ARML 2021 Contest Packet PDF

ARML Competition 2021
George Reuter, Head Writer

Chris Jeuell, Lead Editor
Evan Chen
Paul Dreyer
Edward Early
Zuming Feng
Zachary Franco
Silas Johnson
Winston Luo
Jason Mutford
Andy Niedermaier
Graham Rosby
June 5, 2021
Sponsored By:
ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
1 Team Problems
Problem 1. Let N be the 20-digit number 20202020202020202020. Compute the sum of the digits of N 2 .
Problem 2. In a regular 20-gon, three distinct vertices are chosen at random. Compute the probability that the
triangle formed by these three vertices is a right triangle.
Problem 3. Starting at (0, 0), a frog moves in the coordinate plane via a sequence of hops. Each hop is either 1 unit
in the x-direction or 1 unit in the y-direction. Compute the minimum number of hops needed for the frog to
land on the line 15x + 35y = 2020.
x
Problem 4. Compute the number of real values of x such that cos(cos(x)) = .
10
Problem 5. A circle passes through both trisection points of side AB of square ABCD and intersects BC at
points P and Q. Compute the greatest possible value of tan ∠P AQ.
Problem 6. Compute the least integer n > 2020 such that (n + 2020)n−2020 divides nn .
Problem 7. The latus rectum of a parabola is the line segment parallel to the directrix, with endpoints on the
parabola, that passes through the focus. Define the latus nextum of a parabola to be the line segment parallel to
the directrix, with endpoints on the parabola, that is twice as long as the latus rectum. Compute the greatest
possible distance between a point on the latus rectum and a point on the latus nextum of the parabola whose
equation is y = x2 + 20x + 20.
Problem 8. Circle Ω1 with radius 11 and circle Ω2 with radius 5 are externally tangent. Circle Γ is internally
tangent to both Ω1 and Ω2 , and the centers of all three circles are collinear. Line ` is tangent to Ω1 and Ω2
at distinct points D and E, respectively. Point F lies on Γ so that F D < F E and m∠DF E = 90◦ . Compute
sin ∠DEF .
Problem 9. Nine distinct circles are drawn in the plane so that at most half of the pairs of these circles intersect.
The circles divide the plane into N regions of finite area. Compute the maximum possible value of N .
Problem 10. For a positive integer k, let S(k) denote the sum of the digits of k. Compute the number of
seven-digit positive integers n such that S(n) + 4S(2n) = S(36n).
ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 1
2 Answers to Team Problems
Answer 1. 112
3
Answer 2.
19
Answer 3. 60
Answer 4. 3
7
Answer 5.
11
Answer 6. 2076
√
3 5
Answer 7.
4
√
5
Answer 8.
4
Answer 9. 39
Answer 10. 610
2 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
3 Solutions to Team Problems
Problem 1. Let N be the 20-digit number 20202020202020202020. Compute the sum of the digits of N 2 .
Solution 1. Let M = 1010101010101010101. Note that N = 20M and that
M 2 = 1020304050607080910090807060504030201.
Squaring N ,
N 2 = 400M 2 = 408121620242832364036322824201612080400.
The sum of the digits of N 2 is thus 4 more than twice the sum of all the digits of the positive multiples of 4
up to 36, which is
2(4 + 8 + 1 + 2 + 1 + 6 + 2 + 0 + 2 + 4 + 2 + 8 + 3 + 2 + 3 + 6) + 4 = 112.
Problem 2. In a regular 20-gon, three distinct vertices are chosen at random. Compute the probability that the
triangle formed by these three vertices is a right triangle.
Solution 2. There are 20

3 = 1140 triangles that can be formed by choosing three distinct vertices of the 20-gon.
Inscribe the 20-gon in a circle. By the Inscribed Angle Theorem, a right triangle is only formed in the case
in which two of the vertices chosen (at random) are diametrically opposite each other. There are 10 choices
for a diameter; after the diameter is chosen, the third vertex can be chosen in 18 ways. Thus the answer is
18 · 10 3
= .
1140 19
Problem 3. Starting at (0, 0), a frog moves in the coordinate plane via a sequence of hops. Each hop is either 1 unit
in the x-direction or 1 unit in the y-direction. Compute the minimum number of hops needed for the frog to
land on the line 15x + 35y = 2020.
Solution 3. The frog can only land on lattice points. Let (a, b) denote the point on the line 15x + 35y = 2020 on
which the frog lands. The frog lands on the line only if 15a + 35b = 2020, and the number of hops required
is |a| + |b|. Note that 15a + 35b = 2020 implies 3a + 7b = 404. This line passes through the points (134 32 , 0)
and (0, 57 75 ).
Consider lines of the form x + y = k. Because x + y = k has a slope of −1 and 15x + 35y = 2020 has a
slope of − 15 3
35 = − 7 , the lattice point that produces the minimum number of hops must be as close as possible
to the y-intercept (0, 57 57 ). Notice that if x = 1 or x = 2 or x = 3, then there is no integer solution to
15x + 35y = 2020 with respect to y. If x = 4, then 15 · 4 + 35y = 2020 → y = 56, which is an integer. Thus the
minimum number of hops needed for the frog to land on the line 15x + 35y = 2020 is 4 + 56 = 60.
x
Problem 4. Compute the number of real values of x such that cos(cos(x)) = .
10
x
Solution 4. Let f (x) = cos(cos(x)) and g(x) = 10 . The function cos(x) is even and ranges from −1 to 1, and so
f (x) ranges between cos(1) and cos(0) = 1 with period π and maxima and minima occurring at kπ 2 for odd
and even integers k, respectively. The graphs of f (x) and g(x) are shown below.
The ranges of f (x) and g(x) are identical for x ∈ [10 cos(1), 10], with g(x) increasing over the entirety of this
interval. Note that
π π π π π π
cos(1) = cos − − 1 = cos cos − 1 + sin sin −1
3 3 √3 3 3 3
1 3 π 1 1 3 π
< + −1 < + < < ,
2 2 3 2 21 5 5
√
with the latter inequalities using the facts that cos(x) ≤ 1 and sin(x) < x for all positive x, 23 < 1, and
3 < π < 22 π
7 . Thus it follows that g(10 cos(1)) = cos(1) < f (10 cos(1)) and f (2π) = cos(1) < 5 = g(2π). Be-
cause f (x) is strictly decreasing between consecutive maxima and minima, there is a single intersection point
of the two curves in the interval (10 cos(1), 2π). By a similar argument, it follows that there is an intersection
point of the two curves in each of the intervals (2π, 5π 5π
2 ) and ( 2 , 3π).
There is no intersection point of the curves in the interval (3π, 10) because
Å π ã Å π ã
f (x) < f 3π + Arccos = cos cos 3π + Arccos
6 6
Å π π ã
= cos cos(3π) cos Arccos − sin(3π) sin Arccos
6 6
π √
3
= cos = < g(3π)
6 2
for all x in the interval, as 10 < 3π + Arccos( π6 ) < 7π

2 .
Thus there are a total of 3 intersection points.
Problem 5. A circle passes through both trisection points of side AB of square ABCD and intersects BC at
points P and Q. Compute the greatest possible value of tan ∠P AQ.
Solution 5. Suppose, without loss of generality, that ABCD has sides of length 1; denote the trisection points of
AB by X and Y so that AX = XY = Y B = 13 . Let P be nearer to B and let Q be nearer to C.
A X Y B
• • • •
•P
• •
D C/Q
By Power of a Point from B, BP · BQ = BY · BX = 31 · 23 = 29 ; thus if BQ = x, then BP = 9x2

. Because BC
intersects the circle in points P and Q and P is between B and Q, it follows that BP < BQ ≤ BC = 1, so
2
9x < x ≤ 1. Note that maximizing tan ∠P AQ is equivalent to maximizing P Q. Geometrically, this value is
2
maximized when Q coincides with C, so x = 1. This also agrees algebraically: to maximize P Q = x − 9x ,x
2
should be as large as possible and 9x should be as small as possible. Given the bounds on x, the expression
2
x − 9x achieves its maximum when x = 1. This implies
2
x − 9x 7
tan ∠P AQ = tan(∠QAB − ∠P AB) = 2 = 11 .
1+ 9
Problem 6. Compute the least integer n > 2020 such that (n + 2020)n−2020 divides nn .
Solution 6. Let m = n + 2020 > 4040. Suppose a prime p divides m. Then p also divides n, and therefore
p | 2020 = 22 · 5 · 101. In other words, m is an integer greater than 4040 with no prime factors other than 2, 5,
and 101. Note the least candidate for m is 4096 = 212 ; indeed, the next multiple of 101 after 4040 is 4141, and
there is no number of the form 2a 5b between 4040 and 4096. So the least candidate for n is 2076. This works,
because (n + 2020)n−2020 = 409656 = 212·56 certainly divides 20762076 . Thus the answer is 2076.
Problem 7. The latus rectum of a parabola is the line segment parallel to the directrix, with endpoints on the
parabola, that passes through the focus. Define the latus nextum of a parabola to be the line segment parallel to
the directrix, with endpoints on the parabola, that is twice as long as the latus rectum. Compute the greatest
possible distance between a point on the latus rectum and a point on the latus nextum of the parabola whose
equation is y = x2 + 20x + 20.
Solution 7. Denote by h the distance between the latus rectum and the latus nextum. Let f denote the focal
2
length (so the latus rectum has length 4f ). The diagram below reveals the critical relation (h + 2f ) =
2 2
h + (4f ) → h = 3f .
Calculate either the length ER or P G shown in the diagram to find that the greatest possible distance is
»
2 2
√
(3f ) + (6f ) = 3 5f.
1
Recall that for a general parabola y = ax2 + bx + c, the focal length is f = . In particular, for this parabola,
√ 4a
1 3 5
the focal length is f = , and thus the final answer is .
4 4
Alternate Solution: Because the parabola y = x2 + 20x + 20 is congruent to the parabola y = x2 , con-
sider instead the parabola y = x2 . The latus rectum is along y = 14 , and its endpoints are ± 12 , 14 . Thus the
» 2 2 » √
3 2 3 2
1 − (− 12 ) + 1 − 14 = 3 5

latus nextum has endpoints at (±1, 1). The answer is thus 2 + 4 = 4 .
Problem 8. Circle Ω1 with radius 11 and circle Ω2 with radius 5 are externally tangent. Circle Γ is internally
tangent to both Ω1 and Ω2 , and the centers of all three circles are collinear. Line ` is tangent to Ω1 and Ω2
at distinct points D and E, respectively. Point F lies on Γ so that F D < F E and m∠DF E = 90◦ . Compute
sin ∠DEF .
Solution 8. Let C be the tangency point of Ω1 and Ω2 , and construct diameters QC and CR of Ω1 and Ω2 , as
shown. Let F1 be the point on Γ such that CF1 ⊥ QR. Let segment QF1 intersect Ω1 at D1 , and let segment
RF1 intersect Ω2 at E1 . It is not difficult to see that the following are similar right triangles: CQD1 , F1 CD1 ,
F1 QC, RCE1 , CF1 E1 , RF1 C, and RQF1 .
F1
D1
E1
Q R
C
Hence CD1 F1 E1 is a rectangle and
m∠CQD1 = m∠F1 QC = m∠RF1 C = m∠CF1 E1 = m∠CD1 E1 .
Thus ∠CQD1 and ∠CD1 E1 both intercept minor arc CD1 in circle Ω1 , implying that line D1 E1 is tangent
←−−→
to Ω1 at D1 . Likewise, line D1 E1 is tangent to Ω2 at E1 , hence ` = D1 E1 . Consider the semicircles with
←→
diameters QC and CR that lie on the same side of QR as point F1 . Because these semicircles have a unique
common exterior tangent line, it follows that D = D1 , E = E1 and F = F1 . Then CQ = 22, CR √= 10, and
√ √ p √ 5
CF = CQ · CR = 2 55. Finally, QF = QC 2 + F C 2 = 8 11, and sin ∠DEF = sin ∠CQF = .
4
Alternate Solution: Denote by A, B, and O the centers of Ω1 , Ω2 , and Γ, respectively. Let C be the tangency
point of Ω1 and Ω2 , and construct diameters QC and CR of Ω1 and Ω2 , as shown.
P
E
Q θ
R
A O C B
Observe there is a positive homothety taking Ω1 to Ω2 mapping D to E, centered along the intersection of ` and
line AB. Thus 4QDC ∼ 4CER and moreover, the corresponding sides are parallel. In particular, because
m∠QDC = m∠CER = 90◦ , the lines QD and ER are perpendicular. Because QR is a diameter of Γ, it follows
←→ ←→
that QD and ER meet at a right angle on Γ at some point G with m∠DGE = 90◦ . Because m∠DGE = 90◦ ,
it follows that G lies on the circle with diameter DE. This circle intersects Γ twice, and the reader can confirm
that G is the intersection point closer to D, and thus G = F . Recalling also that DC ⊥ CE by the same pair
of similar triangles, it follows that CDF E is a rectangle. Next, notice that
1
m∠DEF = m∠CDE = m∠CQD = m∠DAC.
2
PA
Finally, construct rectangle DEBP so that P lies on AD. Then cos ∠DAC = cos ∠P AB = AB , and this yields
1 1
sin ∠DEF = sin ∠DAC = sin ∠P AB
2 2
√
PA 1 − 11−5
…
1 − cos ∠P AB 1 − AB 11+5 5
= = = = .
2 2 2 4
Problem 9. Nine distinct circles are drawn in the plane so that at most half of the pairs of these circles intersect.
The circles divide the plane into N regions of finite area. Compute the maximum possible value of N .
Solution 9. First, notice that if any circles are concurrent or tangent, then one may slightly increase the radius of
any such circle while increasing the number of regions formed, without violating any conditions. Thus assume
henceforth that no three circles are concurrent and no two circles are tangent. Let k ≤ 21 92 = 18 denote
the number of pairs of circles that intersect in two points. The idea is to use the following version of Euler’s
Formula on the graph G whose vertices are intersection points and whose edges are arcs, and c is the number
of connected components of G:
V − E + F = 2 + (c − 1).
For convenience, if there are any circles with no intersections at all, then consider this as a connected component
with 0 vertices and 0 edges (in which case the formula still holds). Notice that V = 2k, because every pair
of intersecting circles yields two vertices. Moreover, because every vertex in G has degree 4, it follows that
E = 2V = 4k. Thus
F = E − V + 2 + (c − 1) = 2k + 2 + (c − 1).
Finally, only one region has infinite area (the unbounded face), thus
N = F − 1 = 2k + 1 + (c − 1).
Now consider the connected components of the graph G, whichcorrespond to groups of circles. If there are at
least four groups, then the maximum possible value of k is 62 + 12 + 12 + 12 = 15. So for 16 ≤ k ≤ 18,

it follows that c ≤ 3, and thus N ≤ 39. For k ≤ 15, it is enough to use c ≤ 9 to obtain N ≤ 39 as well.
The following example shows that N = 39 indeed works because it satisfies c = 3 and k = 18. (Note that
k = 72 − 3 = 18 because the three red circles do not intersect.) Hence the final answer is N = 39.

Problem 10. For a positive integer k, let S(k) denote the sum of the digits of k. Compute the number of
seven-digit positive integers n such that S(n) + 4S(2n) = S(36n).
Solution 10. The given equation can be rewritten as

S(10n) + S(20n) + 3S(2n) = S(36n).
The solution relies on the following lemma: for any integers a and b,
S(a + b) ≤ S(a) + S(b),
with equality if and only if there are no carries in the addition of a and b. This follows from the general fact
that S(a + b) = S(a) + S(b) − 9c, where c is the number of carries in the addition of a and b. Notice that
10n + 20n + 2n + 2n + 2n = 36n.
Using the lemma, the problem is equivalent to finding the number of integers n such that the five-term addition
10n + 20n + 2n + 2n + 2n has no carries. Let 2n = d1 d2 . . . dk and let 10n = e0 e1 . . . ek−1 0 , where the di ’s and
ei ’s are digits. Note that d1 could possibly be zero. The addition table for the desired sum is shown below.
2n d1 d2 ... dk−2 dk−1 dk
2n d1 d2 ... dk−2 dk−1 dk
2n d1 d2 ... dk−2 dk−1 dk
10n e0 e1 e2 ... ek−2 ek−1 0
20n d1 d2 d3 ... dk−1 dk 0
It immediately follows that di ∈ {0, 1, 2} for each i and this is assumed henceforth. Then for i = 0, 1, . . . , k − 1,
note that 

 0 if di 6= 1 and di+1 6= 2

1 if d 6= 1 and d
i i+1 = 2
ei =


 5 if di = 1 and di+1 6= 2

6 if di = 1 and di+1 = 2.
It follows that a choice of di ∈ {0, 1, 2} for each i is valid if and only if whenever di = 1, di+1 6= 2. In other
words, the set of valid choices of 2n consists of even integers with digits only 0, 1, 2, such that the string 12 never
appears. Letting Fi denote the ith term of the Fibonacci sequence, the following results can be established by
induction on d ≥ 1:
• the number of such d-digit strings starting with 0 (allowing leading zeros) is F2d−1 ;
• the number of such d-digit strings starting with 1 is F2d−2 ; and
• the number of such d-digit strings starting with 2 is F2d−1 .
Finally, if n is an r-digit number, then 2n must be either an r-digit number beginning with 2, or an (r + 1)-digit
number beginning with 1. There are F2r−1 + F2r = F2r+1 such numbers that satisfy the desired constraints,
and with r = 7, the answer is F15 = 610.
4 Power Question 2021: Complex Triangles
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part. To
receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “compute,”
you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show your work
or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a problem
says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered items
may be used in solutions to later numbered items, but not vice versa.
After the round is over, your team will tell the proctor their answers for the “list” or “compute” questions, and
the proctor will enter them directly into the answer sheet. For the other questions, write the following information
at the top right of each page that you are submitting for credit: Team Number, Problem Number (or part), Page k
of n, where n is the total number of pages in your submission for that problem only. Do not identify the team in any
other way.
For each distinct problem solution you are submitting, you should create a separate PDF file using Adobe Scan.
If you are solving several parts of the same question they should appear in the same file. The file should have the
name Team XXX Problem Y , where XXX is your team number, and Y is the problem (or problem part) be-
ing solved in the document. The link to the PDF file should be emailed to your proctor and to armlpower@arml.com.
Detailed instructions for submitting Power Question solutions are available here.
A complex triangle (hereafter called a c-triangle) is an unordered list of three complex numbers, [a, b, c], with abc 6= 0.
The values a, b, and c are the side lengths of the c-triangle. Two c-triangles are considered the same (or congruent) if
and only if their side lengths are the same, regardless of order. For example, [1, 1, 3] and [1, 3, 1] represent congruent
c-triangles. A complex angle (hereafter called a c-angle) is an ordered triple (a, b, c) whose elements are complex
numbers, again with abc 6= 0. The c-cosine of a c-angle is defined as
b2 + c2 − a2
ccos(a, b, c) = .
2bc
Two c-angles are equal if and only if they have the same c-cosine. In particular, note that the c-angles (a, b, c)
and (a, c, b) are equal.
Every c-triangle [a, b, c] has three c-angles, which are denoted by ∠A = (a, b, c), ∠B = (b, a, c), and ∠C = (c, a, b).
A c-angle ∠P is called proper if |ccos(∠P p )| ≤ 1, and improper otherwise. (Note: The absolute value of a complex
number x + yi, where x and y are real, is x2 + y 2 .)
The squarea (“squared area”) of a c-triangle is defined by the formula

a+b+c
sqar[a, b, c] = s(s − a)(s − b)(s − c), where s = .
2
1. a. For the c-triangle [2, 3, 4], compute the c-cosines of its three c-angles. [3 pts]
b. Suppose a c-triangle [i, 1, c] exists with ccos(i, 1, c) = 1 and ccos(1, i, c) = 1. Compute c. [2 pts]
2. For each of the following properties, list an example of a c-triangle [a, b, c] satisfying that property.
a. ccos(∠A) = 0 [1 pt]
b. ccos(∠A) = 2 [1 pt]
c. ccos(∠A) = i [1 pt]
d. sqar[a, b, c] = 36 [1 pt]
e. sqar[a, b, c] = −36 [1 pt]
f. sqar[a, b, c] = − 43 [1 pt]
3. Find the set of all complex numbers z for which (1, 1, z) is a proper c-angle. [2 pts]
4. List two noncongruent c-triangles [a1 , b1 , c1 ] and [a2 , b2 , c2 ] such that a1 = a2 , b1 = b2 , and ∠C1 = ∠C2 . (Doing
this will show that “SAS congruence” does not hold for c-triangles.) [3 pts]
5. Find an example of each of the following.
a. A c-triangle that has two improper c-angles and one proper c-angle. [2 pts]
b. A c-triangle that has two proper c-angles and one improper c-angle. [2 pts]
6. Show that if every c-angle of the c-triangle [a, b, c] is proper, then |b| + |c| ≥ |a|. [3 pts]
7. a. Show that sqar[a, b, c] = 41 b2 c2 (1 − (ccos(∠A))2 ). [4 pts]
b. Show that for any c-triangle [a, b, c],
1 − (ccos(∠A))2 1 − (ccos(∠B))2
2
= . [2 pts]
a b2
8. Show that if two c-triangles [a1 , b1 , c1 ] and [a2 , b2 , c2 ] have corresponding c-angles equal (that is, ∠A1 = ∠A2 ,
a2 b2
∠B1 = ∠B2 , and ∠C1 = ∠C2 ), and none of those c-angles has c-cosine equal to ±1, then = . (That is,
a1 b1
one could call these c-triangles “similar”.) [4 pts]
A c-configuration consists of a finite set S, whose elements will be called c-points and denoted by capital letters.
In addition, for each pair of distinct c-points in a c-configuration, a nonzero complex number called the distance is
assigned to the pair. The distance between the unordered c-points A, B ∈ S will be denoted by either d(A, B) or
d(B, A), interchangeably.
If A, B, and C are c-points, then the c-triangle [d(A, B), d(A, C), d(B, C)] will be denoted by 4ABC, and the
c-angle d(A, C), d(A, B), d(B, C) by ∠ABC. Note that A, B, and C denote c-points, not c-angles.
9. Let A, B, and C be c-points in a c-configuration. Then the ordered triple (A, B, C) is called collinear if
d(A, B) + d(B, C) = d(A, C). Show that if (A, B, C) is collinear, then ccos(∠ABC) = −1. [3 pts]
10. Suppose that A, B, C, and D are c-points in a c-configuration, and (B, D, C) is collinear. Prove that
ccos(∠ABD) = ccos(∠ABC) if and only if ccos(∠ACD) = ccos(∠ACB). [6 pts]
If A, B, C, D are c-points in a c-configuration, then define the ordered pair (A, D) to be a c-cevian of 4ABC
if (B, D, C) is collinear and the two equations in Problem 10 are satisfied.
11. Given that (A, D) is a c-cevian of 4ABC, prove that ccos(∠ADC) = − ccos(∠ADB). [4 pts]
12. Given that (A, D) is a c-cevian of 4ABC and ccos(∠ADC) = 0, prove that
1 2 2
sqar 4ABC = d(A, D) · d(B, C) . [4 pts]
4
5 Solutions to Power Question
32 +42 −22 7 22 +42 −32 11
1. a. The three c-cosines are ccos(2, 3, 4) = 2·3·4 = 8, ccos(3, 2, 4) = 2·2·4 = 16 , and ccos(4, 3, 2) =
32 +22 −42
2·3·2 = − 14 .
2 2 2
b. The answer is c = 1 + i. By the definition of c-cosine, ccos(i, 1, c) = 1 +c2c −i = 1 and ccos(1, i, c) =
i2 +c2 −12
2ic = 1. This implies 12 + c2 − i2 = 2c and i2 + c2 − 12 = 2ic. Add to obtain 2c2 = 2c(1 + i). Because
c 6= 0, divide both sides by 2c to obtain c = 1 + i. It can be verified that c = 1 + i does indeed satisfy the
conditions of the problem.
2. There are many possible c-triangles for each part. At least one example is given below for each.
a. [5, 3, 4], or in general, any c-triangle in which a2 = b2 + c2
√
b. [ 2i, 1, 1], or in general, any c-triangle in which b2 + c2 − 4bc = a2
√
c. [ 2, i, 1], or in general, any c-triangle in which b2 + c2 − 2ibc = a2
d. [3, 4, 5]
√ √
2 2
e. [3α, 4α, 5α], where α = 2 + 2 i. (Note that α4 = −1.)
f. [1 + i, 1 + i, 1 + i]
2 2 2
−1
3. By definition, ccos(1, 1, z) = 1 +z
2·1·z = 12 z. Therefore S is the set of all complex z 6= 0 with |z| ≤ 2; in other
words, the closed disk of radius 2 centered at 0 in the complex plane, excluding the origin.
4. One such pair of c-triangles is [1, 1, 3] and [1, 1, −3]. In fact, because the formula for c-cosine is an even function
in the first argument, it follows that ccos(c, a, b) = ccos(−c, a, b). Therefore any two c-triangles of the form
[a, b, c] and [a, b, −c] will satisfy the given condition. In fact, by setting ccos(c1 , a, b) = ccos(c2 , a, b), it can be
shown that these are the only examples.
5. a. One such c-triangle is [1, i, 2]. The c-cosines of its three c-angles are 1+(−1)−4
2i = 2i, 1+4−(−1)
4 = 32 , and
4+(−1)−1 1
4i = − 2 i. The first two c-angles are improper, and the third is proper.
1+(−1)−1 1+1−(−1)
b. One such c-triangle is [1, 1, i]. The c-cosines of its three c-angles are 2i = 12 i (twice) and 2 =
3
2 . The first two c-angles are proper, and the third is improper.
6. First, note that for any complex numbers x and y, the standard triangle inequality holds (that is, |x + y| ≤
|x| + |y|). Applying the triangle inequality to y and x − y gives |x| ≤ |y| + |x − y|, so |x − y| ≥ |x| − |y|. (This
will be referred to as the subtractive triangle inequality.)
Now consider a c-triangle [a, b, c], all of whose angles are proper. By definition, the c-angle (a, b, c) is proper,
so |b2 + c2 − a2 | ≤ |2bc|. By reversing the sign in the first absolute value expression, |a2 − b2 − c2 | ≤ |2bc|.
Using the subtractive triangle inequality from above twice, |a2 − b2 − c2 | ≥ |a2 − b2 | − |c2 | ≥ |a2 | − |b2 | − |c2 |.
Therefore
|a|2 − |b|2 − |c|2 ≤ 2 · |b| · |c|.
Rearranging and factoring gives |a|2 ≤ (|b| + |c|)2 . Then because both sides are positive real numbers, taking
the square root yields |a| ≤ |b| + |c|, as desired.
7. a. Observe that 16 sqar[a, b, c] = (a + b + c)(−a + b + c)(a − b + c)(a + b − c). Expand the right-hand side by
first multiplying out the first two and last two factors to obtain:
16 sqar[a, b, c] = (−a2 + b2 + c2 + 2bc)(a2 − b2 − c2 + 2bc) = 4b2 c2 − (b2 + c2 − a2 )2 .
Now factor out 4b2 c2 from the right-hand side, yielding:

(b2 + c2 − a2 )2
Å ã
16 sqar[a, b, c] = 4b2 c2 1 − = 4b2 c2 (1 − (ccos(∠A))2 ).
(2bc)2
Dividing both sides by 16 yields the desired equality.
b. From the result of part (a),
4 sqar[a, b, c] = b2 c2 (1 − (ccos(∠A))2 ) = a2 c2 (1 − (ccos(∠B))2 ).
Dividing both sides by a2 b2 c2 yields the desired equality.

8. Rearranging the result of Problem 7(b) and relying on the assumption that ccos(∠B1 ) 6= ±1 yields:
1 − (ccos(∠A1 ))2 a21

= .
1 − (ccos(∠B1 ))2 b21
However, because ccos(∠A1 ) = ccos(∠A2 ) and ccos(∠B1 ) = ccos(∠B2 ), it follows that
a21 a22 a2 b2
= , so =± .
b21 b22 a1 b1
a2
Similar arguments show that a1 = ± cc21 , so for some k 6= 0, a2 = ka1 , b2 = ekb1 , and c2 = f kc1 , where e = ±1
and f = ±1.
It remains to show that e = 1, for then aa12 = bb21 = k. Because ccos(∠C1 ) = ccos(∠C2 ) and e2 = f 2 = 1,
it follows that
a21 + b21 − c21 (ka1 )2 + (ekb1 )2 − (f kc1 )2 a2 + b21 − c21
= = 1 .
2a1 b1 2(ka1 )(ekb1 ) 2ea1 b1
As long as a21 + b21 6= c21 , it follows immediately that e = 1.
In the remaining case where a21 + b21 = c21 , note that a21 + c21 6= b21 and b21 + c21 6= a21 , so the argument above
suffices to show that
a2 c2 b2 c2
= and = ,
a1 c1 b1 c1
from which the desired conclusion follows.
9. For convenience, let x = d(A, B) and y = d(B, C), so d(A, C) = x + y. Then by definition,
x2 + y 2 − (x + y)2 −2xy
ccos(∠ABC) = = = −1.
2xy 2xy
10. For convenience, let a = d(B, C), b = d(A, C), c = d(A, B), x = d(A, D), y = d(B, D), and z = d(C, D).
Claim: Both given equations hold if and only if the following equation holds:
c2 z + b2 y = ayz + x2 a. (∗)
Proof: The following argument shows that the above equation is equivalent to ccos(∠ABD) = ccos(∠ABC); the
other equivalence follows by exchanging all instances of B and C, b and c, and y and z in this argument. Because
performing the exchange of b with c and y with z results in (∗), it follows that ccos(∠ABD) = ccos(∠ABC) if
and only if ccos(∠ACD) = ccos(∠ACB), as desired.
First, apply the definition of the c-cosine of a c-angle, to obtain

c2 + y 2 − x2 a2 + c2 − b2
ccos(∠ABD) = ccos(∠ABC) ⇐⇒ = .
2cy 2ac
Now, multiply by 2acy, which yields
ccos(∠ABD) = ccos(∠ABC) ⇐⇒ c2 a + y 2 a − x2 a = a2 y + c2 y − b2 y.
Rearrange the terms in the second equation and factor, to obtain
ccos(∠ABD) = ccos(∠ABC) ⇐⇒ c2 (a − y) + b2 y = ay(a − y) + x2 a.
Finally, because (B, D, C) is collinear, it follows from the definition that a = y + z, and substituting a − y = z
gives
ccos(∠ABD) = ccos(∠ABC) ⇐⇒ c2 z + b2 y = ayz + x2 a,
as desired. Note that all of these algebraic steps are reversible, so the equivalence is preserved at each stage.
11. Define a new c-configuration on four c-points {A0 , B 0 , C 0 , D0 }, having all of its distances the same as those of the
corresponding pairs of c-points in the original c-configuration with {A, B, C, D} (that is, d(A, B) = d(A0 , B 0 ),
and so forth), except that d(C 0 , D0 ) = −d(C, D).
Now note that ccos(∠A0 B 0 C 0 ) = ccos(∠ABC), because 4ABC and 4A0 B 0 C 0 are congruent c-triangles. Addi-
tionally, 4A0 B 0 D0 and 4ABD are congruent, so ccos(∠A0 B 0 D0 ) = ccos(∠ABD). Because (A, D) is a c-cevian
of 4ABC, it follows that ccos(∠ABC) = ccos(∠ABD), and therefore ccos(∠A0 B 0 C 0 ) = ccos(∠A0 B 0 D0 ).
Furthermore, (B 0 , C 0 , D0 ) is collinear because d(B 0 , D0 ) = d(B, D) = d(B, C) − d(C, D) = d(B 0 , C 0 ) + d(C 0 , D0 ).

Together with the above, it follows that (A0 , C 0 ) is a c-cevian of 4A0 B 0 D0 . Therefore ccos(∠A0 D0 C 0 ) =
ccos(∠A0 D0 B 0 ).
Borrowing the notation from the solution to Problem 10,
(−z)2 + x2 − b2
ccos(∠A0 D0 B 0 ) = ccos(∠A0 D0 C 0 ) = = − ccos(∠ADC).
2x(−z)
Again, from the congruence of 4ABD and 4A0 B 0 D0 , it follows that ccos(∠A0 D0 B 0 ) = ccos(∠ADB). Therefore
ccos(∠ADB) = − ccos(∠ADC), as desired.
12. Using the result of Problem 7(a) and the assumption that (A, D) is a c-cevian,
1 1
sqar 4ABC = d(A, C)2 · d(B, C)2 (1 − (ccos(∠ACB))2 ) = d(A, C)2 · d(B, C)2 (1 − (ccos(∠ACD))2 ).
4 4
From Problem 7(b) and the assumption that ccos(∠ADC) = 0, it follows that
d(A, D)2 d(A, D)2

1 − (ccos(∠ACD))2 = (1 − ccos(∠ADC)) · = .
d(A, C)2 d(A, C)2
Therefore
1 d(A, D)2 1
sqar 4ABC = d(A, C)2 · d(B, C)2 · = d(A, D)2 · d(B, C)2 ,
4 d(A, C)2 4
as desired.
6 Individual Problems
Problem 1. The sum of the degree measures of the angles of a convex n-gon is 1 more than the number of
diagonals of the n-gon. Compute n.
Problem 2. Compute the number of positive integers b ≥ 5 for which the base-ten number 2020 ends in the digit 4
when it is expressed in base b.
Problem 3. A microwave accepts three digits a, b, and c, not all zero, as input. The microwave will then run for
60a + 10b + c seconds. Compute the number of positive integers k such that the microwave can run for exactly
k seconds.
Problem 4. Let T RIAN GLE be an equilateral octagon with side length 10, and let α be the acute angle whose
tangent is 43 . Given that the measures of the interior angles of T RIAN GLE alternate between 180◦ − α and
90◦ + α, compute [T RIAN GLE].
Problem 5. A deck of cards consists of 2 purple cards and 6 orange cards. Zak draws a hand of three cards at
random from the deck. If all three cards are orange, he returns them to the deck, shuffles the deck, and repeats
the entire process until he has a hand containing at least one purple card. Then he draws one additional card.
Compute the probability that this last card is purple.
Problem 6. In 4ABC, AB = 6, BC = 7, and CA = 8. The circle that passes through each of the vertices of
4ABC is centered at point O. The circle that passes through each of the vertices of 4AOB intersects BC at
point D, different from B. Compute BD.
Problem 7. Let a1 , a2 , a3 , a4 , . . . be the increasing sequence 4, 5, 44, 45, . . . that consists of all positive integers
each of whose digits is either 4 or 5. Compute a100 .
Problem 8. For a given function f , define f 1 (x) = f (x), and for k ≥ 2, define
f k (x) = f (f k−1 (x)).
For example, f 3 (x) = f (f (f (x))). Of the 10! functions g whose domain and range are the set S = {1, 2, 3, . . . , 10},
compute the number of functions such that g g(n) (n) = n for all n ∈ S.
Problem 9. Compute the greatest integer that is the least common multiple of two distinct sets of four nonzero
digits.
Problem 10. Compute the number of ordered pairs (a, b) of integers such that |a| ≤ 2020, |b| ≤ 2020, and
… a 3 b
3 a + b
= + .
2 2 2
7 Answers to Individual Problems
Answer 1. 361
Answer 2. 32
Answer 3. 639
Answer 4. 480
1
Answer 5.
6
13 2
Ñ é
Answer 6. or 1
11 11
Answer 7. 544545
Answer 8. 46
Answer 9. 840
Answer 10. 27
8 Solutions to Individual Problems
Problem 1. The sum of the degree measures of the angles of a convex n-gon is 1 more than the number of
diagonals of the n-gon. Compute n.
Solution 1. In degrees, the sum of the measures of the angles of a convex n-gon is 180(n − 2) and the number of
n(n − 3) n(n − 3)
diagonals of a convex n-gon is . The goal is to find an n ≥ 3 such that 180(n − 2) = + 1. This
2 2
2
implies n − 363n + 722 = 0, so (n − 361)(n − 2) = 0. Because n > 2, it follows that the answer is n = 361.
Problem 2. Compute the number of positive integers b ≥ 5 for which the base-ten number 2020 ends in the digit 4
when it is expressed in base b.
Solution 2. For the base-b representation of N to end in the digit 4, b must be greater than 4, and N must be
congruent to 4 modulo b. In other words, N − 4 must be congruent to 0 modulo b, i.e., b must be a factor of
N − 4 that is greater than 4. For N = 2020, N − 4 = 2016 = 25 · 32 · 71 , so 2016 has (5 + 1)(2 + 1)(1 + 1) = 36
factors, including 1, 2, 3, and 4. Thus there are 36 − 4 = 32 possible values of b corresponding to the remaining
factors of 2016 greater than 4.
Problem 3. A microwave accepts three digits a, b, and c, not all zero, as input. The microwave will then run for
60a + 10b + c seconds. Compute the number of positive integers k such that the microwave can run for exactly
k seconds.
Solution 3. The least amount of time that the microwave can run is 1 second, corresponding to a = b = 0,
c = 1. The greatest amount of time that the microwave can run corresponds to a = b = c = 9, which gives
9(60 + 10 + 1) = 639 seconds. Note that every multiple of 10 between 10 and 9(60 + 10) = 630 can be expressed
in the form 60a + 10b, where a and b are digits, so all integer values between 1 and 639 inclusive are achievable.
Hence the answer is 639.
Problem 4. Let T RIAN GLE be an equilateral octagon with side length 10, and let α be the acute angle whose
tangent is 43 . Given that the measures of the interior angles of T RIAN GLE alternate between 180◦ − α and
90◦ + α, compute [T RIAN GLE].
←→ ←
→
Solution 4. Suppose that m∠T RI = 180◦ − α; then m∠RIA = 90◦ + α. Extend T R and IA to intersect at F ,
and note that the measures of the exterior angles F RI and F IR are α and 90◦ − α, respectively, so 4F IR is
a 6−8−10 right triangle.
E L
• •
T•
•G
180◦ − α
R•
•N
90◦ + α
F• • •
I A
Accordingly, the octagon can be inscribed in a 24 × 24 square with four of the octagon’s sides coincident with
the sides of the square. Thus the area of the octagon is
1
242 − 4 · · 6 · 8 = 576 − 96 = 480.
2
Problem 5. A deck of cards consists of 2 purple cards and 6 orange cards. Zak draws a hand of three cards at
random from the deck. If all three cards are orange, he returns them to the deck, shuffles the deck, and repeats
the entire process until he has a hand containing at least one purple card. Then he draws one additional card.
Compute the probability that this last card is purple.
Solution 5. Zak can draw 63 = 20 hands of 3 orange cards, 62 · 2 = 30 hands of 2 orange cards and 1 purple

card, and 6 hands of 1 orange and 2 purple cards. If Zak draws one of the latter two types of hands, he keeps
that hand, and each of those 36 hands is equally likely to be drawn. The final card he draws will be purple if
and only if exactly one of his initial cards is purple, and in that case, the final card is orange with probability
1 30 1 1
, hence the probability of this occurring is · = .
5 36 5 6
Problem 6. In 4ABC, AB = 6, BC = 7, and CA = 8. The circle that passes through each of the vertices of
4ABC is centered at point O. The circle that passes through each of the vertices of 4AOB intersects BC at
point D, different from B. Compute BD.
Solution 6. Note that m∠AOB = 2m∠ACB because ∠ACB intercepts arc AB in the circle centered at O. Also,
m∠AOB = m∠ADB because both angles intercept arc AB in the circle passing through each of the vertices
of 4AOB. By the Exterior Angle Theorem, m∠CAD + m∠ACD = m∠ADB. Thus
m∠CAD + m∠ACD = m∠ADB → m∠CAD + m∠ACB = m∠AOB = 2m∠ACB

→ m∠CAD = m∠ACB = m∠ACD.
Therefore 4ADC is isosceles and AD = CD. Let M be the midpoint of AC. Then the perpendicular bisector
←−→
of AC is DM , as shown.
C
64
11
O
D
4
13
11
A 6 B
By the Law of Cosines, it follows that
CM 82 + 72 − 62 11
= cos ∠C = = .
CD 2·8·7 16
64 13
Because CM = 4, it follows that CD = and BD = BC − CD = .
11 11
Problem 7. Let a1 , a2 , a3 , a4 , . . . be the increasing sequence 4, 5, 44, 45, . . . that consists of all positive integers
each of whose digits is either 4 or 5. Compute a100 .
Solution 7. Note that the number of k-digit numbers in the sequence is 2k , and that there is a bijection between
the set of numbers {0, 1, . . . , 2k − 1} and the set containing the 2k k-digit numbers in the given sequence. By
converting the numbers 0, 1, . . . , 2k − 1 to binary and allowing for leading zeros, the bijection is given by
0 7→ 4 and 1 7→ 5. (∗)
The first k-digit number in the sequence is a2k −1 . Because 26 − 1 ≤ 100 < 27 − 1, it follows that a100 has six
digits. There are 100 − 63 + 1 numbers among 63, 64, . . . , 100, and so a100 is the 38th smallest six-digit number
in the sequence. The value of a100 can thus be obtained by converting 37 to binary (noting that 37 is the 38th
term in the sequence 0, 1, 2, . . . , 36, 37) and applying the bijection in (∗). Because 3710 = 1001012 , it follows
that a100 = 544545.
Problem 8. For a given function f , define f 1 (x) = f (x), and for k ≥ 2, define
f k (x) = f (f k−1 (x)).
For example, f 3 (x) = f (f (f (x))). Of the 10! functions g whose domain and range are the set S = {1, 2, 3, . . . , 10},
compute the number of functions such that g g(n) (n) = n for all n ∈ S.
Solution 8. This solution will focus on the idea of a permutation. A permutation is a rearrangement of the
elements of a set, thus giving rise to a one-to-one correspondence. This idea is important when considering the
problem at hand. To illustrate this point, consider the following function g, which maps S to S.
n 1 2 3 4 5 6 7 8 9 10
g(n) 7 9 10 2 5 6 4 3 1 8
By considering the row of values of n as an ordered list, note that the row of values of g(n) is a permutation
of the row of values of n. This is true for all functions g whose domain and range are both the set S.
It follows that at least one of the elements of {g(k), g 2 (k), g 3 (k), . . . , g 10 (k)} is k. The problem statement
is therefore equivalent to asking how many functions g have the property that a “cycle” of g(n) iterations of g
on n produces the output n.
To analyze how elements cycle back to themselves, cycle notation will be used. If a sequence of distinct
numbers a1 , . . . , ak satisfies g(ai ) = ai+1 for i = 1, . . . , k − 1, and g(ak ) = a1 , this sequence will be called a
k-cycle of g and denoted by (a1 , a2 , . . . , ak ). Using this notation, any permutation can be expressed in terms
of its cycles. To illustrate, recall the example from above:
n 1 2 3 4 5 6 7 8 9 10
g(n) 7 9 10 2 5 6 4 3 1 8
Then g can be expressed as the “product” of cycles: (1, 7, 4, 2, 9)(3, 10, 8)(5)(6). Notice that in this example,
g n (1) = 1 exactly when n is a multiple of 5, because 1 is in a cycle of length 5. In general, if g g(n) (n) = n, then
g(n) must be a multiple of the cycle length of n in g.
The condition in the problem is equivalent to the condition that all elements in a cycle in a permutation
of {1, 2, 3, . . . , 10} must be divisible by that cycle’s length. Therefore a permutation that satisfies the property
can contain no cycle of length greater than 3 (as there are not, for example, four distinct elements of S that
are multiples of 4).
Suppose the permutation contains a 3-cycle, which must consist of the elements 3, 6, and 9. Then there
may be zero, one, or two 2-cycles consisting of some or all of the elements 2, 4, 8, and 10, with all remaining
elements in 1-cycles. In this case there are 2 possible 3-cycles, 3 ways to have exactly two 2-cycles, 6 ways to
have exactly one 2-cycle, and 1 way to have zero 2-cycles, resulting in 2(3 + 6 + 1) = 20 permutations that
satisfy the conditions of the problem and contain a 3-cycle.
Suppose instead that the permutation does not contain a 3-cycle. Then there may be zero, one, or two
2-cycles consisting of some of the elements 2, 4, 6, 8, and 10, with all remaining elements in 1-cycles. There
are 21 52 32 = 15 ways to select two pairs of elements to appear in 2-cycles, 52 = 10 ways to select one pair of

elements to appear in a 2-cycle, and 1 way to have zero 2-cycles, resulting in 15 + 10 + 1 = 26 permutations
that satisfy the conditions of the problem and do not contain a 3-cycle.
Thus the final answer is 20 + 26 = 46.
Problem 9. Compute the greatest integer that is the least common multiple of two distinct sets of four nonzero
digits.
Solution 9. The least common multiple of 1, 2, 3, 4, 5, 6, 7, 8, 9 equals 5 · 7 · 8 · 9 = 2520, and thus the answer
is a factor of 2520. Check the factors of 2520 from greatest to least. Note that 2520 = lcm(5, 7, 8, 9), but 2520
is not the least common multiple of any other set of four nonzero digits. Also note that 2520/2 = 1260 =
lcm(4, 5, 7, 9), but 1260 is not the least common multiple of any other set of four nonzero digits. Finally, note
that 840 = 2520/3 = lcm(3, 5, 7, 8) = lcm(5, 6, 7, 8). Thus the answer is 840.
Problem 10. Compute the number of ordered pairs (a, b) of integers such that |a| ≤ 2020, |b| ≤ 2020, and
… a 3 b
3 a + b
= + .
2 2 2
a
Solution 10. Let x = 2 and y = 2b , so that the given equation is equivalent to
√
3
x + y = x3 + y, (∗)
where x and y are either integers or half-integers which are at most 1010 in absolute value. There are two main
observations. First, after cubing both sides of (∗), the resulting equation is cubic in y. Second, y = −(x3 + x)
is a solution to (∗). It follows that
(y + x3 )3 − (y + x) = y 3 + 3x3 · y 2 + (3x6 − 1)y + (x9 − x)

= y + (x3 + x) y 2 + (2x3 − x)y + (x2 − 1)(x4 + 1) .

Using the quadratic formula results in three solutions for y (possibly nonreal) in terms of x:
y = −(x3 + x),
√
4 − 3x2
3 x+
y = −x + , and
√2
x − 4 − 3x2
y = −x3 + .
2
Consider first the case where x ∈ −1, − 12 , 0, 12 , 1 . Checking each value x in this set, note that there are 9

different ordered pairs:
• for x = −1, the pairs are (−1, 0), (−1, 1), (−1, 2);
• for x = 0, the pairs are (0, −1), (0, 0), (0, 1);
• for x = 1, the pairs are (1, 0), (1, −1), (1, −2).
For |x| ≥ 32 , the latter two solutions for y are imaginary, so y = −(x3 + x). Now |y| ≤ 1010 if and only if
|x| ≤ 10, and y is an integer or a half-integer only when x is an integer. Therefore there is one additional
solution for each x ∈ {−10, −9, −8, . . . , −2} ∪ {2, 3, . . . , 10}, for an additional 9 + 9 = 18 solutions. Thus the
final answer is 9 + 18 = 27.
√
Alternate Solution: Multiply both sides of the given√ equation by 8 to obtain 4 3 4a + 4b = a3 + 4b. Because
a and b are integers, so is a3 + 4b, and therefore 4a + 4b must also be an integer. Therefore 4a + 4b is a
3
perfect cube. Furthermore, 4a + 4b is even, so 4a + 4b = (2x)3√

→ a + b = 2x3 for some integer x, which implies
3
b = 2x − a. Substituting and rearranging, it follows that 4 4a + 8x3 − 4a = a3 + 8x3 − 4a, which implies
3
8x − 8x3 = a3 − 4a. Factoring yields −(2x − 2)(2x)(2x + 2) = (a − 2)(a)(a + 2).
Define the function f (t) = (t − 2)(t)(t + 2). Then it follows that −f (2x) = f (a). Note that f is an odd
function, and so f (−2x) = f (a). Further note that f is an increasing function over (−∞, −2) ∪ (2, ∞), and
each output of f on this interval corresponds to a unique real input. Thus if x ∈ / [−1, 1], it follows that a = −2x
and hence b = 2x3 + 2x. The only values of x ∈ / [−1, 1] that give solutions within the bounds |a|, |b| ≤ 2020 are
x ∈ {±2, ±3, . . . , ±10}, for a total of 18 solutions. Now, consider x ∈ {−1, 0, 1}. Because f (a) = f (−2x) = 0
for these x-values, it follows that a ∈ {−2, 0, 2}. These nine ordered pairs (a, x) give nine distinct ordered pairs
(a, b) that satisfy the given equation, so the total number of solutions is 18 + 9 = 27, as above.
9 Collaborative Relay Problems
Problem 1A-1. Compute the number of eight-digit positive integers whose digits from left to right are strictly
decreasing.
Problem 1A-2. Let T be the number you will receive. Compute the number of digits in the decimal expansion
of 20T .
Problem 1A-3. Let T be the number you will receive. The figure below results from adjoining two a × b rectangles,
where a and b are positive integers, a < b, and one of the smaller sides of one rectangle fully overlaps with a
larger side of the other rectangle. Given that the perimeter of the figure is 2T , compute the number of possible
ordered pairs (a, b).
Problem 1B-1. The sum of twenty consecutive integers is 310. Compute the least of these twenty integers.
Problem 1B-2. Let T be the number you will receive. Compute the least three-digit number T B C such that when
the three-digit number is increased by the sum of its digits, the result is a perfect square.
Problem 1B-3. Let T be the number you will receive. Compute the number of elements n of the set {1, 2, 3, . . . , T }
for which the product (n2 − n + 3)(n2 + n + 3) is a multiple of 5.
Problem 1C-1. Compute the least two-digit positive integer, neither of whose digits is 0, that is 27 greater than
the two-digit number formed by reversing its digits.
Problem 1C-2. Let T be the number you will receive. Given isosceles triangle ABC with AB = 18 and BC = T ,
compute the area of 4ABC.
Problem 1C-3. Let T be the number you will receive. Given that N is a real number such that the region in the
xy-plane defined by |y| + |2x| ≤ N has area T , compute N .
Problem 1D-1. The mean of four numbers is 6. The mean of the least three numbers is 5. The mean of the greatest
three numbers is 7. Compute the quotient when the greatest number is divided by the least number.
1
Problem 1D-2. Let T be the number you will receive. Let U = 1 + . Compute U T /U .
T −1
Problem 1D-3. Let T be the number you will receive. Let m and n be real numbers with m < n. Given that
1 1
m2 − 4m = n2 − 4n = T , compute 2 − 2 .
m n
Problem 1E-1. Each interior angle of a regular n-gon has a measure that is 4 times the measure of each exterior
angle of the n-gon. Compute n.
Problem 1E-2. Let T be the number you will receive. An isosceles trapezoid has parallel sides of length 16 and 28
and non-parallel sides of length T . Compute the area of the trapezoid.
Problem 1E-3. Let T be the number you will receive. Compute the number of positive integers n less than T for
which jnk jnk jnk n n n
+ + = + + .
3 4 5 3 4 5
Problem 2A-1. Compute the sum of all positive integers n such that 2021 · n is an n-digit number.
Problem 2A-2. Let T be the number you will receive. Paul notices that 2T years from now he will be twice as old
as he was T years ago. Compute Paul’s current age.
Problem 2A-3. Let T be the number you will receive. Sally draws an n-gon and notices that the degree measure
of each interior angle is an integer multiple of 3T /2 and that not all interior angles have the same measure.
Compute the least possible value of n.
Problem 2B-1. Compute the positive integer b for which the base-10 number 2021 is equivalent to the base-b
number 3745.
Problem 2B-2. Let T be the number you will receive. Circles Ω1 and Ω2 are externally tangent. Circle Ω1 is
centered at point L and has radius T , and circle Ω2 is centered at point O and has radius T + 1. Point E lies
←→
on circle Ω1 so that EO is tangent to circle Ω1 . Compute EO.
Problem 2B-3. Let T be the number you will receive. Compute the least positive real number x such that
x + bxc
= T.
x − bxc
Problem 2C-1. Compute the greatest value of x that satisfies x3 − 62x2 − 4x + 248 = 0.
Problem 2C-2. Let T be the number you will receive. A rectangle has perimeter T and a diagonal of length 25.
Compute the area of the rectangle.
Problem 2C-3. Let T be the number you will receive. Compute the units digit of the sum
T + (T + 1) + (T + 2) + · · · + (3T − 1) + (3T ).
Problem 2D-1. Compute the greatest of 43 consecutive integers whose sum is 2021.
Problem 2D-2. Let T be the number you will receive. Compute the number of positive integers less than or equal
to 2021 that are relatively prime to T .
Problem 2D-3. Let T be the number you will receive. Consider the set of all lines with slope m and y-intercept
(0, b) such that 3m + 2b = T . Every line in the set has one particular point (h, k) in common. Compute h · k.
Problem 2E-1. Cara opens a book and notes the two consecutive page numbers. Given that the difference of the
squares of the page numbers is 33, compute the lesser page number.
Problem 2E-2. Let T be the number you will receive. In a circle, T chords are drawn. Compute the maximum
possible number of points of intersection of the chords.
T
Problem 2E-3. Let T be the number you will receive. A set U has 10 elements. Compute the sum of |S| over all
subsets S ⊆ U. (Note: For a finite set A, the notation |A| represents the number of elements of A.)
10 Collaborative Relay Answers
Answer 1A-1. 45
Answer 1A-2. 59
Answer 1A-3. 10
Answer 1B-1. 6
Answer 1B-2. 614
Answer 1B-3. 492
Answer 1C-1. 41
Answer 1C-2. 360

√
Answer 1C-3. 6 10
Answer 1D-1. 3
9 1
Ñ é
Answer 1D-2. or 2 or 2.25

4 4
320 77
Ñ é
Answer 1D-3. or 3
81 81
Answer 1E-1. 10
Answer 1E-2. 176
Answer 1E-3. 2
Answer 2A-1. 9
Answer 2A-2. 36
Answer 2A-3. 5
Answer 2B-1. 8
Answer 2B-2. 15
8 1
Ñ é
Answer 2B-3. or 1
7 7
Answer 2C-1. 62
Answer 2C-2. 168
Answer 2C-3. 2
Answer 2D-1. 68
Answer 2D-2. 952
Answer 2D-3. 714
Answer 2E-1. 16
Answer 2E-2. 120
Answer 2E-3. 24576
11 Collaborative Relay Solutions
Problem 1A-1. Compute the number of eight-digit positive integers whose digits from left to right are strictly
decreasing.
Solution 1A-1. For any integer N satisfying the conditions of the problem, exactly two of the ten digits 0, 1, . . . , 9
do not appear in N . There are 10

2 = 45 ways to choose the two digits that do not appear in N , and there is
exactly one way to order the remaining eight digits to form an integer whose digits are strictly decreasing from
left to right. Hence the answer is 45.
Alternate Solution: Because any integer satisfying the conditions of the problem has eight digits, the leftmost
digit must be greater than 6. Now consider three cases.
• If the leftmost digit is 7, there is only one possibility: 76543210.
• If the leftmost digit is 8, then the rightmost seven digits are seven of the eight digits 7, 6, 5, 4, 3, 2, 1, 0.
There are 8 ways to choose which of these eight digits is not used, and then there is exactly one way to
order the remaining seven digits to form an integer whose digits are strictly decreasing from left to right.
• If the leftmost
digit is 9, then the rightmost seven digits are seven of the nine digits 8, 7, 6, 5, 4, 3, 2, 1, 0.
There are 92 = 36 ways to choose the two digits that are not used, and then there is exactly one way to
order the remaining seven digits to form an integer whose digits are strictly decreasing from left to right.
Thus the answer is 1 + 8 + 36 = 45.
Problem 1A-2. Let T be the number you will receive. Compute the number of digits in the decimal expansion
of 20T .
Solution 1A-2. Notice that 210 = 1024 ≈ 103 so 245 = 25 · (210 )4 ≈ 32 · 1012 , which has 14 digits. Multiplying by
1045 simply appends 45 zeroes, so 2045 has 14 + 45 = 59 digits.
Problem 1A-3. Let T be the number you will receive. The figure below results from adjoining two a × b rectangles,
where a and b are positive integers, a < b, and one of the smaller sides of one rectangle fully overlaps with a
larger side of the other rectangle. Given that the perimeter of the figure is 2T , compute the number of possible
ordered pairs (a, b).
Solution 1A-3. The perimeter of the figure is 2(a + b) + 2b = 2a + 4b. Hence 2a + 4b = 2T , so a + 2b = T . From
a < b, it follows that T = a + 2b < 3b, so b > T /3. Moreover, if T is odd, then the greatest possible value of b
is bT /2c, and if T is even, then the greatest possible value of b is T −2
2 . Hence the greatest possible value of b
is dT /2e − 1. Given a fixed value of b, the value of a is uniquely determined and the number of integral values
of b in the interval (T /3, dT /2e − 1] is dT /2e − 1 − (bT /3c + 1) + 1 = dT /2e − bT /3c − 1. With T = 59, the
answer is 30 − 19 − 1 = 10.
Problem 1B-1. The sum of twenty consecutive integers is 310. Compute the least of these twenty integers.
Solution 1B-1. The twenty consecutive integers have a sum of 310 = 10 · 31, so it follows that the sum of the least
and greatest integers is 31. If the least integer is x, then x + (x + 19) = 31 → x = 6.
Problem 1B-2. Let T be the number you will receive. Compute the least three-digit number T B C such that when
the three-digit number is increased by the sum of its digits, the result is a perfect square.
Solution 1B-2. The conditions of the problem imply that 100T + 10B + C + (T + B + C) = 101T + 11B + 2C is
a perfect square. With T = 6, the sum is 606 + 11B + 2C. The least perfect square greater than 606 is 625,
which can be obtained by taking B = 1 and C = 4, so the answer is 614.
Problem 1B-3. Let T be the number you will receive. Compute the number of elements n of the set {1, 2, 3, . . . , T }
for which the product (n2 − n + 3)(n2 + n + 3) is a multiple of 5.
Solution 1B-3. Notice that (n2 − n + 3)(n2 + n + 3) = n4 + 5n2 + 9. For this to be a multiple of 5, it follows that
n4 + 9 ≡ 0 (mod 5) → n4 ≡ 1 (mod 5). This can be verified to be true if and only if n 6≡ 0 (mod 5) (it is
also true by Fermat’s Little Theorem). The number of positive integers less than or equal to T that are not
multiples of 5 is T − T5 . With T = 614, this is 614 − 122, so the answer is 492.

Problem 1C-1. Compute the least two-digit positive integer, neither of whose digits is 0, that is 27 greater than
the two-digit number formed by reversing its digits.
Solution 1C-1. Suppose the number has two digits and is equal to 10t + u. The problem statement implies
10t + u = 10u + t + 27 → 9(t − u) = 27. Thus t − u = 3. Because t and u must be nonzero, the least positive
two-digit integer satisfying the conditions of the problem will have u = 1 and t = 4. Thus the answer is 41.
Problem 1C-2. Let T be the number you will receive. Given isosceles triangle ABC with AB = 18 and BC = T ,
compute the area of 4ABC.
√ p
Solution 1C-2. The altitude drawn to the base of the isosceles triangle has length either T 2 − 92 or 182 − (T /2)2 ,
√ T
p
depending on which side is the base. This implies that the area is either 9 T − 9 or 2 18 − (T /2)2 . With
2 2 2
T = 41, note that the lengths

√ 18, 18, 41 cannot be the lengths of the sides of a triangle because 18 + 18 < 41,
so the desired area is 9 412 − 92 = 9 · 40 = 360.
Problem 1C-3. Let T be the number you will receive. Given that N is a real number such that the region in the
xy-plane defined by |y| + |2x| ≤ N has area T , compute N .
Solution 1C-3. The region is a rhombus whose diagonals are on the coordinate√axes and√have lengths 2N and N .
The area of the region is thus N 2 = T . With T = 360, the answer is N = 360 = 6 10.
Problem 1D-1. The mean of four numbers is 6. The mean of the least three numbers is 5. The mean of the greatest
three numbers is 7. Compute the quotient when the greatest number is divided by the least number.
Solution 1D-1. The sum of all four numbers is 6 · 4 = 24. The sum of the least three numbers is 5 · 3 = 15, so the
greatest number is 24 − 15 = 9. The sum of the greatest three numbers is 7 · 3 = 21, so the least number is
24 − 21 = 3. The desired quotient is 9 ÷ 3 = 3.
1
Problem 1D-2. Let T be the number you will receive. Let U = 1 + . Compute U T /U .
T −1
Solution 1D-2. Note that U = T

T −1 , so the requested expression simplifies to U T −1 . With T = 3, U = 23 , and the
requested expression equals ( 32 )2 = 94 .
Problem 1D-3. Let T be the number you will receive. Let m and n be real numbers with m < n. Given that
1 1
m2 − 4m = n2 − 4n = T , compute 2 − 2 .
m n
Solution 1D-3. First note that m and n are the two zeroes of the polynomial f (x) = x2 − 4x − T . Also note that
1 1 n2 − m2
− = . To find n2 − m2 , note that n2 − m2 = (n + m)(n − m) = 4(n − m) by Vieta’s Formulas.
m2 n2 (mn)2 √
Because (n − m)2 = (n + m)2 − 4mn = 42 + 4T and n > m, it follows that n − m = 16 + 4T . With T = 49 ,
the answer is 320
81 .
Problem 1E-1. Each interior angle of a regular n-gon has a measure that is 4 times the measure of each exterior
angle of the n-gon. Compute n.
360◦
Solution 1E-1. Solve 180◦ n−2 360

n =4· n to obtain 180 = 5 · n → n = 10.
Problem 1E-2. Let T be the number you will receive. An isosceles trapezoid has parallel sides of length 16 and 28
and non-parallel sides of length T . Compute the area of the trapezoid.
Solution 1E-2. The trapezoid √ can be broken up into two triangles with bases that measure (28 − 16)/2 = 6 and
with heights h of length T 2 − 62 and a 16 × h rectangle. The area of the trapezoid is h2 (16 + 28) = 22h. With
T = 10, it follows that h = 8 and the area is 22 · 8 = 176.
Problem 1E-3. Let T be the number you will receive. Compute the number of positive integers n less than T for
which jnk jnk jnk n n n
+ + = + + .
3 4 5 3 4 5
Solution 1E-3. Note that bxc ≤ x for any real number x, with equality holding if and only if x is an integer. Thus
for any integer k, bn/kc ≤ n/k, with equality holding if and only if n is an integral multiple of k. The given
equation is therefore satisfied if and only if bn/3c = n/3, bn/4c = n/4, and bn/5c = n/5, i.e., n is a multiple of
lcm(3, 4, 5) = 60. The desired number of integers n is therefore the number of multiples of 60 between 1 and T
exclusive. With T = 176 = 2 · 60 + 56, the answer is 2.
Problem 2A-1. Compute the sum of all positive integers n such that 2021 · n is an n-digit number.
Solution 2A-1. Because any positive multiple of 2021 will have at least four digits, n must be at least 4. Note that
2021 · 4 = 8084 is a four-digit number, and 2021 · 5 = 10105 is a five-digit number. The following argument
establishes that no greater values of n can work. First note that if 10n−1 < 2021n < 10n , then 2021n has n
digits. Because 10n−1 > 2021n for n = 6 and the exponential function on the left grows faster than the linear
function on the right for n ≥ 6, it follows that 2021n has fewer than n digits for n ≥ 6. Thus the answer is
4 + 5 = 9.
Problem 2A-2. Let T be the number you will receive. Paul notices that 2T years from now he will be twice as old
as he was T years ago. Compute Paul’s current age.
Solution 2A-2. Let P denote Paul’s current age. Then P + 2T = 2(P − T ), so P = 4T . Substituting T = 9,
P = 36.
Problem 2A-3. Let T be the number you will receive. Sally draws an n-gon and notices that the degree measure
of each interior angle is an integer multiple of 3T /2 and that not all interior angles have the same measure.
Compute the least possible value of n.
Solution 2A-3. The sum of the measures of the interior angles of an n-gon is 180◦ (n−2), which must equal (3T k/2)◦
for some positive integer k. Rewrite this as 120 k
T = n−2 . Thus a candidate for the minimum value of n will occur
when n − 2 is the denominator of 120 T in lowest terms. Such a value must be verified to be achievable, i.e., it
must be possible to construct an n-gon, all of whose interior angle measures are integral multiples of (3T /2)◦ ,
and are not all equal. With T = 36, look for solutions with minimal n to 10 k
3 = n−2 . Because k and n are
positive integers, it follows that n − 2 ≥ 3 → n = 5. Note that a pentagon satisfying the problem’s conditions
can be drawn with angle measures 54◦ , 54◦ , 108◦ , 162◦ , and 162◦ , so the answer is 5.
Problem 2B-1. Compute the positive integer b for which the base-10 number 2021 is equivalent to the base-b
number 3745.
Solution 2B-1. Solve 3b3 + 7b2 + 4b + 5 = 2021 to obtain b(3b2 + 7b + 4) = 2016 = 8 · 252. Note that if b = 8, then
3b2 + 7b + 4 = 192 + 56 + 4 = 252, so the answer is 8.
Problem 2B-2. Let T be the number you will receive. Circles Ω1 and Ω2 are externally tangent. Circle Ω1 is
centered at point L and has radius T , and circle Ω2 is centered at point O and has radius T + 1. Point E lies
←→
on circle Ω1 so that EO is tangent to circle Ω1 . Compute EO.
Solution 2B-2. Let ` = EO. Because LE ⊥ EO, it follows that ` and T are the lengths √ of the legs of a right
triangle with hypotenuse T + (T + 1) = 2T + 1, so `2 + T 2 = (2T + 1)2 , hence ` = 3T 2 + 4T + 1. With T = 8,
it follows that EO = ` = 15.
Problem 2B-3. Let T be the number you will receive. Compute the least positive real number x such that
x + bxc
= T.
x − bxc
Solution 2B-3. First note that x cannot be an integer because otherwise, the denominator of the left-hand side of
the given equation would be 0. By multiplying and rearranging, note that the given equation is equivalent to
x(T − 1) = bxc(T + 1). To find the least possible value of x, consider the possible values of bxc. If bxc = 0,
then T = 1, which implies there is no minimum value of x. If bxc = 1, then x = TT −1
+1
. With T = 15, the answer
8
8
is 7 (and indeed, 7 = 1).
Problem 2C-1. Compute the greatest value of x that satisfies x3 − 62x2 − 4x + 248 = 0.
Solution 2C-1. The given equation is equivalent to x2 (x − 62) − 4(x − 62) = 0, which implies (x2 − 4)(x − 62) = 0,
so the solutions are ±2, 62, the greatest of which is 62.
Problem 2C-2. Let T be the number you will receive. A rectangle has perimeter T and a diagonal of length 25.
Compute the area of the rectangle.
Solution 2C-2. Let b and h be the base and height of the rectangle, respectively. Because the perimeter of the
rectangle is T , it follows that b + h = T2 . Because the diagonal has length 25, it follows that b2 + h2 = 625.
(T /2)2 −625
Squaring the first equation and then subtracting the second equation gives 2bh = ( T2 )2 −625, so bh = 2 .
For T = 62, this equals 168 (and in fact, {b, h} = {7, 24}).
Problem 2C-3. Let T be the number you will receive. Compute the units digit of the sum
T + (T + 1) + (T + 2) + · · · + (3T − 1) + (3T ).
Solution 2C-3. There are 3T − T + 1 = 2T + 1 terms in the given arithmetic series, and the sum is (2T +1)(4T 2
)
=
2T (2T + 1). Note that the sum is even for all T . For T = 1, 2, 3, . . ., the units digits of the sums are 6, 0, 2, 2,
0, 6, 0, 2, 2, 0, . . . in a repeating cycle of length 5. With T = 168 ≡ 3 (mod 5), the units digit of the sum is 2.
(Note that the actual sum is 113232, which confirms the answer.)
Problem 2D-1. Compute the greatest of 43 consecutive integers whose sum is 2021.
Solution 2D-1. Because 2021 = 43 · 47, the middle integer in the ordered list of 43 consecutive integers is 47. With
21 integers on either side of the middle integer, it follows that the greatest integer in the list is 47 + 21 = 68.
Problem 2D-2. Let T be the number you will receive. Compute the number of positive integers less than or equal
to 2021 that are relatively prime to T .
Solution 2D-2. Note that T = 68 = 22 · 17. There are b2021/2c = 1010 multiples of 2 that share a factor
with 68. There are b2021/17c = 118 multiples of 17 that share a factor with 68. However, this overcounts by
b2021/34c = 59, so there are 1010 + 118 − 59 = 1069 positive integers less than or equal to 2021 that share a
factor with 68. Thus the answer is 2021 − 1069 = 952.
Problem 2D-3. Let T be the number you will receive. Consider the set of all lines with slope m and y-intercept
(0, b) such that 3m + 2b = T . Every line in the set has one particular point (h, k) in common. Compute h · k.
Solution 2D-3. The lines in the given set have equation y = mx + T −3m
2 , which is equivalent to y − T2 = m(x − 32 ).
These lines all pass through the point 2 , 2 . With T = 952, the answer is 3T
3 T

4 = 714.
Problem 2E-1. Cara opens a book and notes the two consecutive page numbers. Given that the difference of the
squares of the page numbers is 33, compute the lesser page number.
Solution 2E-1. Let the lesser page number be n. Because (n + 1)2 − n2 = 33, it follows that 2n + 1 = 33 → n = 16.
Problem 2E-2. Let T be the number you will receive. In a circle, T chords are drawn. Compute the maximum
possible number of points of intersection of the chords.
T
T (T −1)
Solution 2E-2. The maximum possible number of points of intersection is given by 2 = 2 . With T = 16,
the answer is 16·15
2 = 120.
T
Problem 2E-3. Let T be the number you will receive. A set U has 10 elements. Compute the sum of |S| over all
subsets S ⊆ U. (Note: For a finite set A, the notation |A| represents the number of elements of A.)
T
. Set U has 1 subset of size 0, t subsets of size 1, 2t subsets of size 2, and so on. Note

Solution 2E-3. Let t = 10
that each subset S of size n has a “complementary” subset of size t − n, namely U − S. Because there are 2t
subsets of U, the desired sum is 21 · 2t · t = t · 2t−1 . With T = 120, it follows that t = 12, and so the answer
is 24576.
12 Tiebreaker Problem
Problem 1. Let ABCDE be a convex pentagon inscribed in a circle. The arcs AB, BC, CD, DE, and EA have
>>>> >
measures 32◦ , 64◦ , 84◦ , 84◦ , and 96◦ , respectively. Lines BC and AD meet at point X, lines AB and DE
meet at point Y , and point M is the midpoint of DY . The degree measure of the acute angle formed by the
intersection of lines XM and AB is T ◦ . Compute T .
13 Tiebreaker Answer
Answer 1. 58
14 Tiebreaker Solution
Problem 1. Let ABCDE be a convex pentagon inscribed in a circle. The arcs AB, BC, CD, DE, and EA have
>>>> >
measures 32◦ , 64◦ , 84◦ , 84◦ , and 96◦ , respectively. Lines BC and AD meet at point X, lines AB and DE
meet at point Y , and point M is the midpoint of DY . The degree measure of the acute angle formed by the
intersection of lines XM and AB is T ◦ . Compute T .
Solution 1. First, note that
m∠XBA = 180◦ − m∠ABC = 180◦ − (84◦ + 84◦ + 96◦ )/2 = 48◦ = m∠XDY,
so quadrilateral BDY X is cyclic. Because m∠DBA = (84◦ + 96◦ )/2 = 90◦ , it follows that m∠DXY = 90◦ .
So XM is a median of right 4DXY , and thus m∠DXM = m∠XDM = 48◦ . On the other hand, m∠Y AX =
m∠BAD = (64◦ + 84◦ )/2 = 74◦ . Thus the answer is 180◦ − (48◦ + 74◦ ) = 58◦ , so the requested value of T
is 58.
X
Y
A
B
M
E C

ARML 2021 Contest Packet PDF

Uploaded by

Copyright:

Available Formats

You might also like

ARML 2021 Contest Packet PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ARML 2021 Contest Packet PDF

Uploaded by

Copyright:

Available Formats

ARML Competition 2021

George Reuter, Head Writer

Answer 10. 610

Solution 1. Let M = 1010101010101010101. Note that N = 20M and that

Solution 2. There are 20

for all x in the interval, as 10 < 3π + Arccos( π6 ) < 7π

Thus there are a total of 3 intersection points.

By Power of a Point from B, BP · BQ = BY · BX = 31 · 23 = 29 ; thus if BQ = x, then BP = 9x2

Hence CD1 F1 E1 is a rectangle and

m∠CQD1 = m∠F1 QC = m∠RF1 C = m∠CF1 E1 = m∠CD1 E1 .

Solution 10. The given equation can be rewritten as

The squarea (“squared area”) of a c-triangle is defined by the formula

16 sqar[a, b, c] = (−a2 + b2 + c2 + 2bc)(a2 − b2 − c2 + 2bc) = 4b2 c2 − (b2 + c2 − a2 )2 .

Now factor out 4b2 c2 from the right-hand side, yielding:

4 sqar[a, b, c] = b2 c2 (1 − (ccos(∠A))2 ) = a2 c2 (1 − (ccos(∠B))2 ).

Dividing both sides by a2 b2 c2 yields the desired equality.

1 − (ccos(∠A1 ))2 a21

First, apply the definition of the c-cosine of a c-angle, to obtain

Rearrange the terms in the second equation and factor, to obtain

ccos(∠ABD) = ccos(∠ABC) ⇐⇒ c2 (a − y) + b2 y = ay(a − y) + x2 a.

Furthermore, (B 0 , C 0 , D0 ) is collinear because d(B 0 , D0 ) = d(B, D) = d(B, C) − d(C, D) = d(B 0 , C 0 ) + d(C 0 , D0 ).

Borrowing the notation from the solution to Problem 10,

d(A, D)2 d(A, D)2

f k (x) = f (f k−1 (x)).

m∠CAD + m∠ACD = m∠ADB → m∠CAD + m∠ACB = m∠AOB = 2m∠ACB

By the Law of Cosines, it follows that

f k (x) = f (f k−1 (x)).

Thus the final answer is 20 + 26 = 46.

(y + x3 )3 − (y + x) = y 3 + 3x3 · y 2 + (3x6 − 1)y + (x9 − x)

perfect cube. Furthermore, 4a + 4b is even, so 4a + 4b = (2x)3√

8x − 8x3 = a3 − 4a. Factoring yields −(2x − 2)(2x)(2x + 2) = (a − 2)(a)(a + 2).

Answer 1B-2. 614

Answer 1B-3. 492

Answer 1C-2. 360

Answer 1D-2. or 2 or 2.25

Answer 1E-2. 176

Answer 2C-2. 168

Answer 2D-2. 952

Answer 2D-3. 714

Answer 2E-2. 120

Answer 2E-3. 24576

T = 41, note that the lengths

Solution 1D-2. Note that U = T

Solution 1. First, note that

You might also like