In the previous section we showed that the probability of both $A$ and $B$ occurring is equal to the product of both individual probabilities:
$$P(A) \ \cdot \ P(B) \ = \ P(A \ \text{and} \ B)$$
However, this is not entirely true. Multiplying individual probabilities is only possible if the events are independent, i.e the probability that $B$ occurs is the same regardless of whether $A$ occurred or not. Flipping a coin twice and getting heads on both flips is an example of two independent events. The outcome of the first flip has no effect on the likelihood of getting heads one the second flip.
But sometimes events are dependent, i.e the probability that $B$ occurs changes depending on whether $A$ occurred or not. For example, if the weather is foggy there is a higher likelihood of ending up in a car accident compared to if the weather is not foggy. When events are dependent we can’t take the product of the individual probabilities $P(A)$ and $ P(B)$ as they are. We must first adjust $P(B)$ depending on whether $A$ occurred or not.
This is what we did in Exercise 3 on the previous page. The probability of choosing a green marble was $\frac{6}{10}$. But after the first green marble had been chosen the likelihood of choosing another green marble was $\frac{5}{9}$. The probability of picking two green marble was therefore not $\frac{6}{10} \cdot \frac{6}{10}$ but $\frac{6}{10} \cdot \frac{5}{9}$. When we have adjusted the probability of $B$ depending on whether $A$ occurred, we can multiply the individual probabilities to determine the likelihood of both events occurring.
The probability that $B$ occurs given that $\pmb{A}$ has already occurred is written as $P(B \ | \ A)$.
$$P(B \ | \ A) \ = \ P(B \ \text{given that} \ A \ \text{has occurred})$$
The probability of picking two green marbles in Exercise 3 could therefore have been written as:
$$P(\text{(green}) \ \cdot \ P(\text{(green | green}) \ = $$
$$\dfrac{6}{10} \ \cdot \ \frac{5}{9} = \frac{1}{3}$$
The general probability multiplication rule is written as:
$$\pmb{P(A) \ \cdot \ P(B \ | \ A) \ = \ P(A \ \text{and} \ B)}$$
or
$$\pmb{P(B) \ \cdot \ P(A \ | \ B) \ = \ P(A \ \text{and} \ B)}$$
We can also rearrange this equation to solve for the individual probabilities:
$$P(A) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(B \ | \ A)} \ \ \ \ \ \ \ P(B) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(A \ | \ B)}$$
The general multiplication rule always holds true. If $A$ and $B$ are independent then $P(A) = P(A \ | \ B)$ and $P(B) = P(B \ | \ A)$, so the equations can be simplified to:
$$P(A) \ \cdot \ P(B) \ = \ P(A \ and \ B)$$
$$P(A) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(B)} \ \ \ \ \ \ \ P(B) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(A)}$$
For independent events we can therefore take the product of the individual probabilities without having to adjust them. If the events are dependent, however, the probability that $B$ occurs is now adjusted thanks to the expression $P(B \ | \ A)$.
Which of the following are examples of independent events?
a) That it's raining and that it's cloudy.
b) To roll a die and get an even number three times in a row.
c) To make a left and a right turn at the same time.
d) To rob a bank and go to jail.
e) Getting stuck in traffic on the way home and winning the lottery.
If $P(A) = P(A \ | \ B)$, is it true that $P(B) = P(B \ | \ A)$?
The general multiplication rule states that:
$$P(A) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(B \ | \ A)} \ \ \ \ \ \ \ P(B) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(A \ | \ B)}$$
If $P(A) = P(A \ | \ B)$, the right version of the rule above can be simplified to:
$$P(B) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(A)}$$
Which can be rearranged to:
$$P(A) \ \cdot \ P(B) \ = \ P(A \ \text{and} \ B)$$
$$P(A) \ = \ \dfrac{P(A \ \text{and} \ B)}{P(B)}$$
If we compare this equation with the left version of the general multiplication rule above we see that $P(B) = P(B \ | \ A)$. It is therefore true that:
$$P(A) = P(A \ | \ B) \ \ \iff \ \ P(B) = P(B \ | \ A)$$