Mafia Go #1: Finished!

fwiffo · **#181**

We've been doing it run-off style, so the folks with the unpopular moves should change their vote.

Kirby · **#182**

OK, I switch to G7.

Joaz Banbeck · **#183**

I'm coming to the sad realization that mafia go as we are playing it is not stable if played rationally. As soon as one team realizes that it has substantially less than 50% chance of winning, the rational play is for the members of that team to make deliberately horrible moves for the opponent. By doing so, they raise their chances back to 50%.
The winning team will of course cluster around some not-so-horrible move. There are then two possibilities: If the teams have the same number of players, then the gamesmaster will decide by coin flip or some similar mechanism, and there is a 50% chance of winning that.
If the teams have not been balanced numerically, there is a 50% chance that the losing team has more players, and can win by sheer voting strength.

Either way, the team that is losing can restore their chances of winning to 50%.

This situation, BTW, is not purely theoretical. It may occur very soon in this game as people start analyzing what happens if black plays G7 and white extends from one stone and then black ataris the other. Unless I have mis-read something, white is screwed. The only rational strategy for white at that point is to nominate 1,1 or similar horrible play for the black move. ( My analysis of the board position could be wrong, but nonetheless, at some point there will be a position that everybody realizes that one side has lost )

Kirby · **#184**

Joaz Banbeck wrote:

I'm coming to the sad realization that mafia go as we are playing it is not stable if played rationally. As soon as one team realizes that it has substantially less than 50% chance of winning, the rational play is for the members of that team to make deliberately horrible moves for the opponent. By doing so, they raise their chances back to 50%.
The winning team will of course cluster around some not-so-horrible move. There are then two possibilities: If the teams have the same number of players, then the gamesmaster will decide by coin flip or some similar mechanism, and there is a 50% chance of winning that.
If the teams have not been balanced numerically, there is a 50% chance that the losing team has more players, and can win by sheer voting strength.

Either way, the team that is losing can restore their chances of winning to 50%.

This situation, BTW, is not purely theoretical. It may occur very soon in this game as people start analyzing what happens if black plays G7 and white extends from one stone and then black ataris the other. Unless I have mis-read something, white is screwed. The only rational strategy for white at that point is to nominate 1,1 or similar horrible play for the black move. ( My analysis of the board position could be wrong, but nonetheless, at some point there will be a position that everybody realizes that one side has lost )

You must be saying this because you're on the white team

In all seriousness, though, I'm for having more clear rules, and some way to pick up the pace. Nobody's even been thrown off, yet.

Joaz Banbeck · **#185**

Kirby wrote:

You must be saying this because you're on the white team

No, this is impartial. I'm sort of neutral about my affiliation now. I mean, what is the point of being in favor of one side or the other when know that the game is unstable, and indeed, unplayable?

topazg · **#186**

Joaz Banbeck wrote:

This situation, BTW, is not purely theoretical. It may occur very soon in this game as people start analyzing what happens if black plays G7 and white extends from one stone and then black ataris the other. Unless I have mis-read something, white is screwed. The only rational strategy for white at that point is to nominate 1,1 or similar horrible play for the black move. ( My analysis of the board position could be wrong, but nonetheless, at some point there will be a position that everybody realizes that one side has lost )

G7 is a fun complicated move, but it isn't that much of a killer - just lots of room for things to go wrong for White. I nominated it because everyone loves a complicated fight. I'm not going to give away what White should do, but for the time being at least I'm going to be genuinely nominating the moves I would play to keep things fun and fighty and, in my mind at least, roughly even.

Kirby · **#187**

Joaz Banbeck wrote:

Kirby wrote:

You must be saying this because you're on the white team

No, this is impartial. I'm sort of neutral about my affiliation now. I mean, what is the point of being in favor of one side or the other when know that the game is unstable, and indeed, unplayable?

I knew what you meant. I was just teasing you

Phelan · **#188**

I was expecting you guys to pick a move at the end of the two days, and carry on with the next, but it seems you haven't, so it falls to me:

Click Here To Show Diagram Code: [go]$$Bc Final Vote $$ ------------------- $$ | . . . . . . . . . | $$ | . . . . . . . . . | $$ | . . , . . b a . . | $$ | . . . c 2 3 . . . | $$ | . . . d 1 4 . . . | $$ | . . . . . . . . . | $$ | . . , . . . , . . | $$ | . . . . . . . . . | $$ | . . . . . . . . . | $$ -------------------[/go]

a (G7): topazg, Marcus, prokofiev, Joaz Banbeck, Kirby
b (F7): Chew Terr
c (D6): Jordus,
d (D5): fwiffo, daniel_the_smith, MountainGo
Not voting:
On symmetries: I'm not consolidating them, no. It is up to the players to pick symmetrical moves or not.

Click Here To Show Diagram Code: [go]$$Bc Move 5 $$ ------------------- $$ | . . . . . . . . . | $$ | . . . . . . . . . | $$ | . . , . . . 5 . . | $$ | . . . . 2 3 . . . | $$ | . . . . 1 4 . . . | $$ | . . . . . . . . . | $$ | . . , . . . , . . | $$ | . . . . . . . . . | $$ | . . . . . . . . . | $$ -------------------[/go]

What now?

Next move will be the most voted in two days, or the first to reach more than 75% of votes(With 10, that's 8 votes).

daniel_the_smith · **#189**

F4 & G5 (and symmetries) seem like the only reasonable moves. I vote for F4.

Marcus · **#190**

This is such an interesting position ... I have to think about what I want White to do.

My biggest thought as White is that :b5:

doesn't really put direct pressure on either White stone. I'm tempted to call for C3 and leave the area for a moment ... is that even reasonable?

I have to consider the possibilities before I actually vote ...

Jordus · **#191**

D4...

*edit* d5....

fwiffo · **#192**

Marcus · **#193**

Ok, upon closer inspection, tenuki is stupid.

I vote for F4.

daniel_the_smith · **#194**

Is d5 a reasonable move at 9k or is Jordus on the black team? :twisted:

Chew Terr · **#195**

daniel_the_smith wrote:

Is d5 a reasonable move at 9k or is Jordus on the black team? :twisted:

D5 is a reasonable 9k first instinct at least. I really need to cull unnecessary ataris from my repertoire. However, F4's better, so it's got my vote.

Kirby · **#196**

F4 ftw.

prokofiev · **#197**

Now the game starts, eh?

I suppose the dan players can figure out which of F4 and G5 is better and then choose that one if they're white and the other if they're black. Then it's up to the rest of us to actually figure out which is better, so we know which team the dan players are on. Or maybe I've spoiled it by saying this?

Edit: I guess I'll play F4 too, so as not to look like I have any original thoughts.

An amusing thought: Various ladders are and have been threatened. I wonder what would happen if one were played out? Would everyone reveal themselves and try to make the other side tenuki before it's too late?

topazg · **#198**

MountainGo · **#199**

All the ataris seem bad here, since White isn't eager to sacrifice either stone. F4 seems like the best move, since it prepares to capture :b1:

in a ladder. (@Phelan: and so that is what I am voting for. ;-)

)

Joaz Banbeck · **#200**

prokofiev wrote:

An amusing thought: Various ladders are and have been threatened. I wonder what would happen if one were played out? Would everyone reveal themselves and try to make the other side tenuki before it's too late?

Yes. That's a particular instance of what I have been saying about the game not being stable. On a board this size, a losing ladder is probably a losing game. So it will crystalize affiliation.

I vote for G5.

EDIT: We may not have to wait for a ladder to compel everyone to come out of their closets. The first vote on ejecting someone should do that. ( I'm assuming 5 vs 5. ) If either team can get all of their members voting together, and the other team does not, then they can reduce the other team by one vote. At that point they have an iron lock on the game: they can force every one of their moves to be a good one, they can force every move by their opponents to be a bad one. And as the next vote comes along, they can decimate the other team. The opposing team has no chance whatsoever.

Mafia Go #1: Finished!

Who is online