All lessons

Module 71: Probability · Lesson 71.1 · about 34 min

Probability

Sample spaces and classical probability, the addition rule and complements, conditional probability, independence, total probability and Bayes theorem, random variables with their mean and variance, and the binomial distribution, up to the game and geometric-trial questions of JEE Advanced.

Key points, meanings and formulas on one page: choose "Save as PDF" in the print box.

Maths · Class 12 · about 34 minutes. By the end you will be able to count outcomes correctly, use the addition, multiplication and complement rules, update a probability when new information arrives (conditional probability and Bayes theorem), and work with random variables and the binomial distribution, including “who wins the game” questions.

Probability is counting with a fair eye: list what can happen, decide which outcomes are equally likely, and count the ones you want. Conditional probability is the same counting done inside a smaller world, the world where you already know something happened.

1. Sample space and classical probability

A random experiment has outcomes you cannot predict in advance. The set of all outcomes is the sample space S, and an event is any subset of S. When all outcomes are equally likely,

P(E)=n(E)n(S)

Two dice give 36 equally likely ordered pairs, not 11 sums: a sum of 7 has 6 pairs, a sum of 2 has only one. Cards: 52 cards, 4 suits of 13, 12 face cards (J, Q, K), 4 aces.

For selections without replacement, count with combinations: drawing 2 red from a bag of 5 red and 3 black has probability 5C28C2=1028.

“There are 11 possible sums, so a sum of 7 has probability 1/11.” The 11 sums are not equally likely. Always count in a sample space whose outcomes are equally likely: for two dice, the 36 ordered pairs.

2. Addition rule and complements

P(A∪B)=P(A)+P(B)−P(A∩B),P(A′)=1−P(A)

Subtracting P(A∩B) removes the outcomes counted twice. For three events, add the singles, subtract the pairs and add back the triple. Mutually exclusive events cannot happen together, so P(A∩B)=0.

The complement is the quickest route for “at least one”: P(at least one six in n throws) = 1−(5/6)n.

Useful pieces: P(only A) = P(A) − P(A ∩ B); P(exactly one of A, B) = P(A) + P(B) − 2P(A ∩ B); P(neither) = 1 − P(A ∪ B).

3. Conditional probability and independence

The probability of A given that B has happened:

P(A|B)=P(A∩B)P(B),P(B)≠0

Think of it as shrinking the sample space to B and asking what fraction of B is also A. Rearranged, it is the multiplication rule: P(A∩B)=P(B)P(A|B), used for one draw after another.

A and B are independent when knowing one does not change the other: P(A∩B)=P(A)P(B). Three events are mutually independent only if every pair and the triple satisfy the product rule.

“Mutually exclusive events are independent.” The opposite is true. If A and B cannot happen together and both have positive probability, then knowing B happened tells you A did not: P(A | B) = 0 ≠ P(A). Exclusive events (with nonzero probabilities) are always dependent.

4. Total probability and Bayes theorem

Suppose the sample space splits into causes E₁, E₂, …, Eₙ that do not overlap and cover everything. An effect A can happen through any of them:

P(A)=∑iP(Ei)P(A|Ei),P(Ek|A)=P(Ek)P(A|Ek)∑iP(Ei)P(A|Ei)
E₁, …, Eₙ
the causes (which bag, which machine, which test result); exactly one of them happens
P(Eᵢ)
prior probability: how likely the cause was before you saw anything
P(A | Eᵢ)
how likely the effect is if that cause is true
P(Eₖ | A)
posterior probability: how likely cause k is now that the effect has been seen
Physical meaning
Bayes is a share of a total: of all the ways A can happen, what fraction came through Eₖ? Drawing a tree with one branch per cause makes this a two-line sum.

Worked example. A test for a disease is 95% accurate on sick people and gives a false alarm for 2% of healthy people. 1% of people are sick. If a person tests positive, P(sick) = (0.01 × 0.95) ÷ (0.01 × 0.95 + 0.99 × 0.02) = 0.0095 ÷ 0.0293 ≈ 0.32. Only about a third: the healthy group is so large that its few false alarms outnumber the true positives.

5. Random variables, mean and variance

A random variable X gives a number to each outcome (number of heads, money won). Its probability distribution lists each value xᵢ with its probability pᵢ, where every pᵢ ≥ 0 and the pᵢ add to 1.

μ=E(X)=∑pixi,Var⁡(X)=∑pixi2−μ2

The mean is the long-run average value; the variance measures the spread around it, and the standard deviation is its square root. Rules: E(aX + b) = aE(X) + b and Var(aX + b) = a²Var(X) (adding a constant does not change spread).

Fair game. A game is fair when the expected gain is zero. Expected gain = (expected amount received) − (fee paid).

6. Bernoulli trials and the binomial distribution

Bernoulli trials: a fixed number n of independent trials, each with two outcomes and the same success probability p (q = 1 − p). The number of successes X is binomial, B(n, p):

P(X=r)=nCrprqn−r,E(X)=np,Var⁡(X)=npq

ⁿCᵣ counts the positions the r successes can take; pʳqⁿ⁻ʳ is the probability of any one such sequence. The most likely value (mode) sits near np. Try it:

B(n, p). The coloured bars are the event you choose; the dashed line is the mean np. Notice the spread npq is largest when p = 0.5.

Given mean and variance. If np = 4 and npq = 3, then q = 3/4, p = 1/4 and n = 16. Variance can never exceed the mean for a binomial, since q < 1.

7. JEE Advanced level: games and repeated trials

Taking turns. A and B throw a die in turn, A first; whoever throws a six first wins. A wins on throw 1, 3, 5, …:

P(A)=p+q2p+q4p+⋯=p1−q2=1/61−25/36=611

First success on trial k (geometric trials): qk−1p. At least one success in n trials: 1−qn; to find the least n that makes this exceed a target, take logarithms.

Geometric probability. When a point is chosen “at random” in a region, probability = favourable length, area or volume ÷ total. Example: two numbers x, y are picked in [0, 1]; P(x + y < 1) is the area of the triangle below the line, 1/2.

“If heads came five times in a row, tails is now due.” Independent trials have no memory. The sixth toss is still 1/2 heads. Streaks feel unusual, but every particular sequence of six tosses has the same probability, 1/64.

Next: Probability: question types and traps.

  • P(E) = n(E)/n(S) only when outcomes are equally likely; for two dice use 36 ordered pairs.
  • P(A ∪ B) = P(A) + P(B) − P(A ∩ B); “at least one” = 1 − P(none).
  • P(A | B) = P(A ∩ B)/P(B); independence means P(A ∩ B) = P(A)P(B). Exclusive events with positive probabilities are dependent.
  • Total probability adds the routes; Bayes takes one route’s share of that total. Draw the tree.
  • E(X) = Σpx, Var(X) = Σpx² − μ²; Var(aX + b) = a²Var(X).
  • Binomial: P(X = r) = ⁿCᵣpʳqⁿ⁻ʳ, mean np, variance npq.
  • Alternate turns: P(first player wins) = p/(1 − q²).

Check yourself

1P(A) = 0.5, P(B) = 0.4 and A, B are independent. What is P(A ∪ B)?
2A fair coin is tossed 4 times. The probability of at least one head is:
3For a binomial variable with n = 10 and p = 0.2, the variance is:

Sources

  • NCERT Mathematics Class 11 (Probability) and Class 12 (Probability): topics followed, no text copied
  • JEE Main and JEE Advanced 2026 syllabi: Statistics and probability

Your ticks stay in this browser. Save them as a file on the course home.

Updated 2026-11-22. Syllabus and exam patterns follow the official notices checked on 2026-11-22; they can change every year, so check the official exam website. All questions are original fibub work, not copied from past papers or books. Nothing you do here is sent anywhere.