feat: 钉钉群聊飞书文档收集与知识图谱系统初始提交

- 通过悟空(dws CLI)拉取dc战略问题研究院+创新组两个群的消息(377条) - 提取190个飞书链接、18个文件附件 - 下载HTML/MD/XLSX等报告文件到output/downloaded-files/ - 构建知识图谱(JSON+HTML可视化) - 生成Obsidian知识库(28个页面,7大主题) - 生成花园世界全量汇总报告 - 所有脚本路径改为相对路径,便于迁移
2026-06-02 20:24:25 +08:00
commit fe5505343e
77 changed files with 20155 additions and 0 deletions
@@ -0,0 +1,740 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+<meta charset="UTF-8">
+<meta name="viewport" content="width=device-width, initial-scale=1.0">
+<title>手游市场战略研究模型</title>
+<style>
+  *, *::before, *::after { box-sizing: border-box; margin: 0; padding: 0; }
+
+  body {
+    font-family: -apple-system, "PingFang SC", "Microsoft YaHei", sans-serif;
+    font-size: 15px;
+    line-height: 1.7;
+    color: #1a1a2e;
+    background: #f7f8fc;
+  }
+
+  /* ── Layout ── */
+  .page { max-width: 960px; margin: 0 auto; padding: 48px 32px 80px; }
+
+  /* ── Header ── */
+  .report-header {
+    background: linear-gradient(135deg, #1a1a2e 0%, #16213e 60%, #0f3460 100%);
+    color: #fff;
+    padding: 52px 48px 44px;
+    border-radius: 16px;
+    margin-bottom: 40px;
+    position: relative;
+    overflow: hidden;
+  }
+  .report-header::before {
+    content: '';
+    position: absolute; inset: 0;
+    background: radial-gradient(ellipse at 80% 50%, rgba(255,255,255,.06) 0%, transparent 60%);
+  }
+  .report-header h1 {
+    font-size: 28px; font-weight: 700; letter-spacing: .5px;
+    margin-bottom: 10px; position: relative;
+  }
+  .report-header .subtitle {
+    font-size: 14px; opacity: .65; position: relative;
+  }
+  .version-badge {
+    display: inline-block;
+    background: rgba(255,255,255,.15);
+    border: 1px solid rgba(255,255,255,.25);
+    color: #fff;
+    font-size: 12px; font-weight: 600;
+    padding: 3px 10px; border-radius: 20px;
+    margin-left: 12px; vertical-align: middle;
+    letter-spacing: .5px;
+  }
+
+  /* ── Section ── */
+  .section { margin-bottom: 44px; }
+  .section-title {
+    font-size: 19px; font-weight: 700;
+    color: #0f3460;
+    padding-bottom: 10px;
+    border-bottom: 2px solid #e2e8f4;
+    margin-bottom: 20px;
+    display: flex; align-items: center; gap: 10px;
+  }
+  .section-title .icon {
+    width: 28px; height: 28px;
+    background: #0f3460;
+    color: #fff;
+    border-radius: 7px;
+    display: flex; align-items: center; justify-content: center;
+    font-size: 13px; flex-shrink: 0;
+  }
+
+  h3 {
+    font-size: 15px; font-weight: 600; color: #16213e;
+    margin: 20px 0 10px;
+  }
+
+  /* ── Card grid ── */
+  .card-grid { display: grid; grid-template-columns: 1fr 1fr; gap: 16px; }
+  .card {
+    background: #fff;
+    border: 1px solid #e2e8f4;
+    border-radius: 12px;
+    padding: 20px 22px;
+  }
+  .card.full { grid-column: 1 / -1; }
+  .card-label {
+    font-size: 11px; font-weight: 700; letter-spacing: 1px;
+    text-transform: uppercase; color: #6b7a99;
+    margin-bottom: 8px;
+  }
+  .card-value {
+    font-size: 22px; font-weight: 700; color: #0f3460;
+  }
+  .card-desc { font-size: 13px; color: #5a6484; margin-top: 4px; }
+
+  /* ── Tables ── */
+  table {
+    width: 100%;
+    border-collapse: collapse;
+    font-size: 14px;
+    background: #fff;
+    border-radius: 10px;
+    overflow: hidden;
+    box-shadow: 0 1px 4px rgba(0,0,0,.06);
+  }
+  th {
+    background: #f0f3fb;
+    color: #4a5568;
+    font-weight: 600;
+    font-size: 12px;
+    letter-spacing: .4px;
+    text-transform: uppercase;
+    padding: 10px 14px;
+    text-align: left;
+    border-bottom: 1px solid #e2e8f4;
+  }
+  td {
+    padding: 10px 14px;
+    border-bottom: 1px solid #f0f3fb;
+    vertical-align: middle;
+  }
+  tr:last-child td { border-bottom: none; }
+  tr:hover td { background: #fafbff; }
+  .num { text-align: right; font-variant-numeric: tabular-nums; }
+
+  /* ── Color tags ── */
+  .tag {
+    display: inline-block;
+    font-size: 11px; font-weight: 600;
+    padding: 2px 8px; border-radius: 12px;
+    white-space: nowrap;
+  }
+  .tag-blue  { background: #e8f0fe; color: #1a56db; }
+  .tag-green { background: #e3f9e5; color: #1a7431; }
+  .tag-red   { background: #fde8e8; color: #c53030; }
+  .tag-gray  { background: #f0f3fb; color: #4a5568; }
+  .tag-orange{ background: #fff3e0; color: #b45309; }
+
+  /* ── Pos/neg values ── */
+  .pos { color: #1a7431; font-weight: 600; }
+  .neg { color: #c53030; font-weight: 600; }
+  .neu { color: #4a5568; }
+
+  /* ── Highlight box ── */
+  .highlight {
+    background: #fff8e1;
+    border-left: 4px solid #f59e0b;
+    border-radius: 0 10px 10px 0;
+    padding: 14px 18px;
+    margin: 16px 0;
+    font-size: 14px;
+  }
+  .highlight strong { color: #92400e; }
+
+  .note {
+    background: #f0f7ff;
+    border-left: 4px solid #3b82f6;
+    border-radius: 0 10px 10px 0;
+    padding: 12px 16px;
+    margin: 14px 0;
+    font-size: 13px;
+    color: #1e40af;
+  }
+
+  /* ── Mechanism cards ── */
+  .mech-grid { display: grid; grid-template-columns: 1fr 1fr; gap: 14px; }
+  .mech-card {
+    background: #fff;
+    border: 1px solid #e2e8f4;
+    border-radius: 12px;
+    padding: 18px 20px;
+    border-top: 3px solid #0f3460;
+  }
+  .mech-card h4 {
+    font-size: 14px; font-weight: 700; color: #0f3460;
+    margin-bottom: 8px;
+  }
+  .mech-card p { font-size: 13px; color: #4a5568; line-height: 1.6; }
+  .mech-card .formula {
+    font-family: "SF Mono", "Fira Code", monospace;
+    font-size: 12px;
+    background: #f0f3fb;
+    padding: 6px 10px;
+    border-radius: 6px;
+    margin-top: 8px;
+    color: #2d3748;
+    word-break: break-all;
+  }
+
+  /* ── Strategy table ── */
+  .strat-name {
+    font-weight: 600; font-size: 14px;
+  }
+  .strat-dot {
+    width: 10px; height: 10px;
+    border-radius: 50%;
+    display: inline-block;
+    margin-right: 7px;
+    vertical-align: middle;
+    flex-shrink: 0;
+  }
+
+  /* ── Result table ── */
+  .winner-row td { background: #f0faf3 !important; }
+  .winner-row .strat-name { color: #1a7431; }
+  .utility-cell {
+    font-size: 15px; font-weight: 700;
+  }
+
+  /* ── Utility bar ── */
+  .bar-wrap { display: flex; align-items: center; gap: 8px; }
+  .bar {
+    height: 8px; border-radius: 4px;
+    background: #3b82f6;
+    min-width: 4px;
+    transition: width .3s;
+  }
+  .bar-neg { background: #ef4444; }
+  .bar-val { font-size: 13px; font-weight: 600; font-variant-numeric: tabular-nums; white-space: nowrap; }
+
+  /* ── Conclusion box ── */
+  .conclusion {
+    background: linear-gradient(135deg, #0f3460, #16213e);
+    color: #fff;
+    border-radius: 14px;
+    padding: 32px 36px;
+    margin-top: 32px;
+  }
+  .conclusion h3 {
+    color: #fff; font-size: 18px; margin: 0 0 18px;
+    border-bottom: 1px solid rgba(255,255,255,.2);
+    padding-bottom: 12px;
+  }
+  .conclusion-grid { display: grid; grid-template-columns: 1fr 1fr; gap: 16px; }
+  .con-item {
+    background: rgba(255,255,255,.08);
+    border: 1px solid rgba(255,255,255,.15);
+    border-radius: 10px;
+    padding: 16px 18px;
+  }
+  .con-item .goal { font-size: 11px; text-transform: uppercase; letter-spacing: .8px; opacity: .6; margin-bottom: 5px; }
+  .con-item .answer { font-size: 15px; font-weight: 700; color: #7dd3fc; }
+  .con-item .reason { font-size: 12px; opacity: .65; margin-top: 4px; }
+
+  /* ── Warning ── */
+  .caveat {
+    background: rgba(255,255,255,.05);
+    border: 1px solid rgba(255,165,0,.3);
+    border-radius: 8px;
+    padding: 12px 16px;
+    margin-top: 16px;
+    font-size: 13px;
+    opacity: .8;
+  }
+  .caveat strong { color: #fcd34d; }
+
+  /* ── Footer ── */
+  .footer {
+    margin-top: 56px;
+    text-align: center;
+    font-size: 12px;
+    color: #9aa3bf;
+  }
+
+  /* ── Responsive ── */
+  @media (max-width: 680px) {
+    .card-grid, .mech-grid, .conclusion-grid { grid-template-columns: 1fr; }
+    .report-header { padding: 32px 24px; }
+    .page { padding: 24px 16px 60px; }
+  }
+</style>
+</head>
+<body>
+<div class="page">
+
+<!-- ══ Header ══════════════════════════════════════════════════════════════ -->
+<div class="report-header">
+  <h1>手游市场战略研究模型 <span class="version-badge">v32</span></h1>
+  <p class="subtitle">基于 Agent-Based Simulation 的开发商与发行商策略演化分析 &nbsp;·&nbsp; 仿真周期 30–50 年 &nbsp;·&nbsp; 10,000 家开发商 × 2,000 家发行商</p>
+</div>
+
+<!-- ══ 1. 研究问题 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">?</div>研究问题</div>
+  <p style="color:#4a5568;line-height:1.8;">
+    在一个存在<strong>学习效应、市场周期、产品类型分化、规模上限、动态分成</strong>的手游市场中，
+    开发商应该选择什么产品策略才能在长达 30 年的市场竞争中存活，并为进入者创造最大的期望收益？
+  </p>
+  <div class="note" style="margin-top:16px;">
+    <strong>核心指标：期望年化收益</strong> = 存活率 × 幸存者年化收益。
+    这是一个随机进入市场的参与者所能期望的真实年化回报——同时惩罚高淘汰率策略和低收益策略。
+  </div>
+</div>
+
+<!-- ══ 2. 市场结构 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">🏢</div>市场结构</div>
+
+  <div class="card-grid">
+    <div class="card">
+      <div class="card-label">开发商</div>
+      <div class="card-value">10,000 家</div>
+      <div class="card-desc">能力 AP ~ N(50, 20)，截断至 [0, 100]<br>
+        稳定性 SD ~ N(20, 10)，截断至 [1, 50]<br>
+        初始资金 ~ N(5C, 2.5C)，最低 0.1C<br>
+        每年最多开发 1 款产品（固定，不允许并行）</div>
+    </div>
+    <div class="card">
+      <div class="card-label">发行商</div>
+      <div class="card-value">2,000 家</div>
+      <div class="card-desc">能力加成 pub_ability ~ N(3, 1)，截断至 [1, 5]<br>
+        初始资金 ~ N(5C, 2.5C)，最低 0.1C<br>
+        并行接单数由资金隐式决定（无硬上限）</div>
+    </div>
+    <div class="card">
+      <div class="card-label">市场进入与退出</div>
+      <div class="card-value full">动态补充</div>
+      <div class="card-desc">每年破产清算后补充缺口 × 50%，使总数向初始规模均值回归。
+        破产后永久退出，进行中的项目继续完成但收益归零。</div>
+    </div>
+    <div class="card">
+      <div class="card-label">破产门槛</div>
+      <div class="card-value">资金不足</div>
+      <div class="card-desc">
+        开发商：无活跃项目 且 资金 &lt; 下一款产品的开发商承担成本<br>
+        发行商：无活跃项目 且 资金 &lt; 0.5C</div>
+    </div>
+  </div>
+</div>
+
+<!-- ══ 3. 财务模型 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">💰</div>财务模型</div>
+
+  <h3>项目周期资金流</h3>
+  <table>
+    <thead><tr><th>时点</th><th>开发商</th><th>发行商</th></tr></thead>
+    <tbody>
+      <tr>
+        <td><strong>启动时</strong></td>
+        <td>扣除 cost × <strong>70%</strong></td>
+        <td>扣除 cost × <strong>30%</strong></td>
+      </tr>
+      <tr>
+        <td><strong>成功完成</strong>（质量 P ≥ Q=80）</td>
+        <td>收回 cost×70% + <em>dev_split</em> × profit</td>
+        <td>收回 cost×30% + <em>(1−dev_split)</em> × profit</td>
+      </tr>
+      <tr>
+        <td><strong>失败完成</strong>（P &lt; Q）</td>
+        <td>损失 cost×70%（不退还）</td>
+        <td>损失 cost×30%（不退还）</td>
+      </tr>
+    </tbody>
+  </table>
+
+  <div style="margin-top:14px;">
+    <span style="font-size:13px;color:#4a5568;">
+      利润公式：<code style="background:#f0f3fb;padding:2px 6px;border-radius:4px;">profit = (P − 80) × 0.2 × rev_mult</code>，
+      其中 <code style="background:#f0f3fb;padding:2px 6px;border-radius:4px;">P = clip(N(eff_AP, eff_SD) + pub_ability, 0, 100)</code>
+    </span>
+  </div>
+
+  <h3>产品类型</h3>
+  <table>
+    <thead><tr><th>类型</th><th>成本</th><th>AP 倍数</th><th>SD 倍数</th><th>特点</th></tr></thead>
+    <tbody>
+      <tr>
+        <td><span class="tag tag-gray">普通</span></td>
+        <td class="num">× 1 C</td><td class="num">× 1.0</td><td class="num">× 1.0</td>
+        <td>基准类型</td>
+      </tr>
+      <tr>
+        <td><span class="tag tag-blue">高投入</span></td>
+        <td class="num">× 2 C</td><td class="num">× 1.2</td><td class="num">× 0.7</td>
+        <td>高成本、高质量均值、低方差</td>
+      </tr>
+      <tr>
+        <td><span class="tag tag-orange">低投入</span></td>
+        <td class="num">× 0.5 C</td><td class="num">× 0.8</td><td class="num">× 1.5</td>
+        <td>低成本、低质量均值、高方差（高风险）</td>
+      </tr>
+    </tbody>
+  </table>
+</div>
+
+<!-- ══ 4. 核心机制 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">⚙</div>四大核心机制</div>
+
+  <div class="mech-grid">
+    <div class="mech-card">
+      <h4>① 学习效应 + 遗忘</h4>
+      <p>同类型连续开发每年 AP 乘数 +1%；切换类型则乘数重置为 1.0（转型惩罚）。
+        每 5 年所有存活开发商的学习增量减半，防止无限复利，均衡稳态约 1.05–1.06×。</p>
+      <div class="formula">multiplier ← 1.0 + (multiplier − 1.0) × 0.5　（每5年）</div>
+    </div>
+    <div class="mech-card">
+      <h4>② 市场红利（双层）</h4>
+      <p><strong>集中度红利</strong>：占比越低的类型 bonus 越高（上限 2.0×，下限 0.5×），奖励逆势进入。<br>
+        <strong>市场趋势冲击</strong>：每 5 年随机切换趋势类型，趋势类型额外 ×1.5。两信号叠乘，方向可能相反。</p>
+      <div class="formula">rev_mult = bonus[type] × (1.5 if trend else 1.0)</div>
+    </div>
+    <div class="mech-card">
+      <h4>③ 市场规模上限</h4>
+      <p>前 15 年为自由增长期，锁定历史最高年总利润为上限。
+        第 16 年起，若当年利润总和超出上限，等比压缩所有成功项目的利润部分（成本回款不受影响）。</p>
+      <div class="formula">scale = market_cap / total_revenue　（超出时）</div>
+    </div>
+    <div class="mech-card">
+      <h4>④ 动态分成</h4>
+      <p>每年根据当前活跃项目的主流产品类型调整利润分成：高投入产品主流时开发商获6:4优势，低投入主流时发行商获6:4优势，均衡时5:5。</p>
+      <div class="formula">
+        高投入主流 → dev : pub = 6 : 4<br>
+        普通主流　 → dev : pub = 5 : 5<br>
+        低投入主流 → dev : pub = 4 : 6
+      </div>
+    </div>
+  </div>
+
+  <div class="highlight" style="margin-top:20px;">
+    <strong>机制互动关键点：</strong>
+    专注高投入策略触发 6:4 分成 → 开发商利润增多 → 市场总利润更快超出上限 → 规模上限更早、更猛地压缩后期收益。
+    两个机制在高投入策略下形成"给了又收回"的抵消效应。
+  </div>
+</div>
+
+<!-- ══ 5. 策略选项 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">🎯</div>开发商策略选项</div>
+
+  <table>
+    <thead>
+      <tr>
+        <th>策略</th><th>决策频率</th><th>逻辑</th><th>学习乘数</th>
+      </tr>
+    </thead>
+    <tbody>
+      <tr>
+        <td><span class="strat-dot" style="background:#4e79a7;"></span><strong>专注普通</strong></td>
+        <td>不切换</td>
+        <td>始终开发普通产品</td>
+        <td><span class="tag tag-green">持续积累</span></td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#2a9d8f;"></span><strong>专注高投入</strong></td>
+        <td>不切换</td>
+        <td>始终开发高投入产品，享受6:4分成</td>
+        <td><span class="tag tag-green">持续积累</span></td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#f4a261;"></span><strong>专注低投入</strong></td>
+        <td>不切换</td>
+        <td>低成本高容错，但质量均值低</td>
+        <td><span class="tag tag-green">持续积累</span></td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#e63946;"></span><strong>追逐红利</strong></td>
+        <td>每年</td>
+        <td>切换至当年 argmax(集中度红利 × 趋势)，追逐最高 rev_mult</td>
+        <td><span class="tag tag-red">每年重置</span></td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#ff9800;"></span><strong>跟随周期</strong></td>
+        <td>每5年</td>
+        <td>仅在市场趋势切换年（每5年）跟随新趋势类型，中间4年稳定积累</td>
+        <td><span class="tag tag-blue">5年积累后重置</span></td>
+      </tr>
+    </tbody>
+  </table>
+</div>
+
+<!-- ══ 6. 仿真结论 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">📊</div>仿真结论</div>
+  <p style="font-size:13px;color:#6b7a99;margin-bottom:16px;">
+    分析场景：开发商 2,000 家 × 发行商 400 家，仿真 30 年，3 个随机种子均值。
+  </p>
+
+  <table>
+    <thead>
+      <tr>
+        <th>策略</th>
+        <th class="num">dev 存活</th>
+        <th class="num">dev 年化</th>
+        <th class="num" style="background:#e8f5e9;">dev 期望年化↑</th>
+        <th class="num">pub 存活</th>
+        <th class="num">pub 年化</th>
+        <th class="num" style="background:#e8f5e9;">pub 期望年化↑</th>
+        <th>分成主态</th>
+      </tr>
+    </thead>
+    <tbody>
+      <!-- 跟随周期 winner -->
+      <tr class="winner-row">
+        <td><span class="strat-dot" style="background:#ff9800;"></span><span class="strat-name">跟随周期</span></td>
+        <td class="num">18.0%</td>
+        <td class="num pos">+4.11%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:70px;"></div>
+            <span class="bar-val pos">+0.701%</span>
+          </div>
+        </td>
+        <td class="num">66.6%</td>
+        <td class="num pos">+3.61%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:100px;background:#16a34a;"></div>
+            <span class="bar-val pos">+2.504%</span>
+          </div>
+        </td>
+        <td><span class="tag tag-blue">混合均衡</span></td>
+      </tr>
+      <!-- 专注高投入 -->
+      <tr>
+        <td><span class="strat-dot" style="background:#2a9d8f;"></span><span class="strat-name">专注高投入</span></td>
+        <td class="num">24.9%</td>
+        <td class="num pos">+2.93%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:66px;"></div>
+            <span class="bar-val pos">+0.658%</span>
+          </div>
+        </td>
+        <td class="num">81.1%</td>
+        <td class="num neg">−6.68%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar bar-neg" style="width:90px;"></div>
+            <span class="bar-val neg">−5.420%</span>
+          </div>
+        </td>
+        <td><span class="tag tag-blue">6:4（96%）</span></td>
+      </tr>
+      <!-- 追逐红利 -->
+      <tr>
+        <td><span class="strat-dot" style="background:#e63946;"></span><span class="strat-name">追逐红利</span></td>
+        <td class="num">14.1%</td>
+        <td class="num pos">+4.22%<sup>*</sup></td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:59px;"></div>
+            <span class="bar-val pos">+0.593%</span>
+          </div>
+        </td>
+        <td class="num">58.3%</td>
+        <td class="num neg">−1.85%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar bar-neg" style="width:18px;"></div>
+            <span class="bar-val neg">−1.072%</span>
+          </div>
+        </td>
+        <td><span class="tag tag-gray">混合波动</span></td>
+      </tr>
+      <!-- 专注普通 -->
+      <tr>
+        <td><span class="strat-dot" style="background:#4e79a7;"></span><span class="strat-name">专注普通</span></td>
+        <td class="num">15.5%</td>
+        <td class="num pos">+3.43%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:53px;"></div>
+            <span class="bar-val pos">+0.537%</span>
+          </div>
+        </td>
+        <td class="num">65.3%</td>
+        <td class="num neg">−3.15%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar bar-neg" style="width:24px;"></div>
+            <span class="bar-val neg">−1.482%</span>
+          </div>
+        </td>
+        <td><span class="tag tag-gray">5:5（100%）</span></td>
+      </tr>
+      <!-- 专注低投入 -->
+      <tr>
+        <td><span class="strat-dot" style="background:#f4a261;"></span><span class="strat-name">专注低投入</span></td>
+        <td class="num">28.9%</td>
+        <td class="num neu">−0.02%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar bar-neg" style="width:4px;"></div>
+            <span class="bar-val neu">−0.006%</span>
+          </div>
+        </td>
+        <td class="num">93.4%</td>
+        <td class="num pos">+2.08%</td>
+        <td class="num">
+          <div class="bar-wrap">
+            <div class="bar" style="width:78px;background:#2563eb;"></div>
+            <span class="bar-val pos">+1.950%</span>
+          </div>
+        </td>
+        <td><span class="tag tag-orange">4:6（96%）</span></td>
+      </tr>
+    </tbody>
+  </table>
+
+  <p style="font-size:12px;color:#9aa3bf;margin-top:8px;">
+    * 追逐红利的幸存者年化（+4.22%）受聚集型破产导致的幸存者偏差影响虚高；期望年化（+0.593%）为真实效用。
+  </p>
+
+  <h3 style="margin-top:28px;">各策略50年演化轨迹（主场景，seed=42）</h3>
+  <table>
+    <thead>
+      <tr>
+        <th>策略</th>
+        <th class="num">年50 开发商AP</th>
+        <th class="num">年50 学习乘数</th>
+        <th class="num">年50 市场压缩</th>
+        <th class="num">幸存者资本</th>
+        <th>市场上限首次触发</th>
+      </tr>
+    </thead>
+    <tbody>
+      <tr class="winner-row">
+        <td><span class="strat-dot" style="background:#ff9800;"></span><strong>跟随周期</strong></td>
+        <td class="num">72.1</td>
+        <td class="num">1.001</td>
+        <td class="num pos">1.000 × <small>（未压缩）</small></td>
+        <td class="num"><strong>35.7 C</strong></td>
+        <td>第 21 年</td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#2a9d8f;"></span>专注高投入</td>
+        <td class="num"><strong>85.0</strong></td>
+        <td class="num">1.054</td>
+        <td class="num neg">0.333 × <small>（打三折）</small></td>
+        <td class="num">26.8 C</td>
+        <td>第 20 年</td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#e63946;"></span>追逐红利</td>
+        <td class="num">70.3</td>
+        <td class="num">1.000</td>
+        <td class="num neg">0.665 ×</td>
+        <td class="num">23.4 C</td>
+        <td>第 16 年</td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#4e79a7;"></span>专注普通</td>
+        <td class="num">69.6</td>
+        <td class="num">1.045</td>
+        <td class="num">0.712 ×</td>
+        <td class="num">24.6 C</td>
+        <td>第 25 年</td>
+      </tr>
+      <tr>
+        <td><span class="strat-dot" style="background:#f4a261;"></span>专注低投入</td>
+        <td class="num">55.2</td>
+        <td class="num">1.048</td>
+        <td class="num pos">0.945 ×</td>
+        <td class="num">7.4 C</td>
+        <td>第 40 年</td>
+      </tr>
+    </tbody>
+  </table>
+</div>
+
+<!-- ══ 7. 核心发现 ══════════════════════════════════════════════════════════ -->
+<div class="section">
+  <div class="section-title"><div class="icon">💡</div>核心发现</div>
+
+  <div class="mech-grid" style="margin-bottom:16px;">
+    <div class="mech-card" style="border-top-color:#ff9800;">
+      <h4>跟随周期：唯一双边共赢策略</h4>
+      <p>dev 期望 <strong>+0.701%</strong>，pub 期望 <strong>+2.504%</strong>，两侧均正。
+        50 年末幸存者资本 35.7C 为全场最高。
+        每5年切换使分成在三档间轮转，平均趋近5:5，规模上限全程近乎不触发。</p>
+    </div>
+    <div class="mech-card" style="border-top-color:#2a9d8f;">
+      <h4>专注高投入：开发商个体最优，发行商深亏</h4>
+      <p>6:4 分成将 dev 存活率推至 <strong>24.9%</strong>，但同时令市场利润更快超出上限——
+        年 50 时收入仅剩 <strong>0.33×</strong>。发行商期望 <strong>−5.42%</strong>，长期无法维持。</p>
+    </div>
+    <div class="mech-card" style="border-top-color:#e63946;">
+      <h4>追逐红利：幸存者偏差陷阱</h4>
+      <p>幸存者年化（+4.22%）看似最高，实为假象：羊群效应导致聚集型破产，
+        批量死亡在均值资本中制造统计跳升。期望年化仅 <strong>+0.593%</strong>，排第三。</p>
+    </div>
+    <div class="mech-card" style="border-top-color:#f4a261;">
+      <h4>专注低投入：高存活，但接近零效用</h4>
+      <p>dev 存活率最高（28.9%），发行商存活率高达 93.4%。
+        然而 dev 期望仅 <strong>−0.006%</strong>，为勉强收支平衡的维持型路线。
+        市场规模压制最轻，属保守仓位而非成长策略。</p>
+    </div>
+  </div>
+
+  <div class="highlight">
+    <strong>机制权衡：</strong>
+    专注高投入的"6:4 分成优势"与"规模上限反噬"相互抵消——高投入利润越多，规模上限触发越早越猛，
+    反而令50年末幸存者资本（26.8C）低于跟随周期（35.7C）。
+    没有一种专一策略能同时绕开这两重约束。
+  </div>
+</div>
+
+<!-- ══ 8. 策略建议 ══════════════════════════════════════════════════════════ -->
+<div class="conclusion">
+  <h3>策略建议（以期望年化为主要指标）</h3>
+  <div class="conclusion-grid">
+    <div class="con-item">
+      <div class="goal">双边市场健康（dev + pub）</div>
+      <div class="answer">跟随周期</div>
+      <div class="reason">唯一双边期望均正，50年末资本最高</div>
+    </div>
+    <div class="con-item">
+      <div class="goal">最大化开发商 dev 期望收益</div>
+      <div class="answer">跟随周期 +0.701%</div>
+      <div class="reason">其次专注高投入 +0.658%</div>
+    </div>
+    <div class="con-item">
+      <div class="goal">最大化发行商 pub 期望收益</div>
+      <div class="answer">跟随周期 +2.504%</div>
+      <div class="reason">其次专注低投入 +1.950%</div>
+    </div>
+    <div class="con-item">
+      <div class="goal">最大化开发商存活率（不计收益）</div>
+      <div class="answer">专注低投入 28.9%</div>
+      <div class="reason">低破产门槛，但期望收益近零</div>
+    </div>
+  </div>
+  <div class="caveat">
+    <strong>注意：</strong>
+    追逐红利的幸存者年化（+4.22%）并非真实效用——大多数参与者在赚到利润前就已破产（仅14.1%存活）。
+    期望年化（+0.593%）才是对随机进入者有意义的指标，追逐红利在此指标下排第三。
+  </div>
+</div>
+
+<!-- ══ Footer ══════════════════════════════════════════════════════════════ -->
+<div class="footer">
+  手游市场战略研究模型 v32 &nbsp;·&nbsp; Agent-Based Simulation &nbsp;·&nbsp; simulation.py
+</div>
+
+</div>
+</body>
+</html>