Perspectives

Dennis Chen

CQV-1

Π1 Overview
Perspectives is a philsophically heavy unit. It’s more important, in the long run, to be able to detect ideas or
problems that feel like Perspectives, rather than remembering every related theorem under the sun. Having
good prototypes of Perspectives is crucial.
So what is Perspectives? It’s as the name suggests — problems that are easier to solve when you consider
them from another perspective. Rather than just being about finding a systematic manner to solve a problem,
Perspectives is about finding the right problem to solve.
Sometimes this means counting an equivalent quantity that is easier to analyze (Bijections). Other times
this means realizing a complex system with many choices actually mostly consists of easily analyzable,
independent choices and few, if any, constrained choices (Freedom). And still other times you want to
cleverly group items together so something can be systematically counted.

Π2 Bijections
Here we discuss some well-known bijections and combinatorial identities.

2.1 Stars and Bars

Stars and Bars is the answer to the question “how many ways can we give identical things to non-identical
people?” It’s also known as sticks and stones or balls and urns. This is a clever trick often used in lower-level
competition mathematics, and it’s best thought of as a clever bijection.

Stars and Bars. The number of ways to distribute 𝑛 indistinguishable items to 𝑘 distinguishable people
𝑛+𝑘−1

is 𝑘−1 .

Proof. Let there be 𝑘 − 1 dividers and 𝑛 items in a line. Then we distribute the items between each set
of dividers (and to the left of the leftmost divider and to the right of the rightmost divider) to the people
in that order. Note that there are 𝑛+𝑘−1


𝑘−1 ways to do this, and this corresponds directly to the number of
ways to directly distribute the items to the people.

★ ★ | ★ | ★ ★ ★ |★

Stars and Bars for 𝑛 = 7 and 𝑘 = 4.

We present a fairly straightforward application of Stars and Bars with no restrictions.

1
 Example (AMC 10A 2003/21). Pat is to select six cookies from a tray containing only chocolate chip,
oatmeal, and peanut butter cookies. There are at least six of each of these three kinds of cookies on the
tray. How many different assortments of six cookies can be selected?

6+3−1
8

Solution. There are six stars and three bars, so the answer is 3−1 = 2 = 28.

You usually will have some sort of restrictions - most commonly, certain people must get a minimum of
the distributed item. (This is evidenced by how hard it was for me to find an example for stars and bars with
no restrictions.) In this case, allot the items "beforehand" and ignore the restrictions, while starting with less
items than before you took care of the restrictions. This is all fairly abstract, so a concrete example will help.

Example (AMC 8 2019/25). Alice has 24 apples. In how many ways can she share them with Becky and
Chris so that each of the three people has at least two apples?

Solution. We distribute 2 apples to each of the 3 people. So we have 18 apples left and no more
restrictions, so the answer is 18+3−1 = 20



3−1 2 = 190.

There will be times when you need to do another clever bijection to make stars and bars easier. Some
example include having a limit on how many items people can receive, and having a number of items very
close to this limit - in this case, we can think about distributing "negative items" - that is, how far away each
person is from receiving the maximum.

2.2 Choosing Unordered Elements

Alternatively titled ascending numbers. We take a look at two generic examples that cover the section pretty
well.

Example (Ascending Numbers). An ascending number is a number whose digits increase from left to right.
How many four digit ascending numbers are there?

Solution. We choose four distinct digits between 1 and 9, and the order is fixed. Thus, the answer is
just 94 = 126.

Now what if the number is just non-decreasing and can stay the same?

Example (Non-descending Numbers). A nondescending number is a number whose digits never decrease
(but may stay the same) from left to right. How many four digit nondescending numbers are there?

Solution. Let the digits be 𝑎, 𝑏, 𝑐, 𝑑. Then we desire 1 ≤ 𝑎 ≤ 𝑏 ≤ 𝑐 ≤ 𝑑 ≤ 9. But note this is also
equivalent to 1 ≤ 𝑎 < 𝑏 + 1 < 𝑐 + 2 < 𝑑 + 3 ≤ 12. So we pick 4 distinct numbers and match them with
𝑎, 𝑏 + 1, 𝑐 + 2, 𝑑 + 3. There are 9−1+4 = 12



4 4 ways to do this, so the answer is 495.

This can also be done with stars and bars - if we let the "baskets" be the digits 1, 2, . . . , 9 and the "stars"
be the 4 digits we choose. We also get the value of 9−1+4


4 = 495.

2
2.3 Combinatorial Arguments
This subsection is about interpreting algebraic sums combinatorially. We present some famous examples of
combinatorial arguments, most famously the Hockey-Stick and Vandermonde identities.
We start with one of the most basic and fundamental examples of using a combinatorial argument to
prove an algebraic identity, and contrast it with an algebraic approach.

𝑛
𝑛

Example (Binomial Theorem). For any positive integer 𝑛, prove that = 2𝑛 .
Í
𝑖
𝑖=0

Solution. This is the number of ways to choose a subset of 𝑛 elements, as each possible size of subsets
is covered.

The algebraic solution involves noticing that (1 + 1)𝑛 is equivalent to the summation.

𝑛
𝑛

Example (Committee Leader). For any positive integer 𝑛, prove that = 𝑛2𝑛−1 .
Í
𝑖
𝑖=0

Solution. The summation represents the number of ways to select a non-empty committee with a
leader, as we first select the committee then choose the leader. But you can also choose the leader out
of 𝑛 people and for the rest of the people, you can choose whether they are in the committee or not, so
there are 𝑛2𝑛−1 ways to choose a committee with a leader.

The fundamental idea is this: the binomial function 𝑛𝑘 can be interpreted as choosing 𝑘 people out of a


group of size 𝑛. No matter how complex the identity, any combinatorial proof will use this.
You can also use an algebraic pairing argument to prove this identity.

Exercise (Shift 1). Prove, combinatorially and algebraically, that for positive integers 𝑛 and 𝑘,

𝑛 𝑛−1
   
𝑘 =𝑛 .
𝑘 𝑘−1

The meat of the subsection is the following two theorems: Vandermonde’s and Hockey-Stick. Typically
AIME problems that employ combinatorial arguments will utilize one of these two theorems, and it will
often be in a very uncreative and straightforward manner.1
1Take, for instance, AIME I 2020/7 and AIME I 2015/12.

3
Hockey-Stick. For positive integers 𝑛, 𝑘,
𝑛  
𝑖 𝑛+1
Õ  
= .
𝑘 𝑘+1
𝑖=𝑘

1
1 1
1 2 1
1 3 3 1
1 4 6 4 1
1 5 10 10 5 1
1 6 15 20 15 6 1
1 7 21 35 35 21 7 1
1 8 28 56 70 56 28 8 1
Diagram from AoPS Wiki.

The Hockey-Stick identity is named such because it looks like a hockey stick in Pascal’s Triangle.

Proof. Have a particle on the lattice grid starting at (0, 0), and allow it to either move 1 unit right or 1 unit
up in each move. Then note that 𝑛𝑖=𝑘 𝑘𝑖 is the sum of the number of ways to get to (𝑘, 0), (𝑘, 1), . . . , (𝑘, 𝑛 −
Í

𝑘), and that 𝑛+1

𝑘+1 is the number of ways to get to (𝑘 + 1, 𝑛 − 𝑘). But note that to get to (𝑘 + 1, 𝑛 − 𝑘), we


go from a point (𝑘, 𝑖) to a point (𝑘 + 1, 𝑖) and then go straight up, which there is always exactly one way
to do once you get to (𝑘, 𝑖). Thus the two values are equal.

3, 4 4, 4

3, 3

3, 2

3, 1

3, 0
Example for 𝑛 = 7 and 𝑘 = 3.

Vandermonde. For positive integers 𝑚, 𝑛, and 𝑘,

𝑘  
𝑚 𝑛 𝑚+𝑛
Õ   
= .
𝑖 𝑘−𝑖 𝑘
𝑖=0

4
 Proof. Note that this is the same as picking a committee of 𝑘 people from 𝑚 + 𝑛 people, since for every
committee, there is some arbitrary number 𝑖 such that we pick 𝑖 from the group of 𝑚 and the rest of the
𝑘 − 𝑖 from the group of 𝑛.

If you understand the proof of Vandermonde’s, this problem should be very straightforward. If you
don’t, you will understand it after you solve this problem.

Exercise (AIME I 2020/7). A club consisting of 11 men and 12 women needs to choose a committee from
among its members so that the number of women on the committee is one more than the number of men
on the committee. The committee could have as few as 1 member or as many as 23 members. Let 𝑁 be
the number of such committees that can be formed. Find the sum of the prime numbers that divide 𝑁 .

Π3 Pairing
Sometimes counting problems will have a large degree of symmetry or regularity in its structure. This
symmetry can be exploited by grouping (typically pairing) off elements.

Example. What is the probability that, after 5 coins are flipped,there are more heads than tails?

The answer is pretty intuitive: 12 . That’s because the probability of getting more heads is the same as the
probability of geting more tails.

Example. What is the probability that, after 6 coins are flipped, there are more heads than tails?

Clearly the probability is not 12 . But the probability of getting more heads is still the probability of getting
more tails. In this case, we just have to consider the probability we get the same number of heads as tails.
Solution. Let the probability of getting more heads be 𝑝. Then by symmetry, the probability of getting
more tails is also 𝑝.
Note the only other possibility is that the number of heads is the same as the number of tails. This
(6) 5
happens with a probability of 236 = 16 . Thus, 2𝑝 + 16
5
= 1, and 𝑝 = 11
32 .

In this case, a pairing argument did not completely finish the problem, but what was left over was easy
to deal with. This is a common pattern with pairing problems.

Exercise (HMMT Feb. Guts 2011/22). Find the number of ordered triples (𝑎, 𝑏, 𝑐) of pairwise distinct
integers such that −31 ≤ 𝑎, 𝑏, 𝑐 ≤ 31 and 𝑎 + 𝑏 + 𝑐 > 0.

The idea is that the number of triples (𝑎, 𝑏, 𝑐) with 𝑎 + 𝑏 + 𝑐 > 0 is the same as with 𝑎 + 𝑏 + 𝑐 > 0. Thus,
you just need to find the number of triples with 𝑎 + 𝑏 + 𝑐 = 0, which is much easier.
Pairing problems can sometimes take an algebraic taste. I wouldn’t really consider them Perspectives
problems, but do keep in mind they exist.
This next example is the epitome of pairing problems and leads perfectly to the next idea: the Principle
of Inclusion-Exclusion.
Example (AIME 1983/13). For {1, 2, 3, . . . , 𝑛} and each of its non-empty subsets a unique alternating
sum is defined as follows. Arrange the numbers in the subset in decreasing order and then, beginning
with the largest, alternately add and subtract successive numbers. For example, the alternating sum for
{1, 2, 3, 6, 9} is 9 − 6 + 3 − 2 + 1 = 5 and for {5} it is simply 5. Find the sum of all such alternating sums
for 𝑛 = 7.

5
 Solution. Note that any subset 𝐴 not containing 7 can be matched with a subset 𝐵 containing 7, and
further note that 𝑆(𝐴) + 𝑆(𝐵) = 7. Since there are 26 = 64 sets 𝐴 (we can treat the empty set as having
alternating sum 0), the answer is 64 · 7 = 448.

3.1 The Principle of Inclusion-Exclusion

Refer to Appendix A if you do not know set notation.
Even though PIE problems may not invoke the idea of pairing, the proof of PIE itself does. Keep this in
mind as you work through these problems.
You’ve probably heard of “Venn Diagram” problems before, and if you haven’t, here’s an example of it.

Example. 20 students are taking Spanish and 30 students are taking French. If everyone takes at least
one language and there are 45 total students, how many students are only taking Spanish?

Solution. Let the amount of students taking Spanish and French be 𝑥. Then note that there are 20 − 𝑥
students only taking Spanish and 30 − 𝑥 students only taking French. We can add the students in all
three of these groups to find the number of students taking at least one of Spanish or French. Since
there are 45 students, 20 − 𝑥 + 30 − 𝑥 + 𝑥 = 45, implying 50 − 𝑥 = 45 or 𝑥 = 5.

The Principle of Inclusion-Exclusion is about splitting the students into groups depending on the exact
number of classes they take (usually the exact classes they take are irrelevant), and making sure you
count each student exactly once. We note that if we count everybody for each time they are in Spanish
and each time they are in French, then we will "double-count" (count twice) the people in both Spanish
and French. This means that we must subtract the people in both Spanish and French. Fortunately, this
overcounting/undercounting behavior is actually quite predictable.

The Two-Set Case. For sets 𝐴1 , 𝐴2 ,

|𝐴1 ∪ 𝐴2 | = |𝐴1 | + |𝐴2 | − |𝐴1 ∩ 𝐴2 |.

Proof. Notice that you count the elements in exactly one set once, but you count the elements in two
sets twice. Thus, we must subtract the elements in both sets to account for this overcounting.

Let’s take a look at the case of 3 sets.

The Three-Set Case. For sets 𝐴1 , 𝐴2 , 𝐴3 ,

|𝐴1 ∪ 𝐴2 ∪ 𝐴3 | =
|𝐴1 | + |𝐴2 | + |𝐴3 | − |𝐴1 ∩ 𝐴2 | − |𝐴2 ∩ 𝐴3 | − |𝐴3 ∩ 𝐴1 | + |𝐴1 ∩ 𝐴2 ∩ 𝐴3 |.

Proof. We note that if we count 𝐴1 , 𝐴2 , 𝐴3 once, then we count everything in exactly two sets twice.
Thus we subtract |𝐴1 ∩ 𝐴2 | + |𝐴2 ∩ 𝐴3 | + |𝐴3 ∩ 𝐴1 |. This means our current value is |𝐴1 | + |𝐴2 | + |𝐴3 | −
|𝐴1 ∩ 𝐴2 | − |𝐴2 ∩ 𝐴3 | − |𝐴3 ∩ 𝐴1 |, but we have counted every person in |𝐴1 | ∩ |𝐴2 | ∩ |𝐴3 | zero times. So
we have to add |𝐴1 ∩ 𝐴2 ∩ 𝐴3 |, giving us our final result as

|𝐴1 ∪ 𝐴2 ∪ 𝐴3 | =
|𝐴1 | + |𝐴2 | + |𝐴3 | − |𝐴1 ∩ 𝐴2 | − |𝐴2 ∩ 𝐴3 | − |𝐴3 ∩ 𝐴1 | + |𝐴1 ∩ 𝐴2 ∩ 𝐴3 |.

Here’s a quick checkup on the three-set case of PIE.

6
 Exercise (AMC 10B 2017/13). There are 20 students participating in an after-school program offering
classes in yoga, bridge, and painting. Each student must take at least one of these three classes, but may
take two or all three. There are 10 students taking yoga, 13 taking bridge, and 9 taking painting. There
are 9 students taking at least two classes. How many students are taking all three classes?

Keeping in mind how many times we count each number, we can generalize PIE. Let’s say we have 𝑛
sets. Then we add the amount of terms in the individual sets, subtract the terms in 2 sets, add the amount of
terms in 3 sets, and so on. Note that this means the amount of terms in at least that many sets, not exactly.2
The general rule is we add the amount of terms in 𝐾 sets if 𝐾 is odd and we subtract the amount of terms in
𝐾 sets if 𝐾 is even. (See the top of the section for a more formalized statement and proof.)
For example, with four sets 𝐴1 , 𝐴2 , 𝐴3 , 𝐴4 , we have
4
Õ Õ Õ
𝐴1 ∪ 𝐴2 ∪ 𝐴3 ∪ 𝐴4 = 𝐴𝑖 − 𝐴1 ∩ 𝐴2 + 𝐴1 ∩ 𝐴2 ∩ 𝐴3 − 𝐴1 ∩ 𝐴2 ∩ 𝐴3 ∩ 𝐴4 .3
𝑖=1 sym sym

Exercise. Write PIE out for five sets.

This exercise is optional, and you should only do it if you feel like you have a very weak grasp of PIE at
this point.
You may have noticed a pattern here: we are adding intersections of an odd number of sets and subtracting
intersections of an even number of sets. The natural question to ask is, “Does this hold in general, and why?”
The answer to the first question is the general Principle of Inclusion-Exclusion, and the second can be
answered with a pairing argument.

The Principle of Inclusion-Exclusion. Given sets 𝐴1 , 𝐴2 , . . . , 𝐴𝑛 ,

 
𝑛 𝑖
 
Ø  Õ Õ Ù 
𝐴 (−1)|𝑆|+1  𝐴𝑠  .

𝑖 =
 
  
𝑖=1 𝑠∈𝑆  𝑗=1
 
∅≠𝑆∈{1,2,...𝑛} 

Don’t get boggled down by the notation. It just means that, over all combinations of intersections of sets
𝑆 𝑖 , its size if added if the number of sets in a certain combination is odd and subtracted if it is even.
The following proof is very similar to AIME 1983/13, which was presented earlier in this section, and
the main idea is proving each element gets counted exactly once. If you’re more experienced, give proving
it on your own a try.

Proof. We prove that each element is counted exactly once. Say some element 𝑋 is in sets 𝑆, 𝑆1 , 𝑆2 , . . . , 𝑆𝑛 .
For every combination of sets not containing 𝑆 with size 𝑖, there is a corresponding combination of sets
containing 𝑆 with size 𝑖 + 1, as 𝑆 can just be inserted into the former combination. After accounting for
{𝑆}, which does not pair with anything, 𝑋 is counted exactly once as desired.

Π4 Freedom
Consider the following problem.
2If this confuses you, just think of the sum of the sizes of all intersections of 𝐾 sets.
3The summations with “sym“ underneath just means all possible combinations of intersections of sets with a certain size. It is an
abuse of notation.

7
 Exercise. Griffin has ten shirts and seven pairs of pants. How many different combinations of shirts and
pants can he wear?

Solving this problem is very easy: you multiply the number of choices you have for each independent
event. In this case, the answer is 10 · 7. However, sometimes it isn’t clear what the independent events are,
or if there even are any at all. Thus, the goal of this section is to develop an intuition for transforming a
combinatorial situation into a set of independent choices.

Exercise. How many ways can four coins be flipped so that an even number of heads come up?

You might naively solve this problem by adding 40 + 42 + 44 , since there are 4





0 ways to flip 0 heads, et
cetera. The answer should make you a little bit suspicious.
Solve this problem for five coins, six coins, and so on, until you see a pattern.

Example. How many ways can 𝑛 coins be flipped such that an even number of heads come up?

At this point, you might suspect the answer is 2𝑛−1 . It seems that every coin added gives us another
independent event with 2 choices, but it’s hard to see how this directly happens.

Solution. Note that regardless of the first 𝑛 − 1 flips, there is exactly 1 choice for the final flipped based
on the parity of the number of heads in the previous 𝑛 − 1 flips.
If the number of heads is odd, then the last flip must be heads, and if the number of heads is even,
then the last flip must be tails.

As a followup, how many ways can 𝑛 coins be flipped such that an odd number of heads come up? Think
about why the numerical answers are the same.
In fact, this is how you prove the identity
𝑛
⌊2⌋  
Õ 𝑛 2𝑛
= .
2𝑖 2
𝑖=0

(Likewise for the odd version.) This is another example of how combinatorial arguments can be used to
solve algebraic problems.
This next problem is a straightforward Freedom problem: the only thing you need to do is find the
number of independent choices.

Example (AMC 10A 2018/20). A scanning code consists of a 7 × 7 grid of squares, with some of its squares
colored black and the rest colored white. There must be at least one square of each color in this grid
of 49 squares. A scanning code is called symmetric if its look does not change when the entire square
is rotated by a multiple of 90◦ counterclockwise around its center, nor when it is reflected across a line
joining opposite corners or a line joining midpoints of opposite sides. What is the total number of possible
symmetric scanning codes?

Solution. Two squares of the board must be identical if they have the same number.

8
 10 9 8 7 8 9 10
9 6 5 4 5 6 9
8 5 3 2 3 5 8
7 4 2 1 2 4 7
8 5 3 2 3 5 8
9 6 5 4 5 6 9
10 9 8 7 8 9 10

Since there are 10 numbers, there are 210 colorings that maintain rotational and reflectional symmetry.
However, we must subtract the all-black and all-white colorings of the board, so the answer is 210 − 2 =
1022.
This final example is a bit different from all the others in that it is a States problem. The canonical reason
I’m including it is to show how Perspectives can overlap with other topics, but it’s really because I think the
problem is cool.

Example (Andy the Unicorn). Andy the unicorn is on a number line from 1 to 2019. He starts on 1. Each
step, he randomly and uniformly picks an integer greater than the integer he is currently on, and goes to
it. He stops when he reaches 2019. What is the probability he is ever on 1984?

Solution. Note that Andy moves from the range 1 − 1983 to the range 1984 − 2019 exactly once, and
furthermore, this is the only way for Andy to reach 1984. Sincce the chance he reaches each number
1
from 1984 − 2019 is equal in this move, the answer is 2019−1984+1 = 36.

There are a couple derivatives of this problem in the States unit, if you are interested.

Π5 MAT 2021/9
In the spring of 2021, I wrote a problem with Aaron Guo that would eventually morph into MAT 2021/9.
Here, we’ll walk through an exploration of the problem and some related problems.

Example (MAT 2021/9, initial). Alexander is juggling balls. At the start, he has four balls in his left hand:
a red ball, a blue ball, a yellow ball, and a green ball. Each second, he tosses a ball from one hand to the
other. If Alexander juggles for 20 seconds and the number of ways for Alexander to toss the balls such
that he has all four balls in the same hand at the end is 𝑁, find 𝑁.

The problem isn’t terribly hard. A naive states bash, which is the first thing I tried when I wrote this
problem, managed to solve it. The Freedom argument is not hard to see either, so I strongly recommend you
try it on your own before reading the solution.

Solution. Notice that after every odd second, Alexander has 3 balls in one hand and 1 ball in the other.
This means that, after 19 seconds, Alexander has 3 balls in one hand and 1 ball in the other, and on the
20th second he must throw the lone ball into the other hand.
Since he can do whatever he wants for the first 19 seconds, and there are 4 ways to choose which
ball to juggle, the answer is 419 .

This initial version of the problem only incorporated Freedom and was not too interesting on its own.
Fortunately, Aaron stumbled onto this variation (which would become the official problem) and came up
with an ingenious pairing argument that convinced us all to use it.

9
 Example (MAT 2021/9). Alexander is juggling balls. At the start, he has four balls in his left hand: a red
ball, a blue ball, a yellow ball, and a green ball. Each second, he tosses a ball from one hand to the other.
If Alexander juggles for 20 seconds and the number of ways for Alexander to toss the balls such that he
has all four balls in his right hand at the end is 𝑁, find 𝑁.

Notice that this time, all balls must end up in his right hand and cannot end in his left hand.

Solution. Let the number of ways for Alexander to end with all balls in his left hand be 𝐿, and let it be
𝑅 for the right hand. Instead of trying to directly find 𝑅, we are going to find 𝐿 + 𝑅 and 𝐿 − 𝑅.
We already found 𝐿 + 𝑅 when solving the first draft of this problem: it’s 419 via a simple Freedom
argument. Now let’s turn our attention to 𝐿 − 𝑅.
Trying to find 𝐿 − 𝑅 by finding 𝐿 and 𝑅 individually is a waste of time, since if we could’ve done
that, we could just find 𝑅 on its own and call it a day. This means we’re going to have to use some kind
of pairing argument.
Note that any set of tosses that ends with all balls in the right hand must reach the state (2, 2) — two
balls in each hand. By symmetry the number of ways to reach (0, 4) — all balls in the right hand — is
the same as the number of ways to reach (4, 0) while passing through (2, 2). This means that 𝐿 − 𝑅 is
the number of ways to (4, 0) without ever reaching (2, 2).
On every even second, the state must be (4, 0), so Alexander can toss whichever of the 4 balls he
wants. On odd seconds, the state is (3, 1), so he only has one choice: pass the ball to the left. Since there
are 10 odd seconds, 𝐿 − 𝑅 = 410 .
Subtracting 𝐿 − 𝑅 from 𝐿 + 𝑅 gives 2𝑅 = 419 − 410 = 238 − 420 , or 𝑅 = 237 − 219 .

An alternate solution I found half a year after the contest is quite similar.

1. Find the number of ways to get to (2, 2) for the first time after 2𝑘 seconds.
2. Find the number of ways to get from (2, 2) to (0, 4) in the remaining 20 − 2𝑘 seconds. (Hint: By
symmetry, it is half of the number of ways to get to (4, 0) or (0, 4).)
3. Sum it up for 1 ≤ 𝑘 ≤ 9.

This is my favorite Perspectives4 problem because it combines two ideas into one: Freedom for 𝐿 + 𝑅 and
Pairing for 𝐿 − 𝑅, and you need to have the meta-understanding of Perspectives to come up with this in the
first place.
This idea of finding 𝐿 + 𝑅 and 𝐿 − 𝑅 can show up in some unexpected places. With this context, try the
following problem.

Exercise. Let 𝑆 = { 12 , 13 , . . . ,
and for any non-empty set 𝑋, let 𝑓 (𝑋) be the product of the elements of
1
100 }
𝑋. Find 𝑓 (𝑋), where 𝑋 spans all of the subsets of 𝑆 with an odd number of elements.
Í

4Arguably it could be generating functions or states. That’s what I love about this problem: there are so many ways to solve it.

10
Π6 Problems
Minimum is [36 p]. Problems denoted with n are required. (They still count towards the point total.)

“I won’t ever allow you to do anything to abandon

these possibilities and choose death!”
Fullmetal Alchemist: Brotherhood

[2p] Problem 1 (AMC 8 2011/6) In a town of 351 adults, every adult owns a car, motorcycle, or both. If 331
adults own cars and 45 adults own motorcycles, how many of the car owners do not own a motorcycle?
[2p] Problem 2 How many integers from 1 to 100 (inclusive) are multiples of 2 or 3?
[3p] Problem 3 (AMC 10A 2018/11) When 7 fair standard 6-sided dice are thrown, the probability that the
sum of the numbers on the top faces is 10 can be written as
𝑛
,
67
where 𝑛 is a positive integer. What is 𝑛?
[2p] Problem 4 There are 3 distinct six-sided dice, one red, white, and blue. How many ways can the
sum of the 15 faces showing on the three die equal 56, if each die orientation is only considered unique if
the sum of its faces that are showing are unique?
[2p] Problem 5 (AMC 10B 2017/13) There are 20 students participating in an after-school program offering
classes in yoga, bridge, and painting. Each student must take at least one of these three classes, but may
take two or all three. There are 10 students taking yoga, 13 taking bridge, and 9 taking painting. There are 9
students taking at least two classes. How many students are taking all three classes?
[3 n] Problem 6 (AMC 10B 2020/23) Square 𝐴𝐵𝐶𝐷 in the coordinate plane has vertices at the points
𝐴(1, 1), 𝐵(−1, 1), 𝐶(−1, −1), and 𝐷(1, −1). Consider the following four transformations:

■ 𝐿, a rotation of 90◦ counterclockwise around the origin;

■ 𝑅, a rotation of 90◦ clockwise around the origin;
■ 𝐻, a reflection across the 𝑥-axis; and
■ 𝑉 , a reflection across the 𝑦-axis.

Each of these transformations maps the squares onto itself, but the positions of the labeled vertices will
change. For example, applying 𝑅 and then 𝑉 would send the vertex 𝐴 at (1, 1) to (−1, −1) and would send
the vertex 𝐵 at (−1, 1) to itself. How many sequences of 20 transformations chosen from {𝐿, 𝑅, 𝐻, 𝑉 } will
send all of the labeled vertices back to their original positions? (For example, 𝑅, 𝑅, 𝑉 , 𝐻 is one sequence of
4 transformations that will send the vertices back to their original positions.
[3p] Problem 7 We have 7 balls each of different colors (red, orange, yellow, green, blue, indigo, violet)
and 3 boxes each of different shapes (tetrahedron, cube, dodecahedron). How many ways are there to place
these 7 balls into the 3 boxes such that each box contains at least 1 ball?
[3p] Problem 8 (AIME II 2009/6) Let 𝑚 be the number of five-element subsets that can be chosen from
the set of the first 14 natural numbers so that at least two of the five numbers are consecutive. Find the
remainder when 𝑚 is divided by 1000.

11
[4p] Problem 9 (AIME II 2002/9) Let 𝒮 be the set {1, 2, 3, . . . , 10}. Let 𝑛 be the number of sets of two
non-empty disjoint subsets of 𝒮. (Disjoint sets are defined as sets that have no common elements.) Find the
remainder obtained when 𝑛 is divided by 1000.
[4p] Problem 10 (Mildorf AIME) Let 𝑁 denote the number of 7 digit positive integers have the property
that their digits are in increasing order. Determine the remainder obtained when 𝑁 is divided by 1000.
(Repeated digits are allowed.)
[4p] Problem 11 (AIME I 2020/9) Let 𝑆 be the set of positive integer divisors of 209 . Three numbers are
chosen independently and at random with replacement from the set 𝑆 and labeled 𝑎 1 , 𝑎2 , and 𝑎3 in the order
they are chosen. The probability that both 𝑎1 divides 𝑎 2 and 𝑎2 divides 𝑎3 is 𝑚𝑛 , where 𝑚 and 𝑛 are relatively
prime positive integers. Find 𝑚.
[6p] Problem 12 (IMO 1981/1) Let 1 ≤ 𝑟 ≤ 𝑛. We consider all 𝑟-element subsets of {1, 2, . . . , 𝑛}. Each of
them has a minimum. Prove that the average of these minima is 𝑛+1
𝑟+1 .

[6p] Problem 13 (AIME II 2013/9) A 7 × 1 board is completely covered by 𝑚 × 1 tiles without overlap;
each tile may cover any number of consecutive squares, and each tile lies completely on the board. Each tile
is either red, blue, or green. Let 𝑁 be the number of tilings of the 7 × 1 board in which all three colors are
used at least once. For example, a 1 × 1 red tile followed by a 2 × 1 green tile, a 1 × 1 green tile, a 2 × 1 blue
tile, and a 1 × 1 green tile is a valid tiling. Note that if the 2 × 1 blue tile is replaced by two 1 × 1 blue tiles,
this results in a different tiling. Find the remainder when 𝑁 is divided by 1000.
[6 n] Problem 14 (AIME I 2015/12) Consider all 1000-element subsets of the set {1, 2, 3, . . . , 2015}. From
𝑝
each such subset choose the least element. The arithmetic mean of all of these least elements is 𝑞 , where 𝑝
and 𝑞 are relatively prime positive integers. Find 𝑝 + 𝑞.
[9p] Problem 15 (AIME 1986/13) In a sequence of coin tosses, one can keep a record of instances in which
a tail is immediately followed by a head, a head is immediately followed by a head, and etc. We denote these
by TH, HH, and etc. For example, in the sequence TTTHHTHTTTHHTTH of 15 coin tosses we observe that
there are two HH, three HT, four TH, and five TT subsequences. How many different sequences of 15 coin
tosses will contain exactly two HH, three HT, four TH, and five TT subsequences?
[13p] Problem 16 (CMIMC 2018) Compute the number of rearrangements 𝑎 1 , 𝑎2 , . . . , 𝑎 2018 of the sequence
1, 2, . . . , 2018 such that 𝑎 𝑘 > 𝑘 for exactly one value of 𝑘.

12
ΠA Set Notation
We define the intersection and union of a pair of sets, and show how this can be extended to multiple sets.

Intersection. Given two sets 𝐴 and 𝐵, their intersection 𝐴 ∩ 𝐵 is the set of all elements in both sets.

Union. Given two sets 𝐴 and 𝐵, their union 𝐴 ∩ 𝐵 is the set of all elements in at least one of the two sets.

As an example, consider the sets {1, 2} and {1, 3}. Then their intersection would be {1} and their union
would be {1, 2, 3}.
There’s a very handy trick to remember which symbol corresponds to the intersection and which
corresponds to the union: the ∪ looks like a U, and union also begins with U. Thus, by process of elimination,
the intersection symbol is ∩.
We can chain together intersections and unions for more than two sets. Try the following exercise
yourself.

Exercise. Find {1, 3, 5} ∩ {1, 3, 4} ∩ {2, 3, 5} and {1, 3, 5} ∪ {1, 3, 4} ∪ {2, 3, 5}.a
a If you’re the type to be bothered by this, assume the order of operations is from left to right.

It turns out that the intersection of multiple sets is the set of all elements in all of the sets, and the union
is the set of all elements in at least one of the sets. The order of operations also does not matter as long as
you only have intersections or unions.

Exercise. Prove that 𝑆1 ∩ 𝑆2 ∩ 𝑆3 ∩ · · · ∩ 𝑆𝑛 is the set of all elements in all the sets, and prove that
𝑆1 ∪ 𝑆2 ∪ 𝑆3 ∪ · · · ∪ 𝑆𝑛 is the set of all elements in at least one of the sets.a
a To prove that two sets 𝐴 and 𝐵 are identical, it suffices to prove that every element of 𝐴 is contained in 𝐵 and every element of

𝐵 is contained in 𝐴.

13