Thoughts on Pigeonhole + Math awards + New contrib
by shiningsunnyday, May 24, 2016, 1:46 PM
Let's start with some good ol' Pigeonhole Principal problems from Interm CP and 102 shall we?
The Theorem
PHP is just common sense. Consider having
pigeons and
holes. Note that
which implies that if we try to stuff holes with
pigeons each, we'll run out of holes first. If we try to stuff holes with
pigeons each, we'll run out of pigeons firsts. Now because
we must have some holes with
pigeons rounded up. In other words, guarantee
pigeons in a single hole.
So it appears that the most pigeons we can guarantee in a hole is
But wait! What if, say,
Uh oh, clearly
does not apply anymore. Fortunately, we can prevent these cases in which
is a multiple of
by slightly altering our formula to
as desired.
First, a direct application:
Solution
Remark
Sometimes, the boxes aren't so easily identifiable. It's up to you to play around a bit. A good way to do so is to partition your "boxes" into groups. Here's a simple application:
Solution
Remark
Sometimes, the boxes are really tricky to quantify:
Solution
Here's a problem with an almost-identical solution, except with an extra twist.
Once again, a counter-example must be given when dealing with maximums/minimums.
My Solution
Book's Solution
There're a few more nice problems which I planned on adding in this post but I didn't get to solve. Perhaps I'll stop being lazy and do them soon?
So yesterday I received my math certificates for this year. It's attached below. Overall, it sums up the misery of my contest season this year pretty well. Ah, how nice it would be had I gotten that JMO qualification... The Canadian ones are pretty self-explanatory. Nothing worth mentioning, honestly. *Cries*
Next year is the big year, I guess. I didn't expect things to turn out this way. I mean, if I had gotten that JMO qualification, my dad wouldn't be hollering me like I've been lollygagging my way through the whole semester, nothing to show besides a 3.5 GPA. *Continues to rant* I mean, so many 'if's. If... if... if... No one gets me at all. Everything just so terr--
*Suddenly, music starts playing in the background.*
Wait, what's that sound?
Hey AoPS, say somethin' to her, holler at her.
Wait, what's going on?
I got one question.
What?
How does she fit all that math, in her head?
She?
Someone new is contribin' to my big fat blog.
Who?
Wiggle wiggle wiggle.
Wiggle wiggle wiggle.
Wiggle wiggle wiggle.
*Music music music music music music music music*
*Body starts to shake uncontrollably*
Kong-tri-buter kong-tri-buter
Of this blog
Got me dazin' at your math skills all day long
If I ask you to be contrib of my blog
I can make you famous on ey-oh P.S.
She's it, baby
Never takes math for granted
Give her a problem and she's right on it
Woah
I can't stand it
Anyways... Y'all know who's contribin' to my nice cool (wait that sounds weird) blog.
Yadayadayada [clearly not out of ideas as to not make the ending of this blog post look awkward]
Let's give it up to our blog's newest contributor, WIII WII WI WI WIIII
GGGGGULLLLL GULLLL ULLL OHHH GIGGLE? WHAT? GGGGOLLLDDD GOOOO LOL
WAMMMMMMMMMMMMMMMMMMMMMMMMMM!!!!!!!!
The Theorem
The Pigeonhole Principal wrote:
If we want to stuff
pigeons into
holes, then we can guarantee at least one hole contains
pigeons.



PHP is just common sense. Consider having








So it appears that the most pigeons we can guarantee in a hole is






First, a direct application:
Interm CP wrote:
A pen-pal club has 12 members. In August, each member of the club sends a letter to 6 of the other members, chosen at random.
(a) Prove that some pair of members of the club will send each other letters.
(b) In September, each member of the club sends a letter to only 5 of the other members. Does the conclusion from part (a) still hold? Why or why not?
(a) Prove that some pair of members of the club will send each other letters.
(b) In September, each member of the club sends a letter to only 5 of the other members. Does the conclusion from part (a) still hold? Why or why not?
Straight-foward PHP. Note that there're
and
letters sent, respectively. Consider the
members as vertices of a graph. Note that there're
edges. An edge corresponds to a letter. Clearly at least one edge will be used twice in the first case, corresponding to a pair of members sending each other letters, but this is not guaranteed in the second case since 
We need also to build a counterexample, but oops I'm lazy so sorry.





We need also to build a counterexample, but oops I'm lazy so sorry.

Remark
When trying to show that a certain property isn't satisfied, it's not sufficient to just apply Pigeonhole, you have to actually construct a counterexample to complete the proof.
Sometimes, the boxes aren't so easily identifiable. It's up to you to play around a bit. A good way to do so is to partition your "boxes" into groups. Here's a simple application:
P6 of 102 wrote:
Twenty five boys and twenty five girls sit around a table. Prove that it is always possible to find a person both of whose neighbors are girls.
Direct PHP. Partition the
chairs into
groups of
consecutive chairs (there'll be two chairs left over, but I mean, you only live once). Because we have
girls and
groups, we can say for sure one group will get
girls. The only possible ways the
girls can sit in those
chairs are as
each of which satisfy the problem condition, so we're done.









Remark
It's worth noting I only got the solution after playing a bit. I knew from the start the girls are the balls and the boxes will be groups of some number of consecutive chairs. It takes some playing to realize that number should be 

Sometimes, the boxes are really tricky to quantify:
Putnam wrote:
A certain college has
students and offers
courses. Each student can enroll in any or all of the
courses, or none at all (which is a real waste of tuition). Prove or disprove: there must exist
students and
courses, such that either all
students are in both courses, or all
students are in neither course.







I was really happy to realize this problem can be solved in almost an identical fashion as the previous one. Let's define the boxes as pairs of classes enrolled in by a student or pairs of classes not enrolled in by a student. There're
boxes for pairs of classes both of which a student is enrolled in, and
boxes for pairs of classes both of which a student isn't enrolled in. Now consider a random student that enrolls in
classes. The student will fill up
of the
boxes. Now note that
so we can guarantee that every student will fill up at least
boxes, so we can guarantee at the end of the day
balls total in the
boxes. But if we want to guarantee there is
balls in a box, then by the Pigeonhole formula, we must have
balls because
while 
Ok, so now that the hard part is done, we just have to build a counterexample to disprove the statement. Clearly we want each student to enroll in
classes, and we want it so that each of the
boxes get exactly
balls, but once again, my laziness prevents me from doing so. Oops













Ok, so now that the hard part is done, we just have to build a counterexample to disprove the statement. Clearly we want each student to enroll in



Here's a problem with an almost-identical solution, except with an extra twist.
HMMT wrote:
Somewhere in the universe,
students are taking a
-question math competition. Their collective performance is called laughable if, for some pair of questions, there exist
students such that either all of them answered both questions correctly or none of them answered both questions correctly. Compute the smallest
such that the performance must be laughable, no matter how the students performed on the competition.




Once again, a counter-example must be given when dealing with maximums/minimums.
Interm CP wrote:
Sam’s band has 6 members, but only 4 of them play together in a concert. Additionally, no 3 members can play together in more than one performance. How many concerts can Sam’s band give, at most?
Here's my sketch. Let the members be
and
WLOG let the first performance be of members
To make sure no three members repeat, we must sub in
and
before the next concert. If we only sub in one, say
for
the
triplet will be repeated. So the next performance will be a two element subset of
WLOG let the second performance be of
Similarly, the next concert must be contain
It can't have
or
in it cause that will conflict with performance
so we see the third performance must be
If we apply our logic once more, the next performance should feature
But we can't feature any of
because that would conflict with either performance
or performance
so the max we can do is
concerts.




















Book's Solution
So the book apparently likes to make things rigorous, so we must achieve a contradiction with the case of
concerts to complete the proof. Ok. So how are we going to achieve a more rigorous contradiction with
concerts? Will, it makes sense to see specifically how many concerts each individual member can be in. WLOG let's look at person
Person
is part of
triplets, but every time he goes to a concert (which has
performers) he will be featured with
triplets. So clearly person
can't participate in all
concerts, so by the same logic each member can go to at most
concerts. Because there're a total of
individual performances, we must have exactly four people participating in
concerts. If we have only three people participating in
concerts, the total individual performances will be at most
contradiction.
So WLOG let the people who perform three times be
and
Note that the only way for this to work is for each concert to feature a three element subset of
Why? Because if not, then we must have at least one performance featuring all four of
and we can easily see that a triplet will be repeated. We must have
to perform in concert
in concert
in concert
in concert
Now remember,
and
must each perform twice, so
must choose two of the concerts to perform in and
will go to the other two. This is an immediate contradiction since whatever two concerts
participate in, there will be a repeated triplet.














So WLOG let the people who perform three times be














There're a few more nice problems which I planned on adding in this post but I didn't get to solve. Perhaps I'll stop being lazy and do them soon?

So yesterday I received my math certificates for this year. It's attached below. Overall, it sums up the misery of my contest season this year pretty well. Ah, how nice it would be had I gotten that JMO qualification... The Canadian ones are pretty self-explanatory. Nothing worth mentioning, honestly. *Cries*
Next year is the big year, I guess. I didn't expect things to turn out this way. I mean, if I had gotten that JMO qualification, my dad wouldn't be hollering me like I've been lollygagging my way through the whole semester, nothing to show besides a 3.5 GPA. *Continues to rant* I mean, so many 'if's. If... if... if... No one gets me at all. Everything just so terr--
*Suddenly, music starts playing in the background.*
Wait, what's that sound?
Hey AoPS, say somethin' to her, holler at her.
Wait, what's going on?
I got one question.
What?
How does she fit all that math, in her head?
She?
Someone new is contribin' to my big fat blog.
Who?
Wiggle wiggle wiggle.
Wiggle wiggle wiggle.
Wiggle wiggle wiggle.
*Music music music music music music music music*
*Body starts to shake uncontrollably*
Kong-tri-buter kong-tri-buter
Of this blog
Got me dazin' at your math skills all day long
If I ask you to be contrib of my blog
I can make you famous on ey-oh P.S.
She's it, baby
Never takes math for granted
Give her a problem and she's right on it
Woah
I can't stand it
Anyways... Y'all know who's contribin' to my nice cool (wait that sounds weird) blog.
Yadayadayada [clearly not out of ideas as to not make the ending of this blog post look awkward]
Let's give it up to our blog's newest contributor, WIII WII WI WI WIIII
GGGGGULLLLL GULLLL ULLL OHHH GIGGLE? WHAT? GGGGOLLLDDD GOOOO LOL
WAMMMMMMMMMMMMMMMMMMMMMMMMMM!!!!!!!!
This post has been edited 1 time. Last edited by shiningsunnyday, May 24, 2016, 1:49 PM