IMO 2012 #3

by Wolstenholme, Oct 29, 2014, 2:29 AM

The liar's guessing game is a game played between two players $A$ and $B$. The rules of the game depend on two positive integers $k$ and $n$ which are known to both players.

At the start of the game $A$ chooses integers $x$ and $N$ with $1 \le x \le N.$ Player $A$ keeps $x$ secret, and truthfully tells $N$ to player $B$. Player $B$ now tries to obtain information about $x$ by asking player $A$ questions as follows: each question consists of $B$ specifying an arbitrary set $S$ of positive integers (possibly one specified in some previous question), and asking $A$ whether $x$ belongs to $S$. Player $B$ may ask as many questions as he wishes. After each question, player $A$ must immediately answer it with yes or no, but is allowed to lie as many times as she wants; the only restriction is that, among any $k+1$ consecutive answers, at least one answer must be truthful.

After $B$ has asked as many questions as he wants, he must specify a set $X$ of at most $n$ positive integers. If $x$ belongs to $X$, then $B$ wins; otherwise, he loses. Prove that:

1. If $n \ge 2^k,$ then $B$ can guarantee a win.
2. For all sufficiently large $k$, there exists an integer $n \ge (1.99)^k$ such that $B$ cannot guarantee a win.

Proof:

So I could only do part $ 1 $, but I'm assuming this would be worth $ 2 $ points at the IMO?

When I first saw part $ 1 $, given that asking yes and no questions corresponds to binary digits and the appearance of $ 2^k $, I immediately saw that converting numbers into binary would be important. Given this motivation, let's proceed with the proof!

For the rest of the proof, given a question $ R $ (where $ R $ is a set) let $ [R] = 1 $ if the answer to $ A $ is yes and let $ [R] = 0 $ otherwise. Also, a number with a line on top of it will be the notation for a string of binary digits. Assume $ N > 2^k $ (if not, the problem is trivial). Consider any $ 2^k $-element subset of $ \{1, 2, 3, \dots N\} .$ Call this subset $ S. $ Identify each element of $ S $ with a different binary string of length $ k. $ Let $ S_i $ for all integers $ 1 \le i \le k $ denote the set of numbers in $ S $ with a $ 0 $ in the $ i $th position (from left to right) of their binary representation. Now, $ B $ ask questions $ S_1, S_2, S_3, \dots, S_k $. After $ A $ answers, let $ a_i = \overline{[S_1][S_2]\dots[S_i](1 - [S_{i + 1}])(1 - [S_{i + 2}])\dots(1 - [S_{k}])} $ for all integers $ 0 \le i \le k $.

Now have $ B $ ask questions $ \{a_0\}, \{a_1\},\dots, \{a_k\} $.

For now, assume that $ A $ answers that $ x \not\in \{a_j\} $ for some $ j $ and that $ j $ is the minimal such number. Then if in reality $ x = a_j $ we would have that $ S_{j + 1}, S_{j + 2}, \dots, S_k, \{a_{0}\}, \{a_{1}\}, \dots, \{a_{j}\} $ were all lies. But this is a sequence of $ k + 1 $ consecutive lies, which is impossible. Therefore we would have that $ x \ne a_j $. We can continue using this method to cancel possibilities for $ x $ until we are left with a set of $ 2^k $ candidates, as desired.

On the other hand, assume that $ A $ answers that $ x \in \{a_i\} $ for all integers $ 0 \le i \le k $. Then one of these must be the truth so we have a set of $ k + 1 \le 2^k $ candidates, as desired.

Therefore $ B $ wins!
This post has been edited 1 time. Last edited by Wolstenholme, Oct 29, 2014, 2:29 AM

Comment

0 Comments

Archives
+ June 2016
+ April 2016
+ March 2016
+ July 2015
+ February 2015
+ June 2014
Shouts
Submit
  • glad to have found a fellow chipotle lover <3

    by nukelauncher, Aug 13, 2020, 6:40 AM

  • the random chinese tst problem is the only thing I read, but I'll assume your blog is nice and give you a shout even though you probably never use aops anymoer

    by fukano_2, Jun 14, 2020, 6:24 AM

  • wolstenholme - op

    by AopsUser101, Jan 29, 2020, 8:27 PM

  • this blog is so hot

    by mathleticguyyy, Jun 5, 2019, 8:26 PM

  • Hi. Nice Blog!

    by User360702, Jan 10, 2019, 6:03 PM

  • helloooooo

    by songssari, Jun 12, 2016, 8:21 AM

  • shouts make blogs happier

    by briantix, Mar 18, 2016, 9:57 PM

  • You were just featured on AoPS's facebook page.

    by mishka1980, Sep 12, 2015, 10:33 PM

  • This is late, but where is the ARML results post?

    by donot, Aug 31, 2015, 11:07 PM

  • "I am Sam"
    "Sam I am"

    by mathwizard888, Aug 12, 2015, 9:13 PM

  • HW$\textcolor{white}{}$

    by Eugenis, Apr 20, 2015, 10:10 PM

  • Uh-oh ARML practice is Thursday... I should start the homework. :P

    by nosaj, Apr 20, 2015, 12:34 AM

  • Yes I am Sam, and Chebyshev polynomials aren't trivial, although they do make some problems trivial :P

    by Wolstenholme, Apr 15, 2015, 10:00 PM

  • How are Chebyshev Polynomials trivial? :P

    by nosaj, Apr 13, 2015, 4:10 AM

  • Are you Sam?

    by Eugenis, Apr 4, 2015, 2:05 AM

  • @Brian: yes, yes I did #whoneedsalgskillz?

    @gauss1181; hey!

    by Wolstenholme, Mar 1, 2015, 11:25 PM

  • hello!!! :D

    by gauss1181, Nov 27, 2014, 12:19 AM

  • Hi Wolstenholme did you actually use calc on that tstst problem :o

    by briantix, Aug 2, 2014, 12:25 AM

18 shouts
Contributors
Tags
About Owner
  • Posts: 543
  • Joined: Mar 3, 2013
Blog Stats
  • Blog created: Apr 3, 2013
  • Total entries: 112
  • Total visits: 35003
  • Total comments: 167
Search Blog