wu :: forums « wu :: forums - Political slugfest » Welcome, Guest. Please Login or Register. Jan 25th, 2022, 3:42am RIDDLES SITE WRITE MATH! Home Help Search Members Login Register
 wu :: forums    riddles    medium (Moderators: william wu, ThudnBlunder, SMQ, Grimbal, towr, Icarus, Eigenray)    Political slugfest « Previous topic | Next topic »
 Pages: 1 Reply Notify of replies Send Topic Print
 Author Topic: Political slugfest  (Read 5162 times)
ecoist
Senior Riddler

Gender:
Posts: 405
 Political slugfest   « on: Nov 1st, 2008, 9:22pm » Quote Modify

4 libertarians, 13 republicans, and 17 democrats gather to argue their political philosophy.  They wander about and debate each other in pairs.  When two of them of different political persuasions debate each other, they become so disallusioned that they both change to the third political persuasion.  Show that it cannot happen that, after awhile, all of them acquire the same political philosophy.
 « Last Edit: Nov 1st, 2008, 9:39pm by ecoist » IP Logged
towr
wu::riddles Moderator
Uberpuzzler

Some people are average, some are just mean.

Gender:
Posts: 13730
 Re: Political slugfest   « Reply #1 on: Nov 2nd, 2008, 8:29am » Quote Modify

Reminds me of a type of lizard.. Let's see if I can find a link

Ah, here we go, and more to the point, here.
 IP Logged

Wikipedia, Google, Mathworld, Integer sequence DB
Hippo
Uberpuzzler

Gender:
Posts: 919
 Re: Political slugfest   « Reply #2 on: Nov 2nd, 2008, 2:09pm » Quote Modify

let = e2i/3 be primitive 3rd root of unity.
Look at (l,r,d)=(l+r2+d)

Operations are additions of (-1,-1,2),(-1,2,-1) or (2,-1,-1) to (l,r,d).

As addition of (-1,-1,-1) does not change , additional (0,0,3),(0,3,0) or (3,0,0) does not change (mod 3).

Starting position (4,13,17) is mod 3 equal (1,1,-1) having = 1 (mod 3).
(34,0,0) = .
(0,34,0) = 2
(0,0,34) = 1

I am not able to finish the proof.

So try to go to 34d.
1) (4,13,17)+=(6,-3,-3)=(10,10,14)
2) (10,10,14)+=(-10,-10,20)=(0,0,34).

At least I have prooved noone else will be able .

(12,13,14)=-+ 1 = i3 (mod 3) and (39,0,0)=(0,39,0)=(0,0,39)=0 (mod 3).
(13,15,17)=-1=-i3(mod 3) and (45,0,0)=(0,45,0)=(0,0,45)=0 (mod 3).

If two of l,r,d are equal (mod 3), we can equalize them as in 1). Than convert to the remaining kind as in 2).
If l,r,d are distinct mod 3 we get (l,r,d) = i3(mod 3) and (l+r+d,0,0)=(0,l+r+d,0)=(0,0,l+r+d)=0 (mod 3).
 « Last Edit: Nov 2nd, 2008, 3:02pm by Hippo » IP Logged
towr
wu::riddles Moderator
Uberpuzzler

Some people are average, some are just mean.

Gender:
Posts: 13730
 Re: Political slugfest   « Reply #3 on: Nov 2nd, 2008, 3:04pm » Quote Modify

on Nov 2nd, 2008, 2:09pm, Hippo wrote:
 So try to go to 34d. 1) (4,13,17)+=(6,-3,-3)=(10,10,14) 2) (10,10,14)+=(-10,-10,20)=(0,0,34).
That bodes well for Obama
 « Last Edit: Nov 2nd, 2008, 3:05pm by towr » IP Logged

Wikipedia, Google, Mathworld, Integer sequence DB
Eigenray
wu::riddles Moderator
Uberpuzzler

Gender:
Posts: 1948
 Re: Political slugfest   « Reply #4 on: Nov 2nd, 2008, 3:53pm » Quote Modify

For the general case, the analysis is simpler if they're allowed to borrow people.  For example, 3 libertarians left to themselves will stay libertarian.  But after a couple rounds with a republican in the room (who then leaves, still a republican), they might just find themselves all democrats.

Find a simple criterion to determine whether one state can turn into another without borrowing.

Hippo: I also thought roots of unity were the right way to go.  But they don't really work for a composite number of parties.  That is, with n parties, and a given number of people, there are nn-1 distinct states (allowing negative people), by looking at all the differences mod n.  But []/(n) has size n(n), where = e2pi i/n is an n-th root of unity.
 « Last Edit: Nov 2nd, 2008, 4:08pm by Eigenray » IP Logged
ecoist
Senior Riddler

Gender:
Posts: 405
 Re: Political slugfest   « Reply #5 on: Nov 2nd, 2008, 5:27pm » Quote Modify

I screwed up, guys!  There is an elementary solution if the total number of people involved is a multiple of 3.  I should have said there are only 3 libertarians.

Thinking about what Eigenray wrote, I came up with the following variation.  Let there be 15 parties with the i-th party having i members, for i=1,...,15.  Whenever 14 of these guys meet, no two belonging to the same party, they all switch to the party none belong to.  Show that it cannot happen that after awhile everyone belongs to the same party.
 IP Logged
Hippo
Uberpuzzler

Gender:
Posts: 919
 Re: Political slugfest   « Reply #6 on: Nov 3rd, 2008, 12:41am » Quote Modify

on Nov 2nd, 2008, 3:53pm, Eigenray wrote:
 For the general case, the analysis is simpler if they're allowed to borrow people.  For example, 3 libertarians left to themselves will stay libertarian.  But after a couple rounds with a republican in the room (who then leaves, still a republican), they might just find themselves all democrats.   Find a simple criterion to determine whether one state can turn into another without borrowing.   Hippo: I also thought roots of unity were the right way to go.  But they don't really work for a composite number of parties.  That is, with n parties, and a given number of people, there are nn-1 distinct states (allowing negative people), by looking at all the differences mod n.  But []/(n) has size n(n), where = e2pi i/n is an n-th root of unity.

Actually I didn't thing about general case and the roots of unity was used in confusing way ... I actualy thought on triangular grid mod 3. I have used 1,, and 2 as a coordinates. You can as well use x,y,z space coordinates and project it to the plane perpendicular to main diagonal (t(1,1,1) line). ... There is 9 positions in the projection mod 3, 7 of them are on projections of axis. Mod 3 just projections of (1,-1,0) and (-1,1,0) are not.

I don't think it can be easily generalised to higher number of parties. At least not from my point of view . ...
Oh yes, it looks well for prime p number of parties, but I have had problems to imagine the mod p operation .

BTW: The greasmonkey eats spaces after escape sequences as . How should it be avoided in posts?
 « Last Edit: Nov 3rd, 2008, 1:18am by Hippo » IP Logged
ecoist
Senior Riddler

Gender:
Posts: 405
 Re: Political slugfest   « Reply #7 on: Nov 11th, 2008, 4:27pm » Quote Modify

Suppose that there are n>2 political parties whose members satisfy two conditions.
a) For any two parties, the numbers of members of each party are incongruent modulo n.
b) If any n-1 people meet, no two of them belonging to the same party, then all n-1 switch to the party none belong to.
Show that it can never happen that everyone belongs to the same party.
 IP Logged
ecoist
Senior Riddler

Gender:
Posts: 405
 Re: Political slugfest   « Reply #8 on: Nov 16th, 2008, 11:44am » Quote Modify

I'm a complete idiot (thx for letting me find out for myself)!  This problem, and its generalization, belongs in easy.  Wowbagger suggested the easy solution, and Hippo noted what was needed to be assumed to make his solution work.

The numbers of members of each party forms a multiset of residues modulo n.  That multiset never changes when party switching occurs (see Hippo's example with (4,13,17).  There are always exactly two parties with the same number of members modulo 3).  The reason for this is that, when switching occurs, it amounts to subtracting 1 from each membership modulo n (adding n-1 members to one party's membership is the same as subtracting 1 modulo n).  Hence, since initially all memberships are incongruent modulo n, they must remain incongruent modulo n; and so the guys can never all belong to the same party.  The only way such a thing could happen is if, initially, n-1 memberships are congruent modulo n.
 IP Logged
 Pages: 1 Reply Notify of replies Send Topic Print

 Forum Jump: ----------------------------- riddles -----------------------------  - easy => medium   - hard   - what am i   - what happened   - microsoft   - cs   - putnam exam (pure math)   - suggestions, help, and FAQ   - general problem-solving / chatting / whatever ----------------------------- general -----------------------------  - guestbook   - truth   - complex analysis   - wanted   - psychology   - chinese « Previous topic | Next topic »