Art of Problem Solving

2025 AMC 12A Problems/Problem 22

Revision as of 19:30, 7 November 2025 by Indiiiigo (talk | contribs) (Adding a 7th solution integrating over pdf of uniforms. Explicitly showing 3 choices of a,b,c instead of a WLOG argument.)

Problem 22

Three real numbers are chosen independently and uniformly at random between $0$ and $1$. What is the probability that the greatest of these three numbers is greater than $2$ times each of the other two numbers? (In other words, if the chosen numbers are $a \geq b \geq c$, then $a > 2b$.)

$\textbf{(A)}~\frac{1}{12}\qquad\textbf{(B)}~\frac19\qquad\textbf{(C)}~\frac18\qquad\textbf{(D)}~\frac16\qquad\textbf{(E)}~\frac14$

Problem 22 (Chinese exams)

Three real numbers are chosen independently and uniformly at random between $0$ and $1$. What is the probability that the greatest of these three numbers is greater than the sum of the other two numbers?

Solution 1

We can solve the problem by approaching it geometrically, where each possible outcome is a coordinate in a 1 by 1 by cube (a, b, c). Now we just have to find the area of the solution set over 1. Let's assume that a is the greatest number, and then multiply by 3 afterwards to account for b or c also possible being the largest. It can be seen that the possible values of a change linearly with b and c changing values, so to find the figure, we can just find the vertices and then connect them. By doing this, we can determine the solution set with a being the biggest value is the volume of the figure with coordinates (1,0,0), (0,0,0), (1,1/2,0), (1,0,1/2), which forms a tetrahedron which the volume can easily be calculated to be 1/12 with the formula V=1/3bh and the base is an isosceles right triangle with side length 1/2, and the height is just the height of the cube which is 1. Now just multiplying this value by 3 to account for b or c also being the maximum gives us the answer of $\boxed{\frac{1}{4},\textbf{E}}.$

~Zak2155

Solution 2 (Integration)

Consider the bounds of $a$, $b$, and $c$. We know that $a\geq b\geq c$ and $a\geq 2b$. For $a$ to be in $[0, 1]$, $b$ must be in $[0, 0.5]$. As such, each of $a$, $b$, and $c$ must be in the following intervals: \begin{align*} 1\geq &a\geq 2b, \\ 0.5\geq &b\geq 0, \\ b\geq &c\geq 0. \end{align*} Now, we can simply integrate over these bounds: \[\int_0^{1/2}\int_{2b}^1\int_0^b dc da db,\] \[\int_0^{1/2}\int_{2b}^1 b da db,\] \[\int_0^{1/2} b(1-2b) db,\] \[\left[\frac{b^2}{2}-\frac{2b^3}{3}\right]_0^{1/2},\] \[\frac{1}{24}.\] Since there are $6$ equally likely permutations of $a$, $b$, and $c$, we multiply $6\cdot\frac{1}{24}$ to obtain our answer of $\boxed{\frac{1}{4},\textbf{E}}.$

Note that this is almost the same solution as Solution 1, where we find the volume of a similar polyhedron using a different method (except we do not need to visualize it). This method would remain effective even if the bounds were non-linear or if there were more than three variables.

Solution 3 (Logic)

Say that, WLOG, $a$ is the maximum. Then notice that we need $b \le \frac{a}{2}$, given that $b \le a$, and similar for $c$. This probability is just $\frac{1}{2}$ for $b$ to be less, and $\frac{1}{2}$ for $c$ to be less (note that this is independent of the value of $a$). Therefore, the answer is $\frac{1}{4}$ (technically, the answer is $\int_{0}^{1} \frac{1}{4} da$ but this is constant).

~ScoutViolet

Solution 4

Define $A$ to be the largest of $a,b,c$. $B$ and $C$ are the two other numbers not ordered yet. We need satisfy P(A > 2B | A>B), and also B > C because of the ordering.

$P(A > 2B | A>B) = \frac{\frac14}{\frac12}=\bf{\frac12}$

Then, because b needs to be larger than c, the probability is 1/2. This is because both numbers are smaller than A, which order one larger than another is 1/2. Hence, the answer is 1/2 * 1/2 = $\boxed{\frac{1}{4},\textbf{E}}.$

(Feel free to correct the LaTex) ~Mitsuihisashi14 (Edited with corrections by ~Zhixing)

Solution 5 (Chinese exams)

Let the three numbers be $a$, $b$, and $c$. By symmetry, the probability that we want to find is exactly

\[3 \Pr(a>b+c)\]

Taking the cube $[0,1]^3$ in a Cartesian plane, we see that the region that corresponds to the event $a>b+c$ is the set of points

\[\{(x, \, y, \, z) \mid 0 \leq x, \, y, \, z \leq 1, x > y+z\}\]

Let $u=1-x$, $v=y$, and $w=z$, then the set above is equivalent to the set

\[\{(u, \, v, \, w) \mid 0 \leq u, \, v, \, w \leq 1, u + v + w \leq 1 \}\]

It is then obvious that the solid given by this set is a tetrahedron formed by cutting the first octant with the plane $u + v + w = 1$. The volume of it can be found by integration:

\[\int\!\!\int\!\!\int_{u, \, v, \, w \geq 0, \, u + v + w \leq 1} du\, dv\, dw =\frac{1}{6}\]

Hence, the answer is

\[3 \Pr(a>b+c) = 3 \cdot \frac{1}{6} = \frac{1}{2}\]

~Bloggish


Solution 6

Consider three independent uniform random variables $X, Y, Z$ on $[0, 1]$. The goal is to find the probability that the maximum value is greater than twice each of the other two values. Equivalently, if we order them as $a \geq b \geq c$, this is the probability that $a > 2b$ (which implies $a > 2c$ since $b \geq c$).

Due to symmetry, compute the probability that $X$ is the maximum and satisfies the condition, then multiply by 3. The condition simplifies to $P(X > 2Y \cap X > 2Z)$, as this ensures $X$ is the maximum (almost surely).

The joint density is 1 over the unit cube. Thus, \begin{align*} P(X > 2Y \cap X > 2Z) &= \int_0^1 \int_0^{x/2} \int_0^{x/2} \, dy \, dz \, dx \\ &= \int_0^1 \left(\frac{x}{2}\right)^2 \, dx \\ &= \int_0^1 \frac{x^2}{4} \, dx \\ &= \frac{1}{4} \cdot \frac{1}{3} \\ &= \frac{1}{12}. \end{align*} The total probability is $3 \times \frac{1}{12} = \frac{1}{4}$.

Alternatively, using order statistics $U < V < W$ with joint density 6 on $0 < u < v < w < 1$, \begin{align*} P(W > 2V) &= \int_0^{0.5} \int_{2v}^1 \int_0^v 6 \, du \, dw \, dv \\ &= 6 \int_0^{0.5} v (1 - 2v) \, dv \\ &= 6 \left[ \frac{v^2}{2} - \frac{2v^3}{3} \right]_0^{0.5} \\ &= 6 \times \dfrac{1}{24} \\ &= \dfrac{1}{4}. \end{align*} No matter what entry points on this problem we choose, we can always get the correct answer: $\color{red}\boxed{\color{black}\dfrac{1}{4}}\color{black}.$

~funkCCP

Solution 7

We have \(a,b,c \sim U(0,1)\). Thus, if \(a\) is the greatest, we must have \(b,c < a/2\), which occurs with probability \(a^2/4\). We integrate this over all \(a\) (note the pdf of \(U(0,1)\) is just \(1\)):

\[\int_0^1 \frac{a^2}{4} \, da=\left[\frac{a^3}{12}\right]_0^1=\frac{1}{12}\]

Since there are \(3\) possibilities for the greatest among \(a,b\), and \(c\), we multiply this by \(3\) to get \(3\cdot 1/12=\boxed{\text{(E) } \frac14}\)

Video Solution 1 by OmegaLearn

https://youtu.be/XxVC6Hx2tIY


See Also

2025 AMC 12A (ProblemsAnswer KeyResources)
Preceded by
Problem 21
Followed by
Problem 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
All AMC 12 Problems and Solutions

These problems are copyrighted © by the Mathematical Association of America.