IB AA: Further Probability with Counting

Probability tells you what could happen, while counting tells you how many ways it can happen. Mastering their combination is essential for solving complex, real-world problems where events are defined by intricate arrangements, selections, and conditions. This topic moves beyond basic probability rules, using the precision of permutations and combinations to analyze sophisticated scenarios with clarity and confidence.

The Combinatorial Bridge to Probability

The foundational link is the classical probability formula: $P (Event) = \frac{Number of favorable outcomes}{Total number of equally likely outcomes}$ . When outcomes are defined by arrangements or selections, permutations (order matters) and combinations (order does not matter) provide the exact counts for this formula.

Consider a committee of 4 people chosen from 10. The total number of possible committees is a combination: $(4 10) = 210$ . If the "favorable" event is selecting a committee containing 2 specific people, you count the favorable outcomes in stages. First, place the 2 specific members on the committee. Then, choose the remaining 2 members from the other 8 people: $(2 8) = 28$ . Thus, the probability is $\frac{28}{210} = \frac{2}{15}$ .

This staged counting is the core technique. You must correctly identify whether each stage involves a permutation or combination. For instance, forming a line-up uses permutations, while forming a group uses combinations. Misidentifying this is a common source of error.

Conditional Probability in a Counting Context

Conditional probability, $P (A ∣ B)$ , asks for the probability of event $A$ given that event $B$ has already occurred. With counting, you can often compute this directly by redefining the sample space. The formula $P (A ∣ B) = \frac{P ( A \cap B )}{P ( B )}$ still holds, but you can frequently compute it as: $P (A ∣ B) = \frac{Number of outcomes in A \cap B}{Number of outcomes in B} .$

Imagine dealing a 5-card poker hand from a standard deck. What is the probability of getting a full house (three of one rank, two of another) given that the hand contains at least two Aces? The condition "contains at least two Aces" becomes your new, restricted sample space. First, count hands with at least two Aces. This is easier by complementary counting: total hands minus hands with 0 or 1 Ace.

Total 5-card hands: $(5 52)$ .
Hands with no Aces: Choose 5 cards from the 48 non-Aces: $(5 48)$ .
Hands with exactly one Ace: Choose 1 Ace from 4, and 4 non-Aces from 48: $(1 4) (4 48)$ .

Therefore, hands with at least two Aces = $(5 52) - (5 48) - (1 4) (4 48)$ .

Next, count the favorable intersection: full houses with at least two Aces. For a full house containing Aces, the Aces must form the "three of a kind" part (you can't have two Aces and three of another rank, as that would be four of a kind). So, choose 3 of the 4 Aces, then choose a different rank for the pair, and choose 2 of its 4 cards: $(3 4) \times (1 12) \times (2 4)$ .

The conditional probability is the ratio of these two counts. This method often avoids the complex fractions of the standard conditional formula.

The Power of Multinomial Coefficients

When you need to arrange or allocate items into more than two groups, the multinomial coefficient is your essential tool. It generalizes the binomial coefficient. The number of ways to arrange $n$ distinct objects where there are $n_{1}$ of type 1, $n_{2}$ of type 2, ..., $n_{k}$ of type $k$ (with $n_{1} + n_{2} + ... + n_{k} = n$ ) is: $(n _{1} , n _{2} , ... , n _{k} n) = \frac{n !}{n _{1} ! \times n _{2} ! \times ... \times n _{k} !} .$

This is also the number of ways to partition $n$ distinct items into $k$ labeled groups of specified sizes. Probability problems often involve dealing cards or distributing people into teams.

Example: A standard deck is shuffled. What is the probability that in a 13-card hand, you get exactly 5 spades, 4 hearts, 3 diamonds, and 1 club?

The total number of 13-card hands: $(13 52)$ .
Count favorable hands using the multinomial framework. First, choose which 5 of the 13 spades you get: $(5 13)$ . Choose 4 of the 13 hearts: $(4 13)$ . Choose 3 of the 13 diamonds: $(3 13)$ . Choose 1 of the 13 clubs: $(1 13)$ .

The number of favorable hands is the product: $(5 13) (4 13) (3 13) (1 13)$ . The probability is this product divided by $(13 52)$ .

Notice the connection: $(13 52)$ itself can be thought of as a multinomial coefficient $(13 , 13 , 13 , 13 52)$ for partitioning the deck into your hand and the remaining 39 cards, but it's simpler to use combinations here. The multinomial concept is most powerful when the groups are defined within the selection.

Constructing Distributions from Combinatorial Arguments

Some problems ask you to find the probability distribution of a discrete random variable defined by a counting scenario. This involves computing $P (X = k)$ for all possible values of $k$ , using combinatorial reasoning.

Classic Scenario: The hypergeometric distribution. Suppose you have a population of $N$ items with $K$ "successes" and $N - K$ "failures." You select $n$ items without replacement. Let $X$ be the number of successes in your sample. The probability mass function is derived purely from combinations: $P (X = k) = \frac{( k K ) ( n - k N - K )}{( n N )}, for max (0, n - (N - K)) \leq k \leq min (n, K) .$

Worked Example: An urn has 8 red balls (successes) and 7 white balls (failures). You draw 5 balls without replacement. What is the probability distribution for $X$ , the number of red balls drawn?

Population: $N = 15$ , $K = 8$ , $n = 5$ .
Possible $k$ values: 0, 1, 2, 3, 4, 5. However, $k$ cannot be 0 because you are drawing 5 from only 7 white balls (you must get at least 1 red). Similarly, $k$ cannot be 5 because you are drawing 5 from only 8 reds (you must get at least 1 white). So $k = 1, 2, 3, 4$ .
Calculate: $P (X = 2) = \frac{( 2 8 ) ( 3 7 )}{( 5 15 )}$ .
Repeat this calculation for $k = 1, 3, 4$ to build the full distribution table. This combinatorial argument is the definition of the hypergeometric distribution, a key link between counting and probability models.

Common Pitfalls

Confusing Permutations and Combinations: The most frequent error. Ask: "Does the order of selection or arrangement change the identity of the outcome?" If yes, use permutations ( $P (n, r)$ or $n P r$ ). If no, use combinations ( $C (n, r)$ or $n C r$ ). In probability, misapplying these will give an incorrect count for both numerator and denominator.

Correction: For a committee, order doesn't matter (combination). For electing a president, vice-president, and secretary from a group, order does matter (permutation).

Misapplying the Fundamental Counting Principle: This principle (multiplying choices for stages) assumes choices at one stage are independent of previous choices. This fails if selections are made without replacement and the pool changes.

Correction: Use combinations for simultaneous selections, or carefully track the changing pool in sequential selections. For example, the probability the second card is an Ace given nothing about the first is $\frac{4}{52}$ , but the probability the second is an Ace given the first was an Ace is $\frac{3}{51}$ .

Incorrect Conditional Probability Setup: When using counting for $P (A ∣ B)$ , students often divide by the original total outcomes instead of the restricted outcomes in event $B$ .

Correction: Mentally (or literally) impose the condition first. Your denominator must be the count of all outcomes that satisfy condition B. Your numerator is the subset of those that also satisfy A.

Overlooking "At Least" Problems: Directly counting outcomes for "at least one" or "at least two" can be very tedious.

Correction: Use complementary counting. $P (at least one) = 1 - P (none)$ . This transforms a complex multi-case count into a single, usually simpler, calculation.

Summary

Counting provides the foundation for precise probability calculations in complex scenarios. Always define your probability as a ratio of carefully counted favorable and total outcomes using permutations, combinations, or the multinomial coefficient.
Conditional probability can be tackled by redefining the sample space according to the given condition and using combinatorial counts within that restricted space.
The multinomial coefficient $\frac{n !}{n _{1} ! n _{2} ! ... n _{k} !}$ is the essential tool for counting arrangements with multiple types or partitions into multiple labeled groups, frequently appearing in card probability problems.
Probability distributions, like the hypergeometric distribution, are built directly from combinatorial arguments, formalizing the process of sampling without replacement.
Success in exam problems hinges on correctly identifying whether order matters and strategically using complementary counting or staged counting to simplify complex event descriptions.

IB AA: Further Probability with Counting

IB AA: Further Probability with Counting

The Combinatorial Bridge to Probability

Conditional Probability in a Counting Context

The Power of Multinomial Coefficients

Constructing Distributions from Combinatorial Arguments

Common Pitfalls

Summary

Write better notes with AI