davepoon/buildwithclaude / plugins/all-skills/skills/mcm-o-guide
mcm-o-guide
VerifiedMCM/ICM 优秀论文写作全流程指南——基于 502 篇历年获奖论文(2004–2024)语料库的结构化分析与评委点评提炼。覆盖:摘要逐句模板、6 类题型模型选择矩阵、25 页页预算分配、图表黄金法则、灵敏度/稳健性分析范式、全文一致性检查清单。当用户准备美赛、写论文、打磨论文、查结构、选模型、做摘要、看评委标准、读优秀论文时使用。也适用于任何需要高质量数学建模论文写作的场景。
playbooks/writing-playbook.md
View on GitHub# 写作手能力强化手册
> 基于 6 篇 O 奖论文的写作结构分析,提炼写作能力的实战模板。
> 目标:摘要优先,学术英文,图表规范,排版专业。
---
## 1 摘要优先策略
### 1.1 为什么摘要优先?
**评委视角:** 初筛阶段(<5 分钟)只读摘要。如果摘要不好,论文直接淘汰。
**O 奖数据(502 篇语料):**
- 普通 O 奖:摘要 29 个数字
- 顶尖 O 奖:摘要 93 个数字(3.2×)
- 2022 C 顶尖论文:摘要 117 个数字
### 1.2 摘要写作工作流
**Day 1(框架):**
```
1. 背景(2 句)
2. 模型命名(1 句)
3. Task 1-6(每 task 1-2 句 + 数字)
4. 灵敏度(1 句)
5. 优劣势(3 vs 3)
6. 关键词
```
**Day 3(完善):**
```
1. 检查数字密度(目标 50+)
2. 检查模型命名("we develop")
3. 检查任务标记("For Task N...")
4. 检查灵敏度结论("robust/sensitive")
5. 检查优劣势数量(3 vs 3)
```
**Day 4(最终):**
```
1. 摘要 ≤ 1 页
2. 词数 400-500
3. 数字 50-100+
4. 无个人信息
5. 无图表(摘要页只放文字)
```
### 1.3 摘要模板(顶尖 O 奖版)
```markdown
While [background sentence]. This paper develops the **[Model Name]**
to [core objective]. Our model consists of four interconnected
subsystems: [1], [2], [3], and [4].
For Task 1, we [action] by [method]. The key result is that
[specific finding with 3-5 numbers, e.g., "the total biomass
decreases by 72.45% under severe drought versus only 18.16%
for multi-species community"].
For Task 2, we [action] using [method]. Under [condition],
[quantitative finding with 3-5 numbers].
For Task 3, we develop the **[Self-Constructed Index Name]**
to [measure what]. Compared to [baseline], our index
[improvement with numbers].
For Task 4-5, sensitivity analysis shows that [parameter] is
[sensitive/robust]: [specific number, e.g., "varying α within
[0.5α₀, 1.5α₀] changes the output by less than 5%"].
For Task 6, we propose [specific recommendation] which achieves
[quantified effect, e.g., "a 94% reduction in [metric] over
20 years"].
The **strengths** of our model are: (1) [S1], (2) [S2],
and (3) [S3]. The **weaknesses** are: (1) [W1], (2) [W2],
and (3) [W3]—all addressable by [direction].
*Keywords:* [k1], [k2], [k3], [k4].
```
---
## 2 美赛固定行文结构
### 2.1 标准结构(25 页)
| 页 | 内容 | 字数 |
|---|---|---|
| 1 | 摘要 | 400-500 词 |
| 2 | 目录 | - |
| 3-4 | Introduction | 500-700 词 |
| 5 | 假设 + 符号 | 200 词 |
| 6-7 | Task 1 模型 | 800-1000 词 |
| 8 | Task 1 结果 | 500 词 |
| 9-10 | Task 2 模型 | 800-1000 词 |
| 11 | Task 2 结果 | 500 词 |
| 12-13 | Task 3 模型 | 800-1000 词 |
| 14-15 | Task 3 结果 | 500 词 |
| 16-17 | Task 4-5 灵敏度 | 800 词 |
| 18 | Task 6 建议 | 500 词 |
| 19 | 优劣势 | 200 词 |
| 20 | 结论 | 200 词 |
| 21 | 参考文献 | - |
| 22-25 | 附录 | - |
### 2.2 各章节写作要点
**Introduction(2 页):**
```
1.1 Background(2 段):从全球/宏观问题 → 具体到题目场景
1.2 Significance(1 段):为什么重要,用统计/经济数据支撑
1.3 Paper Structure(1 段):结构预告("Section X addresses Task Y")
1.4 Data Description(1 段):数据来源、时间范围、可信度
1.5 References(3-5 篇):核心文献
```
**Assumptions + Notations(1 页):**
```
假设 4-7 条,每条带 justification
Notation 表格:符号 | 含义 | 单位
```
**Task 章节(每 task 3-4 页):**
```
N.1 Model Development("we develop")
N.2 Governing Equations(编号方程)
N.3 Results and Discussion(图 + 结论)
```
**Sensitivity Analysis(1-2 页):**
```
独立章节,明确出现 "sensitivity" 一词
至少 3 个参数扫描
结论句:"The model is robust/sensitive w.r.t. X"
```
**Strengths & Weaknesses(1 页):**
```
3 条 Strengths(每条对应一个创新点)
3 条 Weaknesses(每条带解决方向)
```
---
## 3 学术英文写作规范
### 3.1 句式库(高频句式)
**模型描述:**
- "We develop a [model name] to [objective]."
- "The [model name] consists of [components]."
- "We improve the [classical model] by [modification]."
**结果描述:**
- "From Figure N, we find that..."
- "The results show that..."
- "As expected, [tendency]..."
**灵敏度分析:**
- "We examine the sensitivity of our model by varying [parameter]..."
- "The model is robust with respect to [parameter]."
- "The model is sensitive to [parameter]..."
**优劣势:**
- "The strengths of our model are..."
- "However, our model has some limitations..."
- "This weakness can be addressed by..."
**参考:** [sentence-patterns.md](../references/sentence-patterns.md)
### 3.2 中式英语避坑
| ❌ 错误 | ✅ 正确 |
|---|---|
| isolately | in isolation |
| classifiy | classify |
| enviroment | environment |
| the rationality | the validity |
| easily concluded | can be concluded |
**检查:** 提交前搜索 `isolately, classifiy, enviroment, rationality, easily`
### 3.3 段落结构
**问题:** 段落无主题句,直接开始写数据
**正确:**
```
[主题句] From Figure 1, we observe that the biomass decreases...
[证据] Specifically, the total biomass drops from 206.5 kg to...
[解释] This is because...
```
---
## 4 图表规范
### 4.1 图注(caption)规范
**模板:**
```
Figure N: [What] vs [What] under [condition]
```
**示例:**
- Figure 1: Total biomass vs time under drought conditions
- Figure 2: Sensitivity of output to parameter α
- Figure 3: Comparison of Model I vs Model II performance
### 4.2 正文引用规范
**问题:** "The results are shown in Figure 5."(无解读)
**正确:**
```
From Figure 5, we draw the following conclusions:
• The biomass of both plants are influenced by periodic weather change...
• During recovery, the H.nepalensis recovers relatively quickly...
```
### 4.3 图表检查清单
- [ ] 每张图有 caption(Figure N: ...)
- [ ] 每张图在正文中被引用("From Figure N...")
- [ ] 每张图有解读段落(结论句)
- [ ] 坐标轴有标签 + 单位
- [ ] 多曲线有图例
- [ ] 分辨率 300 dpi
---
## 5 LaTeX 排版规范
### 5.1 关键配置
```latex
\documentclass[12pt]{article}
\usepackage[margin=1in,a4paper]{geometry}
\usepackage{fancyhdr}
\usepackage{booktabs}
\usepackage{graphicx}
\usepackage{amsmath,amssymb}
\usepackage{enumitem}
\usepackage[hidelinks]{hyperref}
% 页眉:Team # XXXXXXX Page X of 25
\fancyhead[C]{Team \# \textbf{XXXXXXX} \quad Page \thepage \ of \pageref{LastPage}}
```
**参考:** [latex-template.tex](../templates/latex-template.tex)
### 5.2 排版检查清单
- [ ] 编译无 warning
- [ ] 页数 ≤ 25(含附录和参考文献)
- [ ] 页眉显示 `Team # XXXXXXX Page X of Y`
- [ ] PDF < 17 MB
- [ ] 文件名 = 队号.pdf
- [ ] 无个人信息(姓名/学校/邮箱)
- [ ] 参考文献编号连续([1][2][3]...)
- [ ] 所有图有 caption 和 label
- [ ] 所有图在正文中被引用过
- [ ] 摘要数字 ≥ 50
- [ ] 摘要显式标记每个 task
- [ ] Strengths 和 Weaknesses 各 3 条
- [ ] 灵敏度独立章节
---
## 6 Memo/Letter 写作规范(F 题)
### 6.1 格式
```
To: [recipient]
From: [your name/team name]
Date: [date]
Re: [subject]
```
### 6.2 内容结构
```
1. 问候(1 句)
2. 目的(1 句)
3. 模型简介(2-3 句)
4. 策略/建议(3-4 句)
5. 风险提示(2-3 句)
6. 名言/引用(可选)
7. 结尾(1 句)
8. 签名(姓名 + 日期)
```
**O 奖示例(2022 C):** 引用 Warren Buffett 名言增加人文关怀
---
## 7 写作手工作流(4 天赛制)
### Day 1(框架 + 摘要)
| 时间 | 任务 |
|---|---|
| 0-2h | 读取题目 + 建模手框架 |
| 2-4h | 摘要框架(模板填充) |
| 4-8h | Introduction 起草 |
| 8-12h | 假设 + 符号 |
| 12-24h | 与编程手协作(图表编号) |
### Day 2(正文 + 图表)
| 时间 | 任务 |
|---|---|
| 0-8h | Task 1-2 模型章节 |
| 8-16h | Task 1-2 结果章节 |
| 16-24h | 图表引用 + 解读 |
### Day 3(灵敏度 + 优劣势)
| 时间 | 任务 |
|---|---|
| 0-8h | Task 3 模型 + 结果 |
| 8-16h | 灵敏度分析章节 |
| 16-24h | 优劣势 + 结论 |
### Day 4(排版 + 检查)
| 时间 | 任务 |
|---|---|
| 0-8h | LaTeX 排版 |
| 8-12h | 一致性检查(check-paper.mjs) |
| 12-16h | 最终校对 |
| 16-24h | 提交准备 |
---
## 8 写作手检查清单
### 8.1 摘要
- [ ] 词数 400-500
- [ ] 数字 50-100+
- [ ] 模型命名("we develop")
- [ ] 每个 task 显式标记
- [ ] 灵敏度结论
- [ ] 优劣势 3 vs 3
- [ ] 关键词
### 8.2 正文
- [ ] 段落有主题句
- [ ] 图表有解读
- [ ] 方程有编号
- [ ] 参数有来源
- [ ] 灵敏度独立章节
- [ ] 优劣势各 3 条
### 8.3 排版
- [ ] 编译无 warning
- [ ] 页数 ≤ 25
- [ ] 页眉正确
- [ ] PDF < 17 MB
- [ ] 无个人信息
- [ ] 参考文献连续
---
## 9 写作手与建模手/编程手的协作
### 9.1 与建模手
**建模手输出:**
- 模型描述("we develop...")
- 参数来源说明
- 结果解读要点
**写作手撰写:**
- 模型章节
- 结果章节
- 灵敏度章节
### 9.2 与编程手
**编程手输出:**
- 代码(附录)
- 图表(PDF/PNG)
- 结果数据(CSV/表格)
**写作手撰写:**
- 图表引用("From Figure N...")
- 代码说明(附录)
- 结果解读
---
## 10 常见写作错误
| 错误 | 后果 | 正确做法 |
|---|---|---|
| 摘要无数字 | 初筛淘汰 | 数字 50-100+ |
| 摘要复述题目 | 初筛淘汰 | 第一句引入背景,第二句给模型 |
| 假设无理由 | 筛选扣分 | 每条带 justification |
| 图表无解读 | 筛选扣分 | 每张图有结论句 |
| 灵敏度敷衍 | 降级风险 | 2+ 方法 + 结论句 |
| 中式英语 | 终评扣分 | 搜索常见错误词 |
**参考:** [common-errors.md](../references/common-errors.md)
---
## 快速索引
| 需求 | 章节 |
|---|---|
| 摘要优先 | §1 |
| 固定结构 | §2 |
| 学术英文 | §3 |
| 图表规范 | §4 |
| LaTeX 排版 | §5 |
| Memo 写作 | §6 |
| 工作流 | §7 |
| 检查清单 | §8 |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |