The Mathematics of Maybe: Introduction to Probability

Why This Matters

Life is full of questions with no fixed answer. Will it rain today? Will your school win tomorrow’s hockey match? Will your name be picked in the lucky draw? You cannot know for sure. But you can still say something useful — that one thing is likely and another is unlikely.

This is exactly what your friends do when they argue about the weather. One says, “The sun is shining brightly, so it probably won’t rain.” Another says, “It’s so hot, it might rain later.” Both are guessing from the same sky. That kind of personal guess is fine for a chat. But for serious things — a cricket toss, an insurance plan, a shopkeeper deciding how many mangoes to buy — we need a fair, number-based way to measure chance.

That fair way is probability. It is a kind of measurement, just like length or weight. But instead of measuring how long or how heavy something is, it measures how likely an event is to happen. By the end of this chapter you will be able to put a clear number on “maybe”, from 0 (impossible) all the way to 1 (certain), and explain why that number can never go outside those limits.

The Big Idea

The Big Idea: Probability turns the word “maybe” into a number between 0 and 1. A probability of 0 means the event is impossible. A probability of 1 means it is certain. Everything in between — 0.25, 0.5, 0.75 — measures how likely the event is, with bigger numbers meaning more likely. We find this number in two honest ways: by experimenting and counting what actually happens, or by reasoning about a fair situation where every outcome has an equal chance.

What Makes Something Random?

Before we measure chance, we need to be clear about what kind of situation has chance in it. The key word is random.

A situation is random when you cannot predict exactly what will happen, even though you know all the things that could happen. Think about two simple actions:

• Tossing a coin. You know it will land Heads or Tails. But you cannot say which one before it lands.
• Rolling a die. You know the result will be 1, 2, 3, 4, 5 or 6. But you cannot say which number before it stops.

In both cases you know the full list of possibilities, yet the single result is a surprise. That surprise is randomness. All you can say is what could happen, not what will happen.

These repeatable random actions — tossing, rolling, drawing a slip — are called experiments or trials. A single performance of the experiment (one toss, one roll) is one trial.

An experiment (or trial) is a random action you can repeat, where the result might be different each time and you cannot know it in advance.

Why is something like rain random? Rain depends on so many things at once — temperature, humidity, wind, air pressure — and it is so sensitive to tiny changes in them that no one can predict it with total certainty. We cannot say “it will rain at 3 p.m.” We can only estimate how likely rain is, using past data and patterns.

There is a neat real-life use of randomness you already know. In a cricket match, the captains toss a coin to decide who bats first. This is considered fair precisely because it is random — the coin gives no advantage to either side. Neither captain can fix the result, so both accept it.

The Probability Scale: From Impossible to Certain

Probability is measured on a scale that runs from 0 to 1. This scale is like a number line for chance.

• A probability of 0 means the event is impossible — it simply cannot happen.
• A probability of 1 means the event is certain — it is sure to happen.
• A probability of 0.5 means the event has an even chance — it is just as likely to happen as not.

If your school’s chance of winning is 0.75, that is a 75% chance — more likely to win than not. If it is 0.5, that is a 50% chance — a coin-flip, equally likely to win or lose. If it is 0, winning is impossible (like winning a match you never played). If it is 1, winning is guaranteed.

Most real events live strictly between 0 and 1. The closer the number is to 1, the more likely the event. The closer to 0, the less likely. Figure 7.1 shows this scale with five everyday examples placed on it.

We will look closely at the scale, so here is the picture first:

Tap to zoom

Here are those examples again as a quick table, so you can connect each event to its place on the scale:

Why Probability Can Never Leave the 0-to-1 Range

NCERT tells you probability lives between 0 and 1. But why? Why can it never be 1.4, or −0.3? This is worth understanding, not just accepting. The reason is built into how we calculate it.

In a fair situation, probability is a fraction:

Probability = (number of favourable outcomes) ÷ (total number of possible outcomes)

Now think carefully about the two parts of this fraction.

The favourable outcomes are the ones we are hoping for. They are always part of the full list of possible outcomes. You cannot have more “wins” than there are total possibilities. For example, on a die there are 6 outcomes in total. The number of faces that are “even” is 3. The number that are “less than 10” is all 6. You can never count more than 6 favourable faces, because there are only 6 faces in all.

So the top of the fraction can be at most as big as the bottom. That means:

• The fraction is at most 1 (when every outcome is favourable — a certain event).
• The fraction is at least 0 (when no outcome is favourable — an impossible event; the top is 0).

You also cannot count a negative number of favourable outcomes — counting always starts at zero. So the value can never dip below 0. Put together, every probability P(E) obeys:

0 ≤ P(E) ≤ 1

That is the whole reason for the range. It is not a rule someone invented. It falls straight out of “favourable outcomes are a part of the total outcomes”. Figure 7.2 below shows this “part-of-the-whole” idea clearly when we look at events.

Measuring Probability Objectively

A personal guess (“I feel it’ll rain”) is called subjective probability. It depends on the person. To be fair and scientific, we want an objective estimate that anyone would agree with. There are two honest ways to get one.

1. From experience (experimental probability). Do the experiment many times, or look at past data, and count what actually happened.

2. From reasoning (theoretical probability). Assume the situation is perfectly fair — every outcome equally likely — and reason out the chance without doing any experiment.

Let’s take each one in turn.

Experimental Probability: Count What Actually Happens

To get an experimental probability, you perform the experiment and record the results. Before we count, we need two words.

An outcome is one possible result of the experiment. The sample space is the set of all possible outcomes, written inside curly brackets and separated by commas. Figure 7.3 shows the sample space for a coin and for a die.

Here is the picture before the words pile up:

Tap to zoom

Now the formula for experimental probability:

Experimental Probability = (number of times the event occurred) ÷ (total number of trials)

Let’s use it on a real count.

The phrase relative frequency just means “how often the event happened compared to the total”. It is the same number as the experimental probability. It is very useful when you are working with real, observed data instead of theory.

Theoretical Probability: Reason About a Fair Situation

Sometimes you don’t need to do any experiment at all. If the situation is perfectly fair — every outcome equally likely — you can reason out the probability directly. This is theoretical probability, written as P(Event).

Theoretical Probability, P(E) = (number of favourable outcomes) ÷ (number of possible outcomes)

It tells you what we expect in an ideal, fair world. No coins are tossed, no dice are rolled — you just reason.

Notice this assumes the die is fair, so each face is equally likely. That equal-likeliness is what lets us just count and divide. Let’s try one with letters.

Using Real-World Data to Estimate Probability

The “collect data and count” method is used everywhere — in business for forecasting sales, in insurance, and in science. The idea is simple: if you have data on what happened, you can estimate the chance of it happening again.

This estimate is powerful. Suppose you must buy fruit for the whole school of 1500 students. You cannot ask all 1500. So you ask a sample of 50 and use their answer to estimate the whole. If 40% of the sample likes mango, you estimate that about 40% of 1500 = 600 students in the school like mango. So you buy roughly 600 mangoes. The smaller group (50 students) is the sample; the whole group (1500 students) is the population. This clever shortcut — studying a small fair sample to learn about a big group — is called sampling.

To trust the estimate more, you would take a bigger and fairer sample — say 100 students from different classes, not just one. A bigger, fairer sample gives a more reliable picture.

Experimental vs Theoretical: Why They Differ, and the Law of Large Numbers

You now have two ways to find probability. They usually give close answers, but not always the same answer. Why?

Theoretical probability of rolling a 4 is exactly 1/6 ≈ 0.167. But in our experiment of 50 rolls, the experimental value came out 0.16. Slightly different. This is normal. Even with a perfectly fair die, the experimental result wobbles around the theoretical one — especially when you do only a few trials.

Here is the beautiful part. As you do more and more trials, the experimental probability creeps closer and closer to the theoretical one. Toss a coin 10 times and you might get 7 heads (0.7, far from 0.5). Toss it 10,000 times and the fraction of heads will sit very near 0.5. This rule has a name: the Law of Large Numbers.

Figure 7.4 below shows exactly why this happens — the experimental line swings wildly at first, then settles onto the theoretical line:

Tap to zoom

So remember the key idea in that “Think and Reflect” from your textbook: even if you have just rolled a 4 eight times in a row, the chance of a 4 on the next roll is still about 0.167. Probability does not promise anything about the very next try. It describes the long run.

The Gambler’s Fallacy: The Coin Has No Memory

This leads to one of the most common mistakes in all of probability. After a fair coin lands heads six times in a row, many people feel that tails is now “due” — that tails must be more likely on the next toss. This feeling is wrong. It is so famous that it has a name: the Gambler’s Fallacy.

Here is the truth. The coin has no memory. It does not know or care what happened on the last six tosses. Every toss starts fresh. So after six heads, the chance of tails on the next toss is still exactly 1/2 — the same as it always was.

The same applies to dice. In Snakes and Ladders, suppose you roll three 6s in a row. You might think, “No way I’ll get a 6 again!” But each roll is independent — it does not depend on the rolls before it. The chance of a 6 stays 1/6 ≈ 0.166 every single time. The Gambler’s Fallacy tricks you into seeing a pattern where there is none. Random events simply have no memory.

A fair (unbiased) coin is symmetrical, so there is no reason for it to land on one side more than the other. A random toss means it is allowed to fall freely, with no interference. Fair and unbiased is what lets us say heads and tails are equally likely.

Sample Spaces and Events, More Carefully

We met the sample space earlier. Let’s nail down the exact rules, because they matter for every problem.

The sample space, written S, is the list of every possible outcome. Each outcome in it is called an element. Three rules:

• S must include every possible outcome (nothing left out).
• No outcome is listed twice.
• The number of outcomes in S is the sample size, written n(S).

Here are some sample spaces:

• Will it rain tomorrow? S = {Rain, No Rain}, so n(S) = 2.
• A match result. S = {Win, Lose, Draw}, so n(S) = 3.
• Tossing one coin. S = {H, T}, so n(S) = 2.
• Rolling one die. S = {1, 2, 3, 4, 5, 6}, so n(S) = 6.
• Tossing two coins together. S = {HH, HT, TH, TT}, so n(S) = 4.

One important point: how detailed the sample space should be depends on the question. For “will it rain?”, S = {Rain, No Rain} is fine. But if you care about how much rain, you would expand it to S = {No Rain, Drizzle, Light Rain, Heavy Rain}. Make the sample space match the level of detail your question needs.

What Is an Event?

An event is any one outcome, or any group of outcomes, that we are interested in. In set language, an event is a subset of the sample space — it is a chosen “part” picked out of all the possibilities. This is the exact idea that made probability stay at or below 1: the event is always a part of the whole. Figure 7.4 below makes that “part-of-the-whole” picture clear.

Tap to zoom

Some more sample spaces with an event picked from each:

• Tossing two coins. S = {HH, HT, TH, TT}. Event “at least one head”: E = {HH, HT, TH}.
• Rolling a die. S = {1, 2, 3, 4, 5, 6}. Event “number greater than 4”: E = {5, 6}.
• Picking fruit. S = {Apple, Banana, Orange}. Event “the fruit is yellow”: E = {Banana}.

To find the probability of an event, count the outcomes inside the event and divide by n(S). For “number greater than 4”: E = {5, 6} has 2 outcomes, S has 6, so P(E) = 2/6 = 1/3.

Tree Diagrams: Mapping Out Multi-Step Experiments

So far our experiments had one step (one toss, one roll). But many experiments have several steps — tossing a coin twice, or rolling a die three times. These are multi-step experiments. Listing all outcomes can get confusing. A tree diagram keeps it organised.

A tree diagram draws a branch for each outcome of each step. From the start point, you branch out for the first step. From the end of each of those branches, you branch out again for the second step, and so on. Each complete path from start to finish is one outcome. Reading off all the paths gives the full sample space.

Figure 7.5 below builds the tree for tossing a coin twice:

Tap to zoom

From the tree, the sample space is S = {HH, HT, TH, TT}, so n(S) = 4. Each outcome is equally likely, so each has probability 1/4 = 0.25. For example, the event “getting Heads twice” is just the single outcome HH, so:

P(HH) = 1/4 = 0.25, or 25%

What about “one head and one tail”? That event is E = {HT, TH} — two of the four outcomes. So P(one head and one tail) = 2/4 = 1/2 = 0.5. The tree makes it easy to see that two of the four paths fit.

Common Mistakes

Probability has a few traps that catch almost everyone. Here are the big ones.

Quick Check

Test yourself before the practice problems.

Practice Problems

Try each one yourself before opening the solution.

Easy

Medium

Challenge

Summary

After this chapter, you can now explain:

• Why some situations are random — you know all the possible outcomes but cannot predict the single result of one trial.
• How probability sits on a scale from 0 (impossible) to 1 (certain), and why it can never leave that range: favourable outcomes are always part of the total, so 0 ≤ P(E) ≤ 1.
• The difference between experimental probability (count what actually happened ÷ total trials) and theoretical probability (favourable outcomes ÷ total possible outcomes, in a fair situation).
• Why the two values differ for small numbers of trials, and how the Law of Large Numbers pulls the experimental value toward the theoretical one as trials increase.
• Why the Gambler’s Fallacy is wrong — random events have no memory, so each toss or roll is independent.
• What a sample space S and an event (a subset of S) are, how to find n(S), and how to use tree diagrams to list all outcomes of a multi-step experiment.
• How to estimate real-world probabilities from a sample and scale them up to a whole population.

What’s Next

You have just learned to measure uncertainty — to put a number on “maybe”. Next, we turn from chance to pattern: numbers that follow a clear, predictable rule. In Chapter 8 — Predicting What Comes Next: Exploring Sequences and Progressions, you will spot the rule behind a list of numbers, predict the next term, and find a formula for any term you like. Where probability dealt with what we cannot be sure of, sequences are all about what we can predict exactly.