定量方法（Quantitative Methods）— 概率论模块 · 第四课
一、本课定位
L117 学习了条件概率和贝叶斯公式——如何在有新信息时更新概率判断。但之前我们一直"假设概率已知"。现在退一步问：概率本身是怎么算出来的？
当所有基本结果等可能时：
$$P(A) = \frac{A\text{包含的结果数}}{所有可能的结果总数}$$
分子分母都依赖计数。本课就学怎么"数"——排列组合。
| 项目 | 说明 |
|---|---|
| 模块 | 2.4 概率论 |
| 前置知识 | L115 概率基础、L116 期望与方差 |
| 后续衔接 | L119 概率分布入门 |
| 难度 | ★★★☆☆ |
| 考试权重 | 中（概念题 + 简单计算，1-2 题） |
| 阅读时间 | 约 12 分钟 |
二、核心概念
1. 乘法原理（Multiplication Principle）
直觉引入：
你去买午餐。有 3 种主食（饭、面、馒头）和 4 种配菜（鱼、鸡、牛、素）。主食和配菜各选一样，有多少种组合？
你不需要掰手指头。每一种主食可以搭配 4 种配菜，3 × 4 = 12 种。
这就是乘法原理的本质：如果一件事分 k 步完成，第 i 步有 nᵢ 种选择，总方案数 = n₁ × n₂ × ... × nₖ。
正式表述：
如果一个任务可以分成 k 个阶段，第 i 个阶段有 nᵢ 种完成方式，且各阶段之间互不影响，则完成整个任务的方式总数为：
$$n_1 \times n_2 \times \dots \times n_k$$
案例 1：投资组合构建
某分析师需要从以下资产中选一只股票、一只债券和一种商品构建迷你组合：
| 资产类别 | 可选数量 |
|---|---|
| 股票 | 50 只 |
| 债券 | 30 只 |
| 商品 | 10 种 |
总组合数：
$$50 \times 30 \times 10 = 15{,}000$$
📊 每新增一个资产类别，组合空间呈乘法级扩大——这就是"维度诅咒"在组合优化中的体现。
案例 2：密码复杂度
一个 4 位数字密码，每位可选 0-9：
$$10 \times 10 \times 10 \times 10 = 10^4 = 10{,}000$$
一个 4 位密码，每位可选 26 个大写字母 + 26 个小写字母 + 10 个数字 = 62 种：
$$62^4 = 14{,}776{,}336$$
📊 从 1 万到 1477 万——只是把字符集从 10 种扩到 62 种。这就是为什么网站要求你密码"包含大小写和数字"。
2. 阶乘（Factorial）
定义：
$$n! = n \times (n-1) \times (n-2) \times \dots \times 2 \times 1$$
$$0! = 1 \quad \text{（约定）}$$
案例： 5 只股票按看好程度排名，有多少种排名方式？
$$5! = 5 \times 4 \times 3 \times 2 \times 1 = 120$$
3. 排列（Permutation）—— 顺序重要
直觉问题：
从 8 个候选人中选出 3 人分别担任 CEO、CFO、COO（三个职位不同）。有多少种选法？
关键判断：职位不同 → 顺序重要 → 排列问题。
公式（选排列）：
$$P(n, r) = {}_nP_r = \frac{n!}{(n-r)!}$$
推导逻辑： - 第 1 个职位：8 人选 - 第 2 个职位：剩 7 人选 - 第 3 个职位：剩 6 人选 - 总数：8 × 7 × 6 = 336
用公式：$P(8, 3) = \frac{8!}{5!} = 8 \times 7 \times 6 = 336$ ✅
案例 3：基金经理排名
某基金评选年度 Top 10 基金经理，从 100 位候选人中选出。第 1、第 2、第 3 名的排列数：
$$P(100, 3) = 100 \times 99 \times 98 = 970{,}200$$
📊 近 100 万种可能！这就是为什么每年基金经理排名"换人"很快——排列的组合空间太大了。
特殊情况：全排列
当 r = n 时：
$$P(n, n) = n!$$
即：把 n 个不同元素排成一排的所有可能方式。
带重复元素的排列：
如果 n 个元素中有 k 类，第 i 类有 nᵢ 个重复元素，则排列数为：
$$\frac{n!}{n_1! \cdot n_2! \cdot \dots \cdot n_k!}$$
案例 4：交易信号序列
某量化策略产生 10 个交易信号：6 个"买入"、4 个"卖出"。这 10 个信号的排列方式（仅区分买入/卖出）有：
$$\frac{10!}{6! \times 4!} = \frac{3{,}628{,}800}{720 \times 24} = 210$$
210 种排列——如果你跟踪一个信号分布为 6:4 的策略，理论上看到的信号序列可能性有 210 种。
4. 组合（Combination）—— 顺序不重要
直觉问题：
从 8 个候选人中选 3 人组成一个委员会（3 人地位平等，没有职务之分）。有多少种选法？
关键判断：委员会内部不区分角色 → 顺序不重要 → 组合问题。
公式：
$$C(n, r) = {}_nC_r = \binom{n}{r} = \frac{n!}{r!(n-r)!}$$
推导逻辑： 1. 先按排列算：选 3 人有 P(8,3) = 336 种方式 2. 但委员会里 3 个人的内部排列数为 3! = 6 3. 这 6 种排列其实对应同一个委员会 → 336 ÷ 6 = 56
$$\binom{8}{3} = \frac{8!}{3! \cdot 5!} = \frac{8 \times 7 \times 6}{3 \times 2 \times 1} = 56$$
🧠 心法：排列 = 先把人选出来（组合），再给每个人安排位置（阶乘）。
$$P(n, r) = C(n, r) \times r!$$
所以：$C(n, r) = P(n, r) / r!$
案例 5：投资组合（选股票不排序）
从 500 只标普成分股中选 10 只构成等权组合。有多少种选法？
$$\binom{500}{10} = \frac{500!}{10! \times 490!}$$
这个数体量巨大（约 $2.46 \times 10^{20}$），远超宇宙中的恒星数量。
📊 这就是分散化的数学基础——组合空间如此巨大，任何"最优组合"都是统计意义上的。
案例 6：用组合算概率
某基金持有 20 只股票。其中 5 只出现重大负面新闻。随机抽查 3 只，恰好抽到 2 只问题股的概率？
思路： - 分母：从 20 只中任选 3 只 → C(20,3) = 1140 - 分子：先选 2 只问题股 C(5,2) = 10，再选 1 只好股 C(15,1) = 15 - 概率 = (10 × 15) / 1140 = 150 / 1140 ≈ 13.2%
三、排列 vs 组合：终极判断法
| 判断标准 | 排列（Permutation） | 组合（Combination） |
|---|---|---|
| 顺序重要吗？ | ✅ 重要 | ❌ 不重要 |
| 典型场景 | 排名、密码、职位安排 | 选人进委员会、选股进组合 |
| 公式 | n!/(n-r)! | n!/[r!(n-r)!] |
| 结果数量 | ≥ 组合（多了一个 r! 因子） | ≤ 排列 |
🧠 一句话判断： 如果选出的 r 个元素互换位置后"不算重复"，那就是排列；如果"算同一种"，那就是组合。
四、CFA 工具箱：常用计数关系
| 组合恒等式 | 含义 |
|---|---|
| C(n,0) = 1 | 一个都不选，只有一种方式 |
| C(n,1) = n | 选 1 个，有 n 种 |
| C(n,n) = 1 | 全选，只有一种方式 |
| C(n,r) = C(n, n-r) | 对称性：选 r 个 = 排除 n-r 个 |
| C(n,0)+C(n,1)+...+C(n,n) = 2ⁿ | n 个元素的所有子集总数 |
📊 C(10,3) = C(10,7) = 120。从 10 只股票选 3 只买入，等价于选 7 只不买。利用对称性可以简化计算。
五、组合的进阶：多项式系数
当把 n 个元素分成 k 组（每组 nᵢ 个），分配方式的数量为：
$$\frac{n!}{n_1! \cdot n_2! \cdot \dots \cdot n_k!}$$
这和"带重复元素的排列"使用同一个公式——同一个数学结构，不同角度解读。
案例 7：资产配置分类
某基金将 12 只持仓股票分为三类：5 只价值型、4 只成长型、3 只混合型。有多少种分类方式？
$$\frac{12!}{5! \times 4! \times 3!} = \frac{479{,}001{,}600}{120 \times 24 \times 6} = 27{,}720$$
六、常见易错点
| 易错 | 正确认识 |
|---|---|
| 排列和组合搞混 ❌ | 先问自己：选出来的东西互换位置，结果变了吗？变了→排列，不变→组合 |
| 0! = 0 ❌ | 0! = 1（约定），这是公式一致性的需要 |
| C(n,r) 直接套公式忘了简化 ❌ | C(n,r) = n×(n-1)×...×(n-r+1) / r!，分子分母各有 r 个因子，可以约分 |
| 忘了带重复元素的排列公式 ❌ | 分母要除每个重复类的阶乘 |
| 用排列算概率忘了分母也是排列 ❌ | 分子和分母的计数方式必须一致（都用排列 or 都用组合） |
七、CFA 考试应考指南
概念题： - 排列 vs 组合的区分标准（顺序是否重要） - 乘法原理的适用条件（各阶段独立） - 阶乘的基本性质（0! = 1） - 组合恒等式的识别
计算题（高频考点）： - 给定 n 和 r，直接计算 P(n,r) 或 C(n,r)——CFA 考试允许使用计算器，用 nPr 和 nCr 按键！ - 用组合算等可能概率（分子分母都用组合） - 带重复元素的排列数 - 多项式系数
八、课后测试题
概念题
Q1： 从 10 个分析师中选 3 人组成项目小组（不区分角色），这是排列还是组合？ A. 排列 B. 组合 C. 两者都可以
Q2： 以下哪项 0! 的值是正确的？ A. 0 B. 1 C. 未定义
Q3： C(7,2) 等于以下哪项？ A. C(7,5) B. P(7,2) / 2 C. 以上都对
计算题
Q4： 6 个候选人面试，最终要录取 2 人，第一名和第二名年薪不同。有多少种结果？ A. 15 B. 30 C. 36
Q5： 从 15 只股票中选 4 只构建等权组合（不区分权重），有多少种选法？ A. 1,365 B. 32,760 C. 1,365 × 24
Q6： 一个"买/卖/持有"的 5 天信号序列，恰好包含 3 个"买"和 2 个"卖"（不含"持有"），有多少种排列？ A. 10 B. 20 C. 60
Q7： 8 人参加会议，要选 1 个主席、1 个秘书、1 个财务（三职不同）。有多少种方式？ A. 56 B. 336 C. 512
Q8： 从一副 52 张标准扑克中抽 5 张。总共有多少种可能的 5 张牌组合？ A. C(52,5) = 2,598,960 B. P(52,5) = 311,875,200 C. 52⁵
九、答案与解析
A1：B — "不区分角色"意味着选出来的三个人内部没有排序，是组合。如果分角色（组长/组员等），才是排列。
A2：B — 0! = 1，这是数学约定，为了保证公式 C(n,0) = n!/(0!·n!) = 1 的一致性。
A3：C — C(7,2) = 7!/(2!5!) = 21。C(7,5) = 21（对称性 ✅）。P(7,2) = 42，42/2 = 21 = C(7,2)（排列 = 组合 × r! ✅）。所以 A、B 都对。
A4：B — "第1名和第2名年薪不同" → 顺序重要 → 排列。P(6,2) = 6×5 = 30。如果用组合 C(6,2)=15 再排列 2!=2 也得 15×2=30。
A5：A — "不区分权重" → 组合。C(15,4) = 15×14×13×12 / (4×3×2×1) = 32,760 / 24 = 1,365。
A6：A — 带重复元素的排列：5!/(3!×2!) = 120/(6×2) = 10。
A7：B — 职位不同 → 排列。P(8,3) = 8×7×6 = 336。也可以理解：C(8,3)×3! = 56×6 = 336。
A8：A — 5 张牌的顺序不重要 → 组合。C(52,5) = 52×51×50×49×48 / (5×4×3×2×1) = 2,598,960。P(52,5) 是排队抽牌，顺序重要。52⁵ 是每次抽完放回，不是扑克。
十、今日小结
| 概念 | 一句话 |
|---|---|
| 乘法原理 | 分步骤，各步选择数相乘 |
| 阶乘 n! | n 个不同东西排队的方式总数 |
| 排列 P(n,r) | 选 r 个还要排序：n!/(n-r)! |
| 组合 C(n,r) | 选 r 个不排序：n!/[r!(n-r)!] |
| 核心关系 | P(n,r) = C(n,r) × r! |
| 带重复排列 | n!/(n₁!·n₂!·...) |
| 计数算概率 | P = 满足条件的组合数 / 总组合数（分子分母同方法） |
🧠 三个核心洞察： 1. 排列 vs 组合的分水岭永远是"顺序重要吗？"——问自己这一句就够了 2. 排列 = 组合 × 排序方式数——C 和 P 本质上是同一棵树的两根枝 3. 用计数算概率时，分子分母的计数方式必须一致——要么都用排列，要么都用组合
明天 L119 将进入概率分布——我们把计数工具和概率概念结合起来，正式进入描述随机变量行为的核心工具。
本内容仅供学习参考，不构成投资建议。
Quantitative Methods — Probability Module · Lesson 4
I. Lesson Positioning
L117 covered conditional probability and Bayes' formula — how to update probability judgments when new information arrives. But so far we've assumed probabilities are "given." Now we step back and ask: how are probabilities calculated in the first place?
When all basic outcomes are equally likely:
$$P(A) = \frac{\text{Number of outcomes in A}}{\text{Total number of possible outcomes}}$$
Both numerator and denominator depend on counting. This lesson teaches how to "count" — permutations and combinations.
| Item | Description |
|---|---|
| Module | 2.4 Probability Theory |
| Prerequisites | L115 Probability Basics, L116 Expected Value & Variance |
| Next Up | L119 Introduction to Probability Distributions |
| Difficulty | ★★★☆☆ |
| Exam Weight | Medium (conceptual + simple calculation, 1–2 questions) |
| Reading Time | ~12 minutes |
II. Core Concepts
1. Multiplication Principle
Intuition:
You go to buy lunch. There are 3 main dishes (rice, noodles, bread) and 4 side dishes (fish, chicken, beef, vegetables). You choose one from each category. How many combinations?
No need to count one by one. Each main dish pairs with 4 sides: 3 × 4 = 12.
This is the essence of the multiplication principle: if a task is completed in k steps, and step i has nᵢ choices, then total outcomes = n₁ × n₂ × ... × nₖ.
Formal Statement:
If a task can be divided into k stages, the i-th stage has nᵢ ways of completion, and stages are independent of each other, then the total number of ways to complete the task is:
$$n_1 \times n_2 \times \dots \times n_k$$
Example 1: Portfolio Construction
An analyst must select one stock, one bond, and one commodity from the following to build a mini portfolio:
| Asset Class | Available Choices |
|---|---|
| Stocks | 50 |
| Bonds | 30 |
| Commodities | 10 |
Total combinations:
$$50 \times 30 \times 10 = 15{,}000$$
📊 Each additional asset class multiplies the portfolio space — this is the "curse of dimensionality" in portfolio optimization.
Example 2: Password Complexity
A 4-digit numeric password, each position 0–9:
$$10 \times 10 \times 10 \times 10 = 10^4 = 10{,}000$$
A 4-character password, each position choosing from 26 uppercase + 26 lowercase + 10 digits = 62 options:
$$62^4 = 14{,}776{,}336$$
📊 From 10K to 14.78M — just by expanding the character set from 10 to 62. This is why websites require "uppercase, lowercase, and numbers" in passwords.
2. Factorial
Definition:
$$n! = n \times (n-1) \times (n-2) \times \dots \times 2 \times 1$$
$$0! = 1 \quad \text{(by convention)}$$
Example: Ranking 5 stocks by conviction, how many ranking orders?
$$5! = 5 \times 4 \times 3 \times 2 \times 1 = 120$$
3. Permutation — Order Matters
Intuitive Question:
From 8 candidates, select 3 to fill positions of CEO, CFO, and COO (three distinct roles). How many ways?
Key judgment: Different positions → order matters → this is a permutation problem.
Formula (r-permutation):
$$P(n, r) = {}_nP_r = \frac{n!}{(n-r)!}$$
Derivation Logic: - Position 1: 8 candidates - Position 2: 7 remaining - Position 3: 6 remaining - Total: 8 × 7 × 6 = 336
Using formula: $P(8, 3) = \frac{8!}{5!} = 8 \times 7 \times 6 = 336$ ✅
Example 3: Fund Manager Rankings
A fund awards Top 10 fund managers from 100 candidates. Number of ways to assign 1st, 2nd, and 3rd place:
$$P(100, 3) = 100 \times 99 \times 98 = 970{,}200$$
📊 Nearly 1 million possibilities! This is why annual fund manager rankings "rotate so frequently" — the permutation space is enormous.
Special Case: Full Permutation
When r = n:
$$P(n, n) = n!$$
That is: all possible ways to arrange n distinct elements in a row.
Permutations with Repeated Elements:
If n elements fall into k categories, with nᵢ repeated elements in category i, the number of distinct permutations is:
$$\frac{n!}{n_1! \cdot n_2! \cdot \dots \cdot n_k!}$$
Example 4: Trading Signal Sequence
A quantitative strategy generates 10 trading signals: 6 "Buy" and 4 "Sell." The number of distinct signal sequences (distinguishing only Buy/Sell) is:
$$\frac{10!}{6! \times 4!} = \frac{3{,}628{,}800}{720 \times 24} = 210$$
210 distinct sequences — if you track a strategy with a 6:4 signal distribution, there are theoretically 210 possible signal sequences.
4. Combination — Order Does NOT Matter
Intuitive Question:
From 8 candidates, select 3 to form a committee (all 3 have equal standing, no distinct roles). How many ways?
Key judgment: Committee members have no ordering → order does NOT matter → this is a combination problem.
Formula:
$$C(n, r) = {}_nC_r = \binom{n}{r} = \frac{n!}{r!(n-r)!}$$
Derivation Logic: 1. Start with permutation: selecting 3 gives P(8,3) = 336 ways 2. But within the committee, the 3 people can be internally arranged in 3! = 6 ways 3. These 6 permutations correspond to the same committee → 336 ÷ 6 = 56
$$\binom{8}{3} = \frac{8!}{3! \cdot 5!} = \frac{8 \times 7 \times 6}{3 \times 2 \times 1} = 56$$
🧠 Mental model: Permutation = Combination × Factorial (sorting within).
$$P(n, r) = C(n, r) \times r!$$
Thus: $C(n, r) = P(n, r) / r!$
Example 5: Stock Portfolio (Unordered Selection)
From 500 S&P 500 constituents, select 10 for an equal-weight portfolio. Number of possible selections:
$$\binom{500}{10} = \frac{500!}{10! \times 490!}$$
This number is astronomically large (approximately $2.46 \times 10^{20}$), far exceeding the number of stars in the universe.
📊 This is the mathematical foundation of diversification — the portfolio space is so vast that any "optimal portfolio" is valid only in a statistical sense.
Example 6: Computing Probabilities with Combinations
A fund holds 20 stocks. Among them, 5 have major negative news. If 3 are randomly inspected, what is the probability of exactly 2 problem stocks?
Approach: - Denominator: choose any 3 from 20 → C(20,3) = 1,140 - Numerator: choose 2 problem stocks C(5,2) = 10, then 1 good stock C(15,1) = 15 - Probability = (10 × 15) / 1,140 = 150 / 1,140 ≈ 13.2%
III. Permutation vs. Combination: The Ultimate Decision Rule
| Judgment Criterion | Permutation | Combination |
|---|---|---|
| Does order matter? | ✅ Yes | ❌ No |
| Typical Scenario | Rankings, passwords, role assignments | Committee selection, stock selection into portfolio |
| Formula | n!/(n−r)! | n!/[r!(n−r)!] |
| Number of Outcomes | ≥ Combination (extra r! factor) | ≤ Permutation |
🧠 One-sentence test: If swapping two selected elements produces a "different" outcome, it's a permutation. If it's "the same," it's a combination.
IV. CFA Toolkit: Common Counting Relationships
| Combinatorial Identity | Meaning |
|---|---|
| C(n,0) = 1 | Selecting none — only one way |
| C(n,1) = n | Selecting 1 — n ways |
| C(n,n) = 1 | Selecting all — only one way |
| C(n,r) = C(n, n−r) | Symmetry: choosing r = excluding (n−r) |
| C(n,0)+C(n,1)+...+C(n,n) = 2ⁿ | Total number of all subsets of n elements |
📊 C(10,3) = C(10,7) = 120. From 10 stocks, choosing 3 to buy is equivalent to choosing 7 NOT to buy. Use symmetry to simplify calculations.
V. Advanced Combinations: Multinomial Coefficient
When partitioning n elements into k groups (with nᵢ elements in group i), the number of partition ways is:
$$\frac{n!}{n_1! \cdot n_2! \cdot \dots \cdot n_k!}$$
This is the same formula as "permutations with repeated elements" — same mathematical structure, different interpretation.
Example 7: Asset Style Classification
A fund classifies 12 holdings into three styles: 5 value, 4 growth, 3 blend. Number of possible classifications:
$$\frac{12!}{5! \times 4! \times 3!} = \frac{479{,}001{,}600}{120 \times 24 \times 6} = 27{,}720$$
VI. Common Pitfalls
| Pitfall | Correct Understanding |
|---|---|
| Mixing up permutation and combination ❌ | Always ask first: if I swap two chosen elements, does the outcome change? Yes → permutation; No → combination |
| 0! = 0 ❌ | 0! = 1 (by convention) — required for formula consistency |
| Plugging C(n,r) directly into formula without simplifying ❌ | C(n,r) = n×(n−1)×...×(n−r+1) / r! — numerator and denominator each have r factors; cancel where possible |
| Forgetting formula for repeated-elements permutation ❌ | Denominator must divide by the factorial of each repeated category |
| Numerator uses combinations but denominator uses permutations ❌ | Counting methods in numerator and denominator must be consistent (both permutations or both combinations) |
VII. CFA Exam Preparation Guide
Conceptual Questions: - Distinguishing permutations vs. combinations (does order matter?) - Applicability of the multiplication principle (independent stages) - Basic factorial properties (0! = 1) - Recognizing combinatorial identities
Calculation Questions (High Frequency): - Given n and r, directly compute P(n,r) or C(n,r) — CFA allows calculator use; use the nPr and nCr keys! - Computing equally-likely probabilities using combinations (numerator and denominator both as combinations) - Number of permutations with repeated elements - Multinomial coefficients
VIII. Practice Questions
Conceptual Questions
Q1: Selecting 3 analysts from 10 to form a project team (no distinct roles). Is this a permutation or combination? A. Permutation B. Combination C. Either works
Q2: Which of the following is the correct value of 0!? A. 0 B. 1 C. Undefined
Q3: C(7,2) equals which of the following? A. C(7,5) B. P(7,2) / 2 C. Both of the above
Calculation Questions
Q4: 6 candidates are interviewed; 2 will be hired. 1st place and 2nd place receive different salaries. How many possible outcomes? A. 15 B. 30 C. 36
Q5: From 15 stocks, select 4 to build an equal-weight portfolio (no weighting distinction). How many ways? A. 1,365 B. 32,760 C. 1,365 × 24
Q6: A 5-day "Buy/Sell/Hold" signal sequence contains exactly 3 "Buy" and 2 "Sell" (no "Hold"). How many distinct arrangements? A. 10 B. 20 C. 60
Q7: 8 people attend a meeting. One chairperson, one secretary, and one treasurer must be selected (three distinct roles). How many ways? A. 56 B. 336 C. 512
Q8: From a standard 52-card deck, 5 cards are drawn. How many possible 5-card hands are there total? A. C(52,5) = 2,598,960 B. P(52,5) = 311,875,200 C. 52⁵
IX. Answers and Explanations
A1: B — "No distinct roles" means the three selected have no internal ordering → combination. If there were distinct roles (team lead / member, etc.), it would be a permutation.
A2: B — 0! = 1, a mathematical convention to ensure consistency: C(n,0) = n!/(0!·n!) = 1.
A3: C — C(7,2) = 7!/(2!5!) = 21. C(7,5) = 21 (symmetry ✅). P(7,2) = 42, 42/2 = 21 = C(7,2) (permutation = combination × r! ✅). So both A and B are correct.
A4: B — "1st and 2nd place have different salaries" → order matters → permutation. P(6,2) = 6×5 = 30. Alternatively: C(6,2)=15 × 2! = 15×2 = 30.
A5: A — "No weighting distinction" → combination. C(15,4) = 15×14×13×12 / (4×3×2×1) = 32,760 / 24 = 1,365.
A6: A — Permutation with repeated elements: 5!/(3!×2!) = 120/(6×2) = 10.
A7: B — Distinct roles → permutation. P(8,3) = 8×7×6 = 336. Alternatively: C(8,3)×3! = 56×6 = 336.
A8: A — Order of cards in a hand does not matter → combination. C(52,5) = 52×51×50×49×48 / (5×4×3×2×1) = 2,598,960. P(52,5) would be drawing cards in sequence (order matters). 52⁵ would be drawing with replacement (not how poker works).
X. Key Takeaways
| Concept | One-Liner |
|---|---|
| Multiplication Principle | Break into steps, multiply choices per step |
| Factorial n! | Number of ways to arrange n distinct items in a row |
| Permutation P(n,r) | Select r AND order them: n!/(n−r)! |
| Combination C(n,r) | Select r WITHOUT ordering: n!/[r!(n−r)!] |
| Core Relationship | P(n,r) = C(n,r) × r! |
| Repeated-Element Permutation | n!/(n₁!·n₂!·...) |
| Counting for Probability | P = favorable combinations / total combinations (same method for numerator and denominator) |
🧠 Three core insights: 1. The dividing line between permutation and combination is always "Does order matter?" — ask yourself this one question and you're set 2. Permutation = Combination × number of ordering ways — C and P are two branches of the same tree 3. When using counting to compute probability, the counting method for numerator and denominator must be consistent — both permutations, or both combinations
Tomorrow, L119 will move into probability distributions — we'll combine counting tools with probability concepts to formally describe the behavior of random variables.
This content is for educational reference only and does not constitute investment advice.