Problem 27
Hard Properties of Probability

Given p=P(A) , q=P(B) , r=P(A+B) . Find P(Atriangle B)…

Given $p=P(A)$, $q=P(B)$, $r=P(A+B)$. Find $P(A\triangle B)$, $P(A\overline{B})$ and $P(\overline{AB})$.

Explanation of Method

We first find the probability of the common part $AB$ from the addition rule $P(A+B)=P(A)+P(B)-P(AB)$. Then we split $A$, $B$ and $\Omega$ into non-overlapping pieces; for non-overlapping (incompatible) events probabilities simply add. Each requested probability is a sum or difference of the piece probabilities. Finally we check the formulas on numbers.

Step 1: The probability of $AB$

The addition rule for two events is

\[ P(A+B) = P(A) + P(B) - P(AB). \]

(Adding $P(A)$ and $P(B)$ counts the common part $AB$ twice, so it is subtracted once.) With our notation $r = p + q - P(AB)$, hence

\[ P(AB) = p + q - r. \]

Step 2: $P(A\overline{B})$ — "$A$ occurs, $B$ does not"

$A$ splits into two incompatible pieces: $A = AB + A\overline{B}$. So $P(A) = P(AB) + P(A\overline{B})$ and

\[ P(A\overline{B}) = P(A) - P(AB) = p - (p+q-r) = r - q. \]

\[ \boxed{P(A\overline{B}) = r - q.} \]

Another way to see it: $A+B = A\overline{B} + B$ (incompatible), so $r = P(A\overline{B}) + q$.

Step 3: $P(A\triangle B)$ — "exactly one occurs"

$A\triangle B = A\overline{B}+\overline{A}B$ with incompatible terms. By symmetry with Step 2, $P(\overline{A}B) = r - p$. So

\[ P(A\triangle B) = (r-q) + (r-p) = 2r - p - q. \]

Check another way: "at least one" = "exactly one" + "both", so $P(A\triangle B) = P(A+B) - P(AB) = r - (p+q-r) = 2r-p-q$. ✓

\[ \boxed{P(A\triangle B) = 2r - p - q.} \]

Step 4: $P(\overline{AB})$ — "not both occur"

By the rule for the opposite event, $P(\overline{X}) = 1 - P(X)$:

\[ P(\overline{AB}) = 1 - P(AB) = 1 - p - q + r. \]

\[ \boxed{P(\overline{AB}) = 1 - p - q + r.} \]

$\Omega$ $A$ $B$ $r-q$ $p+q-r$ $r-p$ $1-r$
  • the number in each piece is its probability
Probabilities of the four pieces: $P(A\overline{B})=r-q$, $P(AB)=p+q-r$, $P(\overline{A}B)=r-p$, $P(\overline{A}\,\overline{B})=1-r$. They add up to $1$.

Step 5: Numerical example

Let $p = 0.5$, $q = 0.4$, $r = 0.7$. Then

QuantityFormulaValue
$P(AB)$$p+q-r$$0.2$
$P(A\overline{B})$$r-q$$0.3$
$P(\overline{A}B)$$r-p$$0.2$
$P(A\triangle B)$$2r-p-q$$0.5$
$P(\overline{AB})$$1-p-q+r$$0.8$
$P(\overline{A}\,\overline{B})$$1-r$$0.3$

Check: the four pieces add up to $0.3+0.2+0.2+0.3=1$ ✓, and $P(A\overline{B})+P(AB)=0.3+0.2=0.5=p$ ✓.

Step 6: Which values of $p,q,r$ are possible?

All piece probabilities must be between $0$ and $1$. This gives

\[ \max(p,q) \le r \le \min(1,\ p+q). \]

For example $p=0.5$, $q=0.4$, $r=0.3$ is impossible: $A+B$ cannot be less likely than $A$ alone.


Result

ProbabilityAnswer
$P(AB)$ (helper)$p+q-r$
$P(A\triangle B)$$2r-p-q$
$P(A\overline{B})$$r-q$
$P(\overline{AB})$$1-p-q+r$