What Is Bayes' Theorem?

Bayes' Theorem is a rule for updating probability when new evidence is known. In ordinary probability questions, you might be asked for the probability of event \(A\). In conditional probability questions, you might be asked for the probability of event \(A\) given that event \(B\) has happened, written \(P(A\mid B)\). Bayes' Theorem is the formula that lets you calculate this updated probability when the information you are given is often the reverse conditional probability, \(P(B\mid A)\).

The theorem is named after Thomas Bayes, but students usually meet it as a practical method rather than a historical result. It appears in medical testing, spam filters, machine learning, genetics, weather prediction, risk analysis, quality control, legal evidence, and exam probability questions. Its power comes from combining a prior probability with new evidence. The prior is what you believe before seeing the evidence; the posterior is what you believe after the evidence is included.

In this formula, \(A\) is usually the event or hypothesis you care about, and \(B\) is the evidence you have observed. For example, \(A\) might mean "the patient has a disease" and \(B\) might mean "the test result is positive." A medical test often gives \(P(B\mid A)\), the probability of a positive result if the patient truly has the disease. But the patient wants \(P(A\mid B)\), the probability of truly having the disease after receiving a positive result. Bayes' Theorem turns the given direction into the direction you need.

If you need a narrower tool for direct conditional probability calculations, use the Conditional Probability Calculator. If you want a broader review of probability ideas before studying this theorem, the Laws of Probability page and the Probability guide are useful foundations.

Bayes' Theorem Formula

The most common form of Bayes' Theorem is:

The formula should be read carefully. The vertical bar means "given that." Therefore \(P(A\mid B)\) means "the probability of \(A\), given that \(B\) has occurred." It does not mean \(P(A)\) divided by \(P(B)\). It is a conditional probability, not an arithmetic fraction.

When \(P(B)\) is not given directly, you usually find it using the law of total probability. If the only alternatives are \(A\) and not \(A\), then:

Here \(A'\) means "not \(A\)" or the complement of \(A\). If \(P(A)=0.01\), then \(P(A')=0.99\). In a medical-test example, \(P(B\mid A')\) would usually be the false positive rate: the chance of a positive test even though the patient does not have the disease.

Some textbooks use \(H\) for the hypothesis and \(E\) for the evidence. The same formula becomes:

This version can be easier to understand in words: posterior equals likelihood times prior, divided by evidence. The labels change, but the logic is identical. In school exams, \(A\) and \(B\) are common. In statistics, machine learning and data science, \(H\) and \(E\) are also common.

Conditional Probability Behind Bayes' Theorem

Bayes' Theorem comes directly from the definition of conditional probability. The probability of \(A\) given \(B\) is:

The probability of \(B\) given \(A\) is:

Rearranging the second equation gives:

Substituting that into the first equation gives Bayes' Theorem:

This derivation matters because it shows that Bayes' Theorem is not a separate trick. It is conditional probability written in a useful direction. If you understand conditional probability, Bayes' Theorem is a way to reverse the condition by using the joint probability \(P(A\cap B)\).

The most common student error is thinking that \(P(A\mid B)\) and \(P(B\mid A)\) are the same. They are not. The probability that a person has a disease given a positive test is different from the probability that the test is positive given the person has the disease. The first is a diagnostic probability; the second is a test accuracy statement. Bayes' Theorem connects them but does not make them equal.

How to Solve Bayes' Theorem Problems

Most Bayes' Theorem questions can be solved with a steady five-step method. The method works for disease tests, bags and balls, pass rates, defective products, spam filtering, and many exam contexts.

1. Define the events. Write what \(A\) means and what \(B\) means. Usually \(A\) is the hidden cause or group, and \(B\) is the observed evidence.
2. Write the required probability. Convert the question into notation. If the question says "given that the test is positive," then the positive test belongs after the vertical bar.
3. List the probabilities given in the question. Identify \(P(A)\), \(P(A')\), \(P(B\mid A)\), and \(P(B\mid A')\). Convert percentages into decimals before calculating.
4. Find the denominator. If \(P(B)\) is not given, calculate it using \(P(B)=P(B\mid A)P(A)+P(B\mid A')P(A')\).
5. Substitute into Bayes' Theorem and interpret the answer. State the result as a decimal, fraction or percentage, and explain what it means in the context of the question.

In an exam, the setup usually earns as much value as the arithmetic. A clear line defining the events can stop you from reversing the conditional probability. For example, if \(D\) means disease and \(+\) means positive test, then the question "What is the probability the person has the disease given a positive test?" is \(P(D\mid +)\), not \(P(+\mid D)\).

Tree diagrams are often the easiest way to organize the denominator. Put the prior probabilities on the first branches, then the conditional probabilities on the second branches. The probability of each final path is found by multiplying along the branches. The total probability of the observed evidence is found by adding all paths that lead to that evidence. This is exactly what the denominator does.

The Probability Formula Booklet and Statistics and Probability Formulas page can help when you want Bayes' Theorem alongside other probability rules such as complements, unions, intersections, independent events and expected values.

Bayes' Theorem Calculator

Use this calculator to check Bayes' Theorem calculations. You can enter probabilities as decimals such as \(0.04\) or as percentages such as \(4\). The calculator has two modes: enter \(P(B)\) directly, or let the calculator find \(P(B)\) from \(P(B\mid A)\), \(P(A)\), \(P(B\mid A')\), and \(P(A')\).

The calculator is intended as a checking tool. For full marks in a written solution, still show your definitions of \(A\) and \(B\), your denominator calculation, and your final interpretation. A numerical answer without context is incomplete in many probability questions.

Worked Example 1: Medical Test

A rare disease affects \(1\%\) of a population. A test correctly gives a positive result for \(95\%\) of people who have the disease. It gives a false positive result for \(4\%\) of people who do not have the disease. A randomly selected person tests positive. What is the probability that the person actually has the disease?

Define the events:

• \(D\): the person has the disease.
• \(+\): the test result is positive.

The question asks for \(P(D\mid +)\). The given probabilities are:

First calculate the total probability of a positive result:

Now apply Bayes' Theorem:

This example is important because it shows the base-rate effect. The true positive rate is high, but the prior probability of the disease is low. When the condition is rare, false positives can make up a large share of all positive results. Bayes' Theorem forces the denominator to include both true positives and false positives.

Worked Example 2: Exam Pass Probability

In a class, \(60\%\) of students studied consistently and \(40\%\) did not. Of those who studied consistently, \(80\%\) passed an exam. Of those who did not study consistently, \(30\%\) still passed. A student is chosen at random and is known to have passed. What is the probability that the student studied consistently?

Let \(S\) mean "studied consistently" and \(P\) mean "passed." The question asks for \(P(S\mid P)\).

Find the total probability of passing:

Apply Bayes' Theorem:

Notice that the posterior equals \(80\%\), but not because the formula always returns the pass rate for studied students. It happens here because \(P(P)=0.60\) and the numerator is \(0.48\). In a different class distribution, the answer would change. Always calculate the denominator rather than guessing from one conditional percentage.

Worked Example 3: Bags and Coloured Balls

Bag A contains 3 red balls and 7 blue balls. Bag B contains 6 red balls and 4 blue balls. A bag is chosen at random, and then one ball is drawn. The ball is red. What is the probability that the bag chosen was Bag A?

Let \(A\) mean "Bag A was chosen" and \(R\) mean "a red ball was drawn." The question asks for \(P(A\mid R)\).

First find the total probability of drawing a red ball:

Then apply Bayes' Theorem:

Worked Example 4: Defective Product

A factory uses two machines. Machine X makes \(70\%\) of the products and Machine Y makes \(30\%\). Machine X has a defect rate of \(5\%\), while Machine Y has a defect rate of \(8\%\). A randomly selected product is defective. What is the probability it was made by Machine X?

Let \(X\) mean "made by Machine X" and \(D\) mean "defective." The question asks for \(P(X\mid D)\).

The total probability of a defective item is:

Therefore:

Bayes' Theorem With Fractions

Bayes' Theorem can be solved using decimals, percentages or fractions. Fractions are often cleaner in bag, dice, card and coin problems because the probabilities come from counts. Suppose a box contains 4 fair coins and 1 two-headed coin. A coin is selected at random and flipped once. It lands heads. What is the probability that the selected coin was the two-headed coin?

Let \(T\) mean "two-headed coin" and \(H\) mean "heads appears."

Find the total probability of heads:

Now calculate the posterior probability:

The answer is \(\frac{1}{3}\), not \(\frac{1}{5}\), because seeing heads increases the chance that the coin is the two-headed coin. It is not \(1\), because a fair coin can also land heads.

Bayes' Theorem, Tree Diagrams and Two-Way Tables

Bayes' Theorem problems can be represented in three equivalent ways: formulas, tree diagrams and two-way tables. The formula is compact, but diagrams and tables often make the denominator easier to see.

In a tree diagram, the first split usually represents the prior probabilities, such as disease or no disease. The second split represents the conditional probabilities, such as positive or negative test. Each final branch is found by multiplying along a path. For example, true positives are \(P(+\mid D)P(D)\), while false positives are \(P(+\mid D')P(D')\). The denominator \(P(+)\) is the sum of all positive paths.

In a two-way table, rows might represent disease status and columns might represent test result. If a population size is not given, choosing a convenient number such as 10,000 can make the percentages concrete. If \(1\%\) of 10,000 people have a disease, then 100 people have it and 9,900 do not. If the true positive rate is \(95\%\), then 95 of the diseased people test positive. If the false positive rate is \(4\%\), then 396 of the healthy people test positive. Out of \(95+396=491\) positive tests, only 95 are true positives, so the posterior probability is \(95/491\approx19.4\%\).

This natural-frequency method is especially helpful when Bayes' Theorem feels counterintuitive. It shows exactly why the denominator includes every way the evidence can appear. Students who struggle with symbols often understand the table first, then return to the formula after the structure is clear.

Real-World Uses of Bayes' Theorem

Bayes' Theorem is widely used because real decisions rarely happen with complete information. New evidence arrives gradually, and each piece of evidence changes what is reasonable to believe.

These applications are different, but the structure is the same: begin with a prior probability, observe evidence, calculate how likely the evidence is under each scenario, and update the probability. The theorem gives a disciplined way to avoid overreacting to evidence without ignoring it.

Common Mistakes in Bayes' Theorem

The most common mistake is reversing the conditional probability. If a question gives the probability of a positive test given disease, that is \(P(+\mid D)\). If the question asks for the probability of disease given a positive test, that is \(P(D\mid +)\). These are different statements and often very different numbers.

The second mistake is using the numerator as the denominator. The denominator \(P(B)\) must include every way \(B\) can happen. In a medical-test problem, positive tests can happen among people with the disease and among people without the disease. In a defective-product problem, defective items can come from Machine X and Machine Y. Omitting one source usually makes the answer too large.

The third mistake is forgetting complements. If \(P(A)=0.3\), then \(P(A')=0.7\), not \(0.3\). If the problem gives the false positive rate, that is \(P(B\mid A')\), not \(P(A')\). Keep event probabilities and conditional probabilities separate.

The fourth mistake is mixing percentages and decimals. The formula works with probabilities, so \(5\%\) should be entered as \(0.05\), not \(5\), unless a calculator explicitly converts percentages. A numerator like \(95\times1\) instead of \(0.95\times0.01\) changes the scale of the result.

The fifth mistake is failing to interpret the result. A final answer of \(0.1935\) is mathematically correct, but a complete response says: "Given a positive test, the probability that the person has the disease is about \(19.4\%\)." Probability answers should always refer back to the event and the evidence.

Bayes' Theorem for GCSE, IGCSE, A-Level, IB and AP Students

Different courses introduce Bayes' Theorem with different levels of formality. At GCSE and IGCSE level, the focus is usually conditional probability, tree diagrams, two-way tables and interpreting "given that" correctly. Bayes' Theorem may appear in a simple form through reverse conditional probability, even when the theorem is not named.

At A-Level, IB and AP level, students are more likely to see the formula explicitly. Questions may involve disease screening, production lines, classification, repeated events or populations split into groups. The calculation is still manageable, but the wording becomes more precise. A strong solution defines the events, writes the desired conditional probability, calculates the total probability of the evidence and then applies the theorem.

For curriculum-linked probability revision, use the A-Level Probability page, the A-Level Statistics page, and the broader IB Mathematics resources. If you are reviewing probability alongside data analysis, the Statistics and Descriptive Statistics pages provide useful support.

Bayes' Theorem also links naturally to distributions and inference. For example, a probability model may give the likelihood of observing data under different assumptions. Bayesian statistics then updates beliefs about those assumptions. Students do not need advanced Bayesian inference to solve school-level examples, but understanding the basic theorem prepares you for later topics in statistics, data science and decision theory.

Base Rates and Why Bayes' Theorem Can Feel Surprising

Many Bayes' Theorem results feel surprising because people naturally focus on the evidence and ignore the base rate. The base rate is the prior probability before the evidence appears. In medical screening, the base rate is the prevalence of the disease. In quality control, it is the share of products made by each machine and the defect rate for each machine. In spam filtering, it is the prior chance that an email is spam before analyzing its words.

Suppose a disease affects only \(1\%\) of people, and a test is very good but not perfect. A positive result is important evidence, but most people do not have the disease. Even a small false positive rate applied to a large healthy group can create more false positives than true positives. That is why the posterior probability can be far lower than the test sensitivity.

Bayes' Theorem prevents this error by forcing the denominator to include the evidence under every relevant scenario. In the medical example, \(P(+)\) includes both true positives and false positives. In the machine example, \(P(D)\) includes defects from every machine. In the bag example, \(P(R)\) includes red balls drawn from every possible bag.

Whenever a Bayes' Theorem answer seems unexpectedly low or high, check the base rate. A strong piece of evidence can move a probability a long way, but it does not erase the prior. This is one reason the theorem is so useful: it balances old information and new information rather than allowing either one to dominate automatically.

Bayes' Theorem Compared With Other Probability Rules

Bayes' Theorem is part of a wider probability toolkit. It is closely related to the multiplication rule, conditional probability, complements, mutually exclusive events and the law of total probability. Understanding those connections makes the formula easier to remember.

The complement rule says:

The multiplication rule for conditional probability says:

The law of total probability says that if \(A_1,A_2,\ldots,A_n\) split the sample space into non-overlapping cases, then:

Bayes' Theorem combines these ideas. It uses the multiplication rule to write a joint probability, then divides by the total probability of the evidence. This is why Bayes' Theorem is often easier after you have studied conditional probability and tree diagrams. For additional review, the Probability Cheat Sheet Table of Contents is a useful entry point.

When Not to Use Bayes' Theorem

Bayes' Theorem is powerful, but it is not needed for every probability problem. If the question directly asks for \(P(A\cap B)\), use the multiplication rule. If the question asks for \(P(A\cup B)\), use an addition rule. If the events are independent and the question only asks for a straightforward product, Bayes' Theorem may be unnecessary.

Use Bayes' Theorem when the problem asks for a reverse conditional probability. A strong clue is wording such as "given that the result is positive, what is the probability the person has the disease?" or "given that the item is defective, what is the probability it came from Machine X?" The evidence is known, and you are asked to infer the likely source, group or hidden condition.

Another clue is when the question gives \(P(B\mid A)\) but asks for \(P(A\mid B)\). If the condition in the given probability and the condition in the required probability point in opposite directions, Bayes' Theorem is usually the right method.

Bayes' Theorem Practice Questions

Try each question before opening the answer. For best practice, define the events, calculate the denominator, and write a sentence interpreting your final probability.

Using Natural Frequencies to Understand Bayes' Theorem

Natural frequencies are one of the clearest ways to understand Bayes' Theorem. Instead of working only with percentages, choose a realistic sample size and count how many cases fall into each group. This method often makes the answer feel less surprising, especially in medical testing and screening examples.

Consider the medical-test example again. The disease affects \(1\%\) of people, the test detects \(95\%\) of true cases, and it gives a false positive for \(4\%\) of healthy people. The percentage calculation gave about \(19.4\%\). To see why, imagine testing \(10{,}000\) people.

Out of 491 positive tests, 95 are true positives. Therefore:

This is the same answer as the formula. The natural-frequency version shows the denominator as a real count: all positive tests, not just true positives. This approach is particularly useful when explaining the result to someone who expects a positive result to mean near-certainty. The test can be accurate and still produce a moderate posterior probability when the condition is rare.

Natural frequencies also help with exam checking. If your Bayes' Theorem answer says the probability is \(95\%\) in this example, the table immediately shows why that cannot be right. There are 95 true positives but 396 false positives, so true positives are not the majority of all positives. A simple count-based check can catch a reversed conditional probability before you submit the answer.

Bayes' Theorem With More Than Two Cases

Many school examples use \(A\) and \(A'\), but Bayes' Theorem can handle more than two possible causes. Suppose a product can come from three machines: \(M_1\), \(M_2\), and \(M_3\). If a defective item is found, you may want the probability that it came from \(M_2\). The denominator must include defects from all three machines.

The general version is:

The symbol \(\sum\) means "add these terms." If there are three machines, the denominator becomes:

For example, suppose \(M_1\) makes \(50\%\) of output with a \(2\%\) defect rate, \(M_2\) makes \(30\%\) with a \(5\%\) defect rate, and \(M_3\) makes \(20\%\) with a \(4\%\) defect rate. A product is defective. The probability it came from \(M_2\) is:

So the probability is about \(45.5\%\). This is not simply the largest machine share or the largest defect rate; it depends on the product of each machine's share and its defect rate. That product is the path probability for "made by this machine and defective."

When there are more than two cases, a table is often safer than a long formula. Make one row for each case, write its prior probability, write the likelihood of the evidence, multiply the two, and then divide the row you need by the total of the multiplied column. This is the same theorem, but the layout reduces mistakes.

How to Check Whether Your Bayes' Theorem Answer Is Reasonable

A Bayes' Theorem answer should always be checked for reasonableness. The first check is the probability range. A probability must be between 0 and 1, or between \(0\%\) and \(100\%\). If your answer is \(1.36\) or \(-0.12\), something is wrong with the setup, the denominator or the percentage conversion.

The second check is comparison with the prior. If the evidence supports \(A\), the posterior \(P(A\mid B)\) should usually be higher than the prior \(P(A)\). If the evidence argues against \(A\), the posterior should usually be lower. For example, a positive disease test should increase the probability of disease compared with the disease prevalence, but it does not necessarily make the probability close to \(1\).

The third check is comparison with competing cases. In the bag example, Bag B has twice the red-ball rate of Bag A. After a red ball is drawn, Bag B should become more likely than Bag A. If your answer says Bag A is more likely, the calculation probably reversed a conditional probability or omitted a branch from the denominator.

The fourth check is the denominator. The denominator should represent the probability of the evidence, not the probability of the hypothesis. In a positive-test problem, the denominator should be the probability of a positive test from every relevant group. In a defective-product problem, it should be the probability of a defective product from every production source.

The fifth check is unit-free consistency. Probabilities do not have units, but percentages and decimals must be consistent. If one part of the calculation uses \(95\) and another uses \(0.01\), the scale is mixed. Convert all percentages to decimals first, or use percentages consistently with a calculator that handles them correctly.

The final check is interpretation. A valid Bayes' Theorem answer should be readable as a sentence: "Given that \(B\) happened, the probability of \(A\) is ..." If you cannot write that sentence, you may not have defined the events clearly enough.

Bayes' Theorem and Independence

Bayes' Theorem is most interesting when the evidence changes the probability of the event. If \(A\) and \(B\) are independent, then knowing \(B\) does not change the probability of \(A\). In notation:

If events are independent, then \(P(B\mid A)=P(B)\). Substituting into Bayes' Theorem gives:

This shows that Bayes' Theorem agrees with independence. Evidence that is unrelated to an event should not update the probability of that event. If a coin is fair, knowing that it rained today does not change the probability of heads on the next flip. The evidence has no relevant connection to the event.

In many real contexts, however, evidence is not independent. A positive medical test is related to disease status. A word in an email is related to whether the email is spam. A defective product is related to which machine made it. A high exam score is related to study habits and course level. Bayes' Theorem is valuable exactly because it handles dependent evidence in a structured way.

Students sometimes use independence too quickly. Do not assume events are independent just because the question does not explicitly say they are dependent. If the question gives conditional probabilities such as \(P(B\mid A)\), that is usually a sign that the events are related. Check the wording carefully before applying independent-event shortcuts.

Bayes' Theorem in Data Science and Classification

Bayes' Theorem is a foundation for many classification problems. A classifier tries to decide which category an item belongs to after observing features. For example, an email classifier observes words, links, sender information and formatting. It then estimates whether the email is spam or not spam. In a simple Bayesian model, the category is the hidden event and the observed features are the evidence.

A naive Bayes classifier uses Bayes' Theorem with an additional simplifying assumption: the features are treated as conditionally independent given the class. This assumption is not always perfectly true, but it often works surprisingly well in text classification. If an email contains "prize," "winner" and "claim," the model combines how often those words appear in spam and non-spam emails to update the probability that the message is spam.

The simplified classification form is:

For school-level probability, you do not need to build a full classifier. Still, the connection is useful because it shows why Bayes' Theorem is not only an exam formula. It is a practical method for making decisions under uncertainty. The same structure appears in recommendation systems, document sorting, fraud detection and risk scoring.

The lesson for students is simple: Bayes' Theorem is about evidence. Evidence may be a test result, a red ball, a defective item, a word in an email, or a data pattern. Once you can identify the prior, likelihood and denominator, the same reasoning applies across many different contexts.

Writing a Full-Mark Bayes' Theorem Solution

A full-mark solution is more than a final percentage. It shows the reasoning clearly enough that another reader can follow the setup. Start by defining the events in words and symbols. For example, write \(D=\) disease and \(+=\) positive test. Then write the probability required by the question, such as \(P(D\mid +)\). This immediately shows that you understand the direction of the conditional probability.

Next, list the known values. Put the prior probabilities and conditional probabilities on separate lines. This prevents a common mix-up between \(P(D)\), \(P(+\mid D)\), and \(P(D\mid +)\). If percentages are given, convert them to decimals or keep them as fractions consistently.

Then calculate the denominator. In most Bayes' Theorem exam questions, this is where the marks are. The denominator is often the total probability of the evidence, and it usually requires the law of total probability. Show the addition of all relevant paths, not just the path in the numerator.

After that, substitute into Bayes' Theorem and simplify. Keep enough decimal places during calculation, then round the final answer appropriately. If the problem uses exact fractions, leave the answer as a fraction unless a decimal or percentage is requested.

Finally, interpret the answer in context. Instead of writing only \(0.625\), write "Given that the student scored above \(80\%\), the probability that the student is in the higher course is \(62.5\%\)." This final sentence confirms the event and the condition.

Mini Review: Notation You Must Know

Bayes' Theorem uses notation that can look compact at first. The following symbols appear often in probability questions and should be read fluently.

Good notation reduces mistakes. In particular, the vertical bar in \(P(A\mid B)\) is the most important symbol on this page. Everything after the bar is already known. Everything before the bar is the event whose probability is being updated.

Useful Probability and Statistics Resources

Bayes' Theorem sits inside a wider statistics and probability toolkit. The resources below are useful when you want to revise related ideas naturally alongside this topic.

Bayes' Theorem FAQ

What is Bayes' Theorem in simple words?

Bayes' Theorem is a formula for updating a probability after new evidence appears. It starts with a prior probability, uses the likelihood of the evidence, and produces a posterior probability. In simple words, it answers: "Now that I know this evidence, how likely is the original event?"

What is the formula for Bayes' Theorem?

The formula is \(P(A\mid B)=\frac{P(B\mid A)P(A)}{P(B)}\). If \(P(B)\) is not given, a common two-case form is \(P(B)=P(B\mid A)P(A)+P(B\mid A')P(A')\).

What is the difference between prior and posterior probability?

The prior probability is the probability before considering the new evidence. The posterior probability is the updated probability after the evidence is included. In \(P(A\mid B)\), the posterior is the probability of \(A\) after \(B\) is known.

Why is \(P(A\mid B)\) not the same as \(P(B\mid A)\)?

They condition on different information. \(P(A\mid B)\) means the probability of \(A\) after \(B\) is known. \(P(B\mid A)\) means the probability of \(B\) after \(A\) is known. A test can be very likely to be positive when a disease is present, while the probability of having the disease after a positive result can still be much lower if the disease is rare.

When should I use the law of total probability?

Use the law of total probability when the denominator \(P(B)\) is not given directly. Add all mutually exclusive ways that the evidence \(B\) can happen. In a two-case problem, this usually means adding the \(A\) path and the \(A'\) path.

Can Bayes' Theorem be solved with a tree diagram?

Yes. A tree diagram is often the clearest method. Multiply along branches to find final path probabilities, then add the paths that match the evidence. The final Bayes calculation is the relevant path divided by the total probability of the evidence.

Why do medical-test examples often give surprisingly low answers?

The answer can be low when the disease is rare. Even a test with high sensitivity can produce many false positives among the much larger healthy population. Bayes' Theorem accounts for both true positives and false positives.

Is Bayes' Theorem only for statistics?

No. It is used in statistics, but also in medicine, computing, engineering, economics, machine learning, search systems, quality control and decision-making. Any setting that updates uncertainty with evidence can use Bayesian reasoning.