---
title: "Basic probability rules"
canonical: "https://modelassist.epixanalytics.com/space/EA/26574944/Basic%20probability%20rules"
format: markdown
---
(review probability notation [here](https://epixanalytics.atlassian.net/wiki/spaces/EA/pages/26574943/))

##### **1. Complementation rule**

For an event A and its complement A<sup>c</sup> :

![image](media://6a8735cd-8c97-4071-9d8f-8429f1ce4987)


> Macro (mathblock)

### ***Example:***

A = throwing a double six with two dice

A<sup>c</sup> = not throwing a double six

P(A) = 1/36

P(A<sup>C</sup>) = 35/36

##### **2. Addition rule**

For mutually exclusive events A<sub>i</sub>:

![image](media://89b4f9ee-8110-4d15-94a5-0d1d165acca6)

  


> Macro (mathblock)

### ***Example:***

A<sub>1</sub> = the next patient in a hospital is suffering from breast cancer

A<sub>2</sub> = the next patient in a hospital is suffering from prostate cancer

P(A<sub>1</sub>) = 3%

P(A<sub>2</sub>)  = 2.5%

P(A<sub>1</sub>EA<sub>2</sub>) = 5.5%

##### **3. Multiplication rule**

![image](media://d4cfecff-db9c-4e7a-b6ce-ab20dddd77b8)


> Macro (mathblock)

### ***Example:***

A = a person walking by a shop enters the shop

B|A = person in shop purchase something

C|A ∩ B = person buying spends >$100

P(A) = 10%

P(B|A) = 40%

P(C|A ∩ B) =  25%

P(A ∩ B ∩ C*) =* 10%*40%*25% = 1%

P(A|B) = 100%

  


Rearranging the formula leads to [Bayes formula](https://epixanalytics.atlassian.net/wiki/spaces/EA/pages/26574955/);

![image](media://ae8c7434-8363-4614-b9ad-7b786f94666e)

  


> Macro (mathblock)

  


This formula is of fundamental importance in statistics and is the cornerstone of [Bayesian inference](https://epixanalytics.atlassian.net/wiki/spaces/EA/pages/26575368/).

(More on [conditional probabilities](https://epixanalytics.atlassian.net/wiki/spaces/EA/pages/26574945/))

##### ***Independence***

If A and B are such that

![image](media://4131437b-63de-471c-9eab-5dd31a11991d)


> Macro (mathblock)

  


they are called *independent* events.

### **Example:**

A = theme park has a bad summer in Miami due to weather

B = theme park has a bad summer in Brussels due to weather

  


If there is no link between weather in Miami and Brussels then these events are independent. It would be important to know if there is any correlation if you owned both theme parks, because you could have a higher (positive correlation) or lower (negative correlation) probability of a bad summer in both at the same time with a greater resultant cashflow uncertainty.

  


  


  


---