COMPARISON MATRIX
-------------------------------------------------
Concept  | Excitement | Feasibility | Uniqueness | Team Fit | Total
-------------------------------------------------
[A]      | /5         | /5          | /5         | /5       | /20
[B]      | /5         | /5          | /5         | /5       | /20
[C]      | /5         | /5          | /5         | /5       | /20
-------------------------------------------------

选出一个最佳概念，作为后续工作的基础。不要对分数取平均值——一个“兴奋度”得分为 5/5、“可行性”得分为 2/5 的概念，与各项都得分 3/5 的概念，值得以不同方式讨论。明确讨论其中的权衡。

---

阶段 3：市场验证

目标：确认这个概念确实有玩家群体，且市场能够支撑它。

缺乏市场意识的热情，往往会让有才华的开发者花两年时间制作一款只卖出 47 份的游戏。Vampire Survivors 的成功，源于 poncle 发现了服务不足的“休闲弹幕地狱”细分市场——这些玩家想体验满屏混乱带来的多巴胺快感，却不想面对传统纵向卷轴射击游戏的操作门槛。Balatro 找到了一个无人知晓其存在的“扑克肉鸽”空白市场。Outer Wilds 则为“知识即成长”开辟了一个全新市场。这些都不是偶然。每位创作者都在开发前了解了市场格局。

练习 3.1——市场规模估算

通过同类游戏估算你的概念可触及的市场：

MARKET SIZING
-------------------------------------------------
Step 1: Identify 3-5 comparable games (same genre, similar scope, similar audience)
  Comparable       | Est. Revenue (SteamSpy/VGInsights) | Peak Players
  [game 1]         | [$X]                                | [N]
  [game 2]         | [$X]                                | [N]
  [game 3]         | [$X]                                | [N]

Step 2: Assess genre health
  Is the genre growing, stable, or declining?
  (Check Steam tag trends, itch.io tag volume, genre-specific subreddit growth)

  Growing genres (2024-2025 signal): roguelikes, cozy sims, extraction shooters
  Oversaturated: battle royales, zombie survival, generic pixel platformers
  Underserved (opportunity): asymmetric co-op, deduction games, tactile
  physics puzzlers, single-player narrative FPS

Step 3: Calculate realistic floor and ceiling
  Floor: Worst-performing comparable * 0.5 (assume you are less known)
  Ceiling: Best-performing comparable * 0.3 (unless you have a marketing edge)
  Realistic estimate: Median of comparables * 0.7

  If the realistic estimate does not cover development costs, the concept
  needs either a smaller scope or a more commercial angle. Do not ignore this.
-------------------------------------------------

练习 3.2——竞争分析

你的游戏并非在真空中发布。绘制竞争格局：

COMPETITIVE ANALYSIS
-------------------------------------------------
Direct competitors (same genre + similar hook):
  [Game]  | Strength         | Weakness         | Your differentiation
  [game1] | [what it nails]  | [where it fails] | [how you exploit that gap]
  [game2] | [what it nails]  | [where it fails] | [how you exploit that gap]

Indirect competitors (different genre, same player motivation):
  [Game]  | Why players might choose it over yours
  [game1] | [reason]
  [game2] | [reason]

The gap: What does your concept offer that NO existing game provides?
  [Specific answer -- "more accessible X" or "Y mechanic in Z context"]

If you cannot articulate the gap in one sentence, your concept is not
differentiated enough. Return to Phase 2 and sharpen the hook.
-------------------------------------------------

Obra Dinn 不只是制作了一款推理游戏——它发现，工作室纷纷放弃推理游戏类型时，玩家对这类游戏的兴趣却随着桌面游戏的发展而不断增长。这个缺口就是它的市场论据。

练习 3.3——概念验证关卡

让你的最佳概念通过三道二元关卡：

VALIDATION GATES
-------------------------------------------------
Gate 1: The Wishlist Test
  "If you saw this game's Steam page right now, would you wishlist it?"
  Ask 5 people who are NOT on your team and NOT your friends.
  If fewer than 3 say yes, the concept is not commercially compelling
  enough in its current form.

Gate 2: The Elevator Test
  Describe the concept in two sentences to someone unfamiliar with games.
  Do their eyes light up, or do they politely nod?
  Light up = hook works. Polite nod = hook is too niche or too generic.

Gate 3: The Comparison Test
  "What is this game competing against for the player's $15 and 20 hours?"
  If the answer includes a free-to-play juggernaut or an established franchise
  with a sequel coming, reassess your positioning. Compete where the giants
  are not looking.
-------------------------------------------------

如果概念未通过两道或更多关卡，就不要进入下一阶段。返回阶段 2 生成新概念，或打磨现有概念，直到它通过验证。创意不值钱，开发时间却很宝贵。

---

阶段 4：核心循环设计

目标：设计贯穿四种时间尺度的游戏脉搏。

核心循环不是功能列表，而是玩家行动与游戏反馈构成的循环，并在整个体验过程中不断重复。核心循环薄弱的游戏，无论内容、故事还是润色多么出色，都无法挽救。核心循环强大的游戏，即使几乎没有其他内容，也能撑下去。

按照四种时间尺度设计循环。每种时间尺度都必须构成完整循环： 行动 --> 反馈 --> 奖励 --> 决策 --> 下一步行动。

时间尺度 1：30 秒循环（微观）

这是游戏玩法的最小单位。任意一个 30 秒的时间段里会发生什么？

30-SECOND LOOP TEMPLATE
-------------------------------------------------
Action:    What the player physically does (input)
Feedback:  What the game immediately communicates back (visual, audio, haptic)
Reward:    What the player gains (progress, resource, information, satisfaction)
Decision:  What choice this creates for the next action
Repeat:    How the cycle feeds back into itself
-------------------------------------------------

CRITICAL TEST: Remove all progression, unlocks, story, and meta-rewards.
Is this 30-second loop still fun on its own?

If YES --> proceed. You have a core loop.
If NO  --> the concept needs mechanical redesign before proceeding.
           No amount of content wrapping fixes a boring core loop.

时间尺度 2：5 分钟循环（单次游玩片段）

一个有意义的子目标，为一段游玩时间提供结构。

5-MINUTE LOOP TEMPLATE
-------------------------------------------------
Objective:  What is the player trying to accomplish in this window?
Escalation: How does tension or complexity build across these minutes?
Resolution: What marks the end of this segment (boss, puzzle, wave, trade)?
Carry-over: What persists into the next 5-minute segment?
-------------------------------------------------

时间尺度 3：单次游戏循环（30-60 分钟）

一次完整的游戏体验，从开始游玩到放下游戏，感觉如何？

SESSION LOOP TEMPLATE
-------------------------------------------------
Session goal:    What draws the player to start a session?
Session arc:     How does difficulty/intensity curve across the session?
Session reward:  What meaningful progress was made by session end?
Session hook:    What makes the player want to come back for another session?
Quit trigger:    What is the natural stopping point? (Never trap players.)
-------------------------------------------------

时间尺度 4：成长循环（完整游戏）

贯穿整个游戏体验的宏观历程。

PROGRESSION LOOP TEMPLATE
-------------------------------------------------
Beginning:  What does the player start with? What is their initial capability?
Middle:     How does capability, complexity, or narrative evolve?
Mastery:    When does the player feel like they truly understand the systems?
Endgame:    What does the late game look like? (Avoid content drought.)
Completion: What marks the end? Is there post-completion content?
-------------------------------------------------

循环可视化

将每个循环画成一个环，而不是列表。无法在视觉上回到起点的循环，其实是线性序列，并非游戏玩法循环。

        +--> [ACTION] --+
        |               |
        |               v
   [DECISION]      [FEEDBACK]
        |               |
        |               v
        +-- [REWARD] <--+

---

阶段 5：支柱与边界

目标：定义游戏是什么，以及它不是什么。 支柱是设计中承重的承诺。每个机制、功能和内容都必须至少对应一个支柱。如果某项功能不服务于任何支柱，它要么是应该删掉的孤儿功能，要么说明你遗漏了一个支柱。

完整的支柱方法论请参阅 docs/game-design-theory.md，并使用 @templates/game-pillars.md 正式记录这些支柱。

练习 5.1——定义 3-5 个支柱

每个支柱都必须：

• 可证伪：你能明确判断某项功能是支持还是违背它
• 彼此独立：任意两个支柱都不应有显著重叠
• 令人难忘：每个支柱用一句短语表达，而不是一段话
• 可指导行动：支柱能直接指导设计决策

PILLAR TEMPLATE
-------------------------------------------------
Pillar          | "Every moment is a meaningful choice"
Supports        | Branching dialogue, build diversity, multiple solutions
Violates        | Linear corridors, optimal builds, single correct answers
Design test     | "Does this feature give the player a real decision?"
-------------------------------------------------

不好的支柱示例（过于宽泛，没有实际指导作用）：

• “有趣的玩法”——任何东西都可以说自己有趣，它无法指导任何事情
• “精美的美术”——这是一种愿景，而非设计约束
• “沉浸式世界”——对你的游戏来说，“沉浸式”具体意味着什么？

好的支柱示例：

• “世界会对一切做出反应”——可证伪、可测试，也能指导设计
• “失败带来教训，而非惩罚”——能直接影响难度调整
• “每种资源都有两种用途”——让每次资源开销都充满抉择

练习 5.2——定义反支柱

明确说明这款游戏拒绝成为什么：

ANTI-PILLAR TEMPLATE
-------------------------------------------------
Anti-Pillar     | "No grinding"
Meaning         | Player power comes from skill and knowledge, never from
                | repeating content for incremental stat increases
Implication     | No XP bars, no level-gated content, no repetitive farming
-------------------------------------------------

反支柱能帮你拒绝范围蔓延，让你有理由说“不”。有人提出功能时，先对照反支柱检查。拒绝比评估更快。

练习 5.3——酒吧测试

你能否在酒吧里用两句话，向一个不玩游戏的陌生人引人入胜地介绍这款游戏？如果不能，说明这个概念要么过于复杂，要么过于缺乏新意，难以快速解释。

BAR TEST TEMPLATE
-------------------------------------------------
"It's a game where you [core verb] [in what context].
 The twist is [what makes it unique]."

Rules:
- No jargon ("roguelike deckbuilder" means nothing to a non-gamer)
- No comparisons ("it's like X meets Y" is a crutch)
- Must provoke a follow-up question ("how does that work?")
-------------------------------------------------

如果酒吧测试得到的反应平淡，说明这个概念需要更鲜明的吸引点。

---

阶段 6：玩家类型验证

目标：验证这个概念是否能满足明确可识别的玩家动机。

一款为所有人设计的游戏，往往谁也吸引不了。应根据玩家的心理动机，而非人口统计特征，确定你的核心受众。

练习 6.1——Bartle 类型映射

根据四种 Bartle 类型映射这个概念（请参阅 docs/game-design-theory.md）：

BARTLE MAPPING
-------------------------------------------------
Type        | How your game serves them     | Priority
-------------------------------------------------
Achievers   | [specific systems]            | [Primary/Secondary/Not served]
Explorers   | [specific systems]            | [Primary/Secondary/Not served]
Socializers | [specific systems]            | [Primary/Secondary/Not served]
Killers     | [specific systems]            | [Primary/Secondary/Not served]
-------------------------------------------------

练习 6.2——自我决定理论检查

这个概念是否满足三种基本心理需求？

SDT VALIDATION
-------------------------------------------------
Need         | How your game delivers it          | Strength
-------------------------------------------------
Autonomy     | [player agency, meaningful choice] | [Strong/Moderate/Weak]
Competence   | [skill expression, mastery curve]  | [Strong/Moderate/Weak]
Relatedness  | [connection, shared experience]    | [Strong/Moderate/Weak]
-------------------------------------------------

概念不必在这三方面都很强，但至少要在一方面表现出色，并且不应主动损害任何一方面。

练习 6.3——明确排除对象

你不打算为哪些人设计？指出他们，并接受相应的取舍。

EXCLUSION ACKNOWLEDGMENT
-------------------------------------------------
"This game is NOT for players who want [specific experience], because
 our pillars explicitly prioritize [alternative experience] instead."
-------------------------------------------------

这不是为了排斥玩家，而是为了聚焦。试图同时满足竞技 PvP 玩家和休闲沙盒玩家的游戏，最终谁也无法满足。

---

阶段 7：范围与可行性

目标：让概念立足现实。实际能做出什么？

热情在这里遇上了现实的计算。阶段 1-5 形成的概念，必须经受团队实际能力、时间安排和技术水平的检验。

练习 7.1——定义垂直切片

垂直切片是能够证明核心循环行得通的最小可玩概念验证。它不是演示版，也不是打磨完善的产品，而是证明值得继续开发这款游戏的基础玩法的证据。

VERTICAL SLICE SCOPE
-------------------------------------------------
Must include:
- The complete 30-second loop, playable
- Enough content to demonstrate the 5-minute loop once
- Placeholder art and audio (programmer art is fine)
- One complete player decision cycle
- Basic game feel (responsive controls, clear feedback)

Must NOT include:
- Full progression system
- Multiple levels/areas
- Polished UI
- Story/narrative beyond setup context
- Monetization features
- Multiplayer (unless the core loop requires it)
-------------------------------------------------

练习 7.2——功能分级

使用 MoSCoW 方法，将所有已确定的功能分为四个级别：

FEATURE TIERS
-------------------------------------------------
MUST-HAVE (pillar features -- the game does not exist without these)
- [feature] --> supports [pillar]
- [feature] --> supports [pillar]

SHOULD-HAVE (enhancers -- significantly better with these, but playable without)
- [feature] --> enhances [pillar]

COULD-HAVE (polish -- nice if time permits, cut without guilt)
- [feature]

WON'T-HAVE (explicitly deferred -- do not build these in v1, no matter what)
- [feature] --> why it's deferred
-------------------------------------------------

“不做”清单与“必须做”清单同样重要。范围蔓延披着巧妙的伪装出现时，这份清单能让团队理直气壮地说：“我们已经决定不做这个了。”

练习 7.3——时间估算

根据团队能力粗略估算时间安排：

TIMELINE ESTIMATE
-------------------------------------------------
Phase             | Duration      | Milestone
-------------------------------------------------
Concept document  | 1 week        | Concept doc complete, reviewed
Vertical slice    | 4-8 weeks     | Core loop playable, pillar tested
Alpha             | [team-dependent] | All Must-Have features functional
Beta              | [team-dependent] | Feature-complete, bug fixing begins
Release           | [team-dependent] | Polished, tested, store-ready
-------------------------------------------------

Apply reality checks:
- Solo dev: multiply estimates by 1.5 (context switching overhead)
- First game: multiply estimates by 2.0 (learning curve)
- Unfamiliar engine: add 2-4 weeks of ramp-up time
- Part-time team: convert to calendar time using actual available hours

练习 7.4——风险识别

列出 5 个主要风险，并说明如何应对：

RISK REGISTER
-------------------------------------------------
Risk              | Probability | Impact | Mitigation
-------------------------------------------------
[technical risk]  | H/M/L       | H/M/L  | [specific plan]
[content risk]    | H/M/L       | H/M/L  | [specific plan]
[scope risk]      | H/M/L       | H/M/L  | [specific plan]
[team risk]       | H/M/L       | H/M/L  | [specific plan]
[market risk]     | H/M/L       | H/M/L  | [specific plan]
-------------------------------------------------

输出格式

此工作流将生成一份游戏概念文档，结构如下。 完整模板请参阅 templates/game-concept.md。

# Game Concept Document

## Identity
- **Working Title:** [name]
- **Genre:** [genre or genre-blend]
- **Platform:** [target platforms]
- **Team Size:** [number and roles]
- **Target Duration:** [expected play time]

## The Pitch
**Elevator Pitch (2 sentences):**
[The bar test sentences]

**Core Fantasy:**
[The central experience the player is buying into]

**Unique Hook:**
[What makes this different from everything else in the genre]

## Emotional Foundation
[Themes and feelings from Phase 1, with references]

## Market Validation
[Market sizing, competitive analysis, and validation gate results from Phase 3]

## Core Loop
[All four timescales from Phase 4, with cycle diagrams]

## Design Pillars
[3-5 pillars with support/violate examples from Phase 5]

## Anti-Pillars
[What the game refuses to be]

## Target Aesthetics (MDA)
[Primary and secondary aesthetics from Phase 2/6]

## Player Type Alignment
[Bartle mapping and SDT validation from Phase 6]

## Feature Tiers
[MoSCoW breakdown from Phase 7]

## Vertical Slice Plan
[Scope and definition from Phase 7]

## Risk Register
[Top 5 risks with mitigation plans]

## Next Steps
[Recommended immediate actions and workflow sequence]

质量标准

一次成功的头脑风暴应满足以下所有条件：

• 创意决策由用户主导，而非助手
• 至少生成 3 个不同的概念后再选定一个
• 30 秒循环通过“没有进度也有趣”的测试
• 市场验证已完成——已找出可比游戏、阐明差异，并通过验证关卡
• 设计支柱可证伪，且确实能指导设计决策
• 范围估算诚实，并明确列出不会包含的内容
• 能用 2 句话解释该概念（通过酒吧测试）
• 垂直切片可在既定约束内完成
• 坦诚列出风险，而不是盲目乐观地轻描淡写
• 概念文档足够完整，可以交给一位未参与头脑风暴的开发者，让其理解要构建什么
• 所有框架引用（MDA、SDT、Bartle、Flow）均按照 docs/game-design-theory.md 中的定义正确使用

示例用例

1. “我想做一款游戏，但完全不知道做什么类型。” 从阶段 1、练习 1.1 开始。“兴奋点清单”会梳理出各种主题和情绪，自然而然地指向不同的类型和机制。

2. “我有一个带重力机制的解谜游戏点子。” 跳到阶段 2，为这个概念生成不同变体，然后继续完成阶段 3–7，对最有潜力的变体进行压力测试、验证和范围规划。

3. “我的团队想一起为下一个项目头脑风暴。” 完整进行 6 个阶段，通过“约束映射”让讨论立足于团队的实际能力，并借助“概念比较矩阵”促进团队决策，避免沦为各说各话。

4. “我总是开始做项目，却半途而废。帮我坚持下去。” 重点关注阶段 5（支柱通过明确方向帮助建立投入）和阶段 7（坦诚规划范围，避免因不堪重负而放弃）。垂直切片的定义会提供一个可完成的首个里程碑。

5. “我很喜欢 Hades 和 Slay the Spire，想做一款类似的游戏。” 使用阶段 1 的“参考作品挖掘”，弄清楚这些游戏具体在哪些方面打动了你；然后使用阶段 2 的“混搭法”，寻找独特切入点，避免意外做出克隆作品。“反灵感”练习可以防止设计流于模仿。