Diplomatic Go

I would like to propose a new Go variant that combines several aspects

  1. Multi-color Go: the game involves 3 or more players, each playing with a unique color of stones.
  2. Simultaneous Moves: instead of taking alternating turns, each player privately submits sealed moves to an arbiter, and the moves are executed simultaneously. This requires considerations of how to address collisions and simultaneous capturing (see mechanics discussion below).
  3. Negotiation and Collusion: before each round of moves, ample time is given for players to discuss and negotiate publicly or in private one-on-one sessions. They may collude and make promises about their future moves and intentions, but no agreements or alliances are binding.

Anyone interested in playing?

Please reply below if you are interested in playing a pilot game. I will serve as the non-playing game arbiter, since I think we will need to use one to help run things. I’m thinking that this could best be implemented as a correspondence game, with maybe 24-48 hours for each round (including negotiation), and be played here on these forums.

Please let me know what you think about the rules below, and if you have any suggestions about the logistics/timing.

Rules Details

Just some initial ideas, but open for critique, suggestions, and amendments.

Basic Rules
  1. Play is on a standard go board. We will use 9x9 for the initial pilot game.
  2. There are 3 or more players, each using a unique color of stone.
  3. The game will be played here on these forums, in another dedicated thread that will be created, and via private messages for private discussion between the players and arbiter.
  4. The game involves multiple rounds of play, where each player may (attempt to) place a single stone in each round.
  5. If a player feels that another has broken the rules, please only contact the game arbiter who will work on how to resolve the issue and serve as the judge to settle any disputes about conduct.
General Order of Play
  1. Each round of play lasts 24 hours. If a player has not submitted moves by the deadline, the deadline may be extended by 24 hours. If this occurs, it will be publicly announced by the arbiter and all players can continue discussion and potentially change their moves for that round. Each player starts with 3 “extra days” which are used used up every time they go over the normal deadline, and an extension will not occur if none of the players that missed the deadline have any extra days left.
  2. A player may choose to intentionally use a “extra day” to postpone the deadline (even if they have already submitted a move) by informing the arbiter (via private PM, to withdraw their move if they have already submitted one) before the normal deadline. Note: the deadline in each round may only be postponed once.
  3. During each round, players can discuss, negotiate, and plan their moves, intentions, alignments, etc. See the “Communication” section below.
  4. Each player should privately submit their move to the arbiter (via private message) before the deadline runs out. Failing to submit a move (after the extended deadline, or normal deadline if it is not extended) will be interpreted as submitting a pass.
  5. Players can change their submitted move up until the deadline at the end of the round.
  6. After the deadline runs out, the arbiter will work out and reveal what happened on the board from the moves submitted by the players, and then the next round will begin. See sections below on move submission and capturing mechanics.
  7. Starting in the fourth round, any player that does not have any stones on the board will be eliminated and must stop participating in the game (including playing moves and any discussion). Eliminated players are considered as having lost and cannot be part of any draws.
  1. During each round, players can discuss publicly in the game thread and/or privately via private messages with other players.
  2. Private messages should only be one-on-one, i.e., no private group chats between more than two players, and one should not add another player to any of these private channels.
  3. Each player will also have a private message with the game arbiter for submitting moves and asking rules questions that they do not wish to publicly ask in the game thread.
  4. All game related discussion should be confined to these official channels, in order clearly distinguish the boundaries between the game and other interaction in these forums.
  5. Players can make promises and offer alliances, but nothing spoken during these discussions is binding on what moves they submit.
  6. All communication about the game should be confined to the official game thread and private message channels.
  7. Please keep communication as civil as possible (the general forum rules still apply).
Submitting Moves and Resolving Collisions
  1. When submitting a move, each player may either submit a pass, or submit a board play.
  2. Submitting a board play involves specifying the coordinates where one wishes to place a stone and a list of contingency placements, should the first choice fails due to a collision.
  3. Players may submit up to 3 different choices, clearly indicating first, second and third choice.
  4. The arbiter will first attempt to place a stone for each player at their first choice.
  5. If two or more players pick the same location with their first choice, none of them get to play at that location, and the arbiter will attempt to place a stone for the colliding players’ second choices.
  6. If those second choices collide with each other or stones previously played as other player’s first choices, those second choices are not played, and the process similarly moves onto their third choices.
  7. If all of a player’s choices collide with those of others, they do not get to place a stone that round.
  8. Player’s second and third choices will not be revealed if any earlier choice successfully placed a stone. However, any collisions that occur will be revealed.
Capturing Mechanics
  1. As a prerequisite, the board plays, collisions, and contingencies are resolved according to the section above, which results in a set of newly placed stones.
  2. The newly placed stones and any chains (of the same color) that they connect to are called “new chains”. The other chains are called “old chains”.
  3. First, any old chains without liberties are removed.
  4. Then, any new chains without liberties are removed.
  5. Note that this allows suicide and for some cases of simultaneous capture.
  6. After resolving captures, the arbiter will reveal the resulting board state, along with the information of where and by whom new stones were placed (to clarify cases where newly placed stones were immediately captured), and the where/who/when of any collisions that occurred.
Ending the Game
  1. Along with submitting a move, in the messages to the arbiter, a player may choose to vote for a “Resign” and/or “Draw”. When submitting a “Draw” vote, one may optionally specify player(s) that they believe should be excluded from the draw (none by default, and this list can include oneself). Eliminated players are always automatically excluded from any draw proposals. Any votes cast will be made revealed at the beginning of the next round. Players may discuss, even publicly, how they intend to vote, but such discussion is non-binding and not official. Votes must be cast with the arbiter to count.
  2. If all players vote “Resign”, everyone is declared a loser.
  3. If all but one players vote “Resign”, the one that did not is declared the winner and everyone else is declared a loser.
  4. If every player submits “Draw”, with unanimous agreement on who is to be excluded (if any), those not excluded will be declared drawn winners, and the rest are declared as losers.
  5. The previous 3 rules are in order of precedence. If none of the above are ever triggered, the game may instead end by mutual passing or repeated cycling.
  6. A mutual pass occurs when every remaining (non-eliminated) player directly submits a pass (rather than being forced to pass due to collisions).
  7. Upon the first occurrence of a mutual pass, the game enters the agreement phase, where every player should indicate publicly which (if any) stones on the board they think are dead and should be removed. If there is unanimous agreement, the dead stones are removed from the board and the game proceeds to scoring. If there is any disagreement, play resumes with a new round beginning.
  8. Upon the second occurrence of a mutual pass, the game enters the scoring phase with no additional agreement phase, or removal of any stones (i.e., all stones remaining on the board after a second mutual pass are treated as living).
  9. The game also immediately ends and proceeds to scoring (without removing any dead stones), if after the resolution of a round of play, the board position is the same as it was after three prior rounds of play (i.e., the whole board position was produced thrice before).
  10. Each player’s score is calculated (after removing the dead stones, if agreement was reached) as their stones on the board plus any empty territory surrounded by only their stones (i.e., area scoring). No komi is applied.
  1. The goal is to win by either having a score larger than all of the other players, or by having everyone else concede to you via a resign.
  2. If one player has the highest score, they are declared the sole winner, and the rest of the players are all declared equal losers.
  3. If two or more players are all tied in having the same highest score, they are declared drawn winners, and the rest of the players are all declared equal losers.
  4. Basically, there should be no playing for a “strong second” (in score), while letting someone else win.
  5. Winning is preferable to being part of a draw, but a draw is preferable to losing.

Index of Games


Indeed, looks like Diplomacy with go mechanics. I’m interested in multiplayer free-for-all diplomatic “game theory” about optimal negotiations.


Count me in :slight_smile:


Would it be good if we just made this post in this thread?

I mean it is for variants and this is one.Or should I just make a link to this thread there?

1 Like

No, it wouldn’t be good. Let’s keep discussions separated for each variant, it’s difficult enough to keep the overview as is.


Just a couple of comments on the design of the collision resolution mechanics, and some of the potential strategic options.

  1. A player can submit placements that collide with already existing stones (those that were already there at the beginning of the round). However, that will automatically collide.
  2. A collision of first choices, does not prevent someone else’s second choice from playing at that location.
  3. The contingency placements should be unique (within each player’s list), and duplicates are simply removed from the list.

An example

Alice submits the move:

  1. E4
  2. B3
  3. C3

Bob submits the move:

  1. E4
  2. E4
  3. B3

Charlie submits the move:

  1. F1
  2. E4
  3. C3

Suppose all of the points named above are empty at the beginning of this round, except for F1

Here is how it gets resolved:

  1. Since Bob’s list contains a duplicate, it gets simplified to 1. E4 2. B3 (and no third choice contingency).
  2. Alice and Bob collide at E4 for their first choice, while Charlie’s first choice collides with a preexisting stone at F1. Thus, none of the first choices are placed.
  3. Moving to the second choice, Alice and Bob collide at B3, so neither of them gets to play that stone. However, Charlie’s second choice of E4 is placed.
  4. Bob does not have a third choice, so he ends up not placing a stone this round.
  5. Alice gets her third choice at C3, since Charlie has already gotten his second choice (and thus does not cause a collision with his third choice, which does not need to be executed).

Analysis of why these moves might have occurred:

  • Charlie may have learned (from discussion) that Alice wanted to play at E4 as her first choice. Charlie would like to play at that point instead, and does not want to simply block her with collision, but actually get a stone there.
  • Charlie may have colluded with Bob to set up that first choice collision, in order to then sneak into that spot via his second choice.
  • Bob submitting E4 for both his first and second choice may have been an error. His move is effectively 1. E4 2. B3. I just added this as an example to explain that submitting duplicates is not a strategic option.

Interesting idea, not sure I can free some time to join but sure i’m not a candidate to be arbiter!
From my experience of the game “diplomacy” itself it’s a hard game if you get too much involved, the rules have to be clear and simple for no future argue. The resolution and play time has to be exposed in public.

The size of the board may vary with the number of players? Like a 13x13 if more as 5 players?


Count me in to play too! I’ve only read through the rules once but it all sounds quite reasonable. Given how much thought you’ve obviously put into specifying the details of the playing mechanics it feels comforting to have you as the arbiter in the first game @yebellz :blush:


Yes, possibly 13x13 may be necessary to avoid feeling too cramped with 5 or more players. I think even with as few as 3 players it would be nice to try 19x19 to have more room, but I’m just suggesting to start with 9x9 as a pilot game to work out some potential issues and understand logistics.

Glad to have you on board, @le_4TC!


if this is correspondence, i’ll in. sounds interesting :slight_smile:


I won’t be playing, but I can offer some links as primers on strategic diplomacy:



If it’s a real pilot, perhaps something even smaller, like 5x5 or 6x6 could be good to test the rules. Otherwise, I’m ready to do a 19x19, but we should be aware that it might take over a year to finish the game.

Or the discussion could be done on something like discord, live, while talking, that will speed it up a lot.


Ok, so it looks like we have three players signed up:

  1. @Vsotvep
  2. @le_4TC
  3. @Haze_with_a_Z

I’m fine with doing any size of game and changing the rules and parameters, if the players want something different.

  1. What board size do you want to play? I’m fine with doing 19x19, but it will involve a long time commitment from the players, so please keep that in mind. As for sizes on the smaller end, I think 9x9 would be close to the minimum without making for a very distorted game. Maybe 7x7 is possible if people want a very short pilot game, but even though it would be a pilot game, I would still want it to be interesting.
  2. What do you want for time controls? Is 48 hours per round okay? Faster? Slower? I think it would have to be a multiple of 24 hours to make it easier to adjudicate on a schedule, but I could do as fast as 24 hours per round if that works for everyone. Another option is to have flexible rounds that are normally 48 hours, but each round could be cut short to 24 hours if everyone unanimously agrees in public discussion within the first half of the round.
  3. What discussion channels do you wish to use? I’m fine with whatever (like Discord as suggested by Vsotvep) as long as it’s all been agreed by all of the players beforehand.
  4. Any other suggestions, changes, concerns to the rules?
  1. I think 9x9 sounds good for a 3 player pilot, but I would also be fine with 7x7 or 11x11. Smaller seems too cramped and bigger would take too long for a pilot I think.
  2. As fast as possible would be preferable for the pilot I think. How about this: 24h/move by default, but if some player hasn’t submitted a move in time, we postpone the deadline by 24h (and this is announced to all players, who can still change their submissions). Each players starts with 3 “extra days” which are used up every time they go over the normal deadline (kind of like byo-yomi). If you have no extra days left, not submitting counts like a pass. This allows for some flexibility if life gets in the way, but still keeps a steady pace of 1 move per day for most moves. (a simple 24h or 48h/move would also work for me)
  3. Either discord or this forum sounds good to me, with a slight preference for discord.

I really like this idea. Let’s do that if it works with @Haze_with_a_Z and @Vsotvep as well. Does this timing work for you two?

@Haze_with_a_Z, are you okay with using Discord for the chat?


What colors should the stones be?

In order to avoid bias, maybe no one should use the traditional black or white? How about these colors as a palette? Are there any color-blindness issues with these choices?



Any other suggestions for the colors? Which color do you want to play as?

For implementing the board, since this is a correspondence game, I will simply edit the attached SVG and post it each round to the game thread.


Red, yellow, green & blue as shown above (maybe green in a somewhat darker version).
And then: who is afraid of red, yellow, green & blue?
I bet you all make a beautiful painting of the game!

1 Like

I agree with @le_4TC on each point. My idea for discord was more to play the game in a live speed, with all players and arbiter present, such that discussion time could be 10 minutes or so. But I agree that it would be nicer for chat in general, I think.

What I find interesting, is that rank disparity will be a smaller problem than in normal go games, as it is good strategy to first makes sure everybody agrees on the current board situation before trying to make diplomatic pacts with anyone.

It would be lovely if we could find a fourth participant.

About colourblindness, I don’t have it, but this site is nice if you want to test things for visibility. I’d like to play with a darkish blue, not too saturated, like #274786



I’ll let the players choose their individual colors. My palette is just a suggestion.

Are you interested in joining the game, @Atorrante?

It would be great to have a game played live, and I highly encourage that. However, I doubt that I would personally be able to arbitrate or participate in a live game, simply due to the time requirements. Even on a very small board, 10 minutes per round would most likely last many hours and probably require multiple sessions over several days.


I see no problem with this :slight_smile:


No, interested in the concept, but no correspondence for me anymore.