Two Minor Variations

Just a couple minor variants, I’ve been thinking about, and haven’t seen mentioned.

Exhaustive Go

  • No resigning.
  • Subvariant A: No passing unless there are no legal moves.
  • Subvariant B: No passing unless there are no legal moves that do not result in self-atari.

Subvariant B doesn’t exactly change the game, except forcing both players to complete filling the board, and I am not sure if it would change the way players approach parts of the game. It may make territory scoring with a komi unfeasible, but I think should operate normally under area scoring.

Subvariant A’s primary effect is to remove life through seki from the game, potentially resulting in jockeying for position to force your opponent to have to play into various seki positions in the endgame.

Jitter Go

Another very simple variant which adds an additional option to the [Play Stone, Pass, Resign] list: Move unencumbered Stone (a single stone with 4 liberties) into one of it’s liberties.

The question with this variant is really: Is moving a stone ever going to be a better move than playing a new stone, and I am not really sure. Playing another stone at first glance seems to me to generally be much much stronger. Maybe making such a variant interesting requires the ability to move at least small groups of stones together? Then the question becomes: How big you allow the groups to be, and how movement propagates through the group (no breaking apart? slithering like a snake?).

Any thoughts on these. They are very minor variations compared to a lot of others, but I kinda like the idea of them at least. Particularly Exhaustive Go, since I love a filled up Go board, and sometimes find it a bit disappointing when a game ends with half the points empty.


Since subvariant A forces players to eventually self-atari, it does create some interesting end game considerations, involving throwing dead stones into your opponent’s territory.

See No Pass Go at Sensei's Library which has some similar analysis and discussion.

1 Like

Reminds me a bit of Slither, a connection game played on a go board which allows you to place a new stone and move an old one in each move.

There are actually a couple different seki situations with non-self-atari moves where playing any move loses, so technically subvariant B is not identical to regular go, even with area scoring!

(it’s a fun exercise to think of all the different types of seki like this - I can think of 3 examples right now but I’m sure there are more)

1 Like

That is interesting, thanks for the link. I’d say the primary difference with No Pass is I was thinking of normal scoring, i.e. the first person who can’t play a legal move doesn’t lose, they simply pass to the opponent, and the results of the opponent’s play may even result in new legal moves opening up.

“Slithering” was actually what I was originally thinking of, before ultimately paring the idea down to “jittering”. I had started thinking of any group stones of a certain size or lower could slide into an open liberty. The idea being smaller groups have some mobility, while larger groups do not. Then I got into the ideas I mentioned in passing in the original post: how many stones should be allowed in mobile groups, how many liberties should the group have open and still allow movement, 1? 2? % of potential liberties?, and how should this movement work. I then started thinking of explaining this and thought “Whoah, one of the beautiful things about Go is how simple the rules are, this is starting to get a little complex”. That is when I decided that a lot of the potential complexity of movement goes away when you restrict it to single stones with 4 liberties. I am simply not sure if the ability to move single stones like this is ever strong enough compared to adding another stone to the board, to be worth playing.

Without allowing groups of stones to move (which I think is an interesting variant on its own), I can imagine that there are very rarely situations where moving a single stone is better than placing a new stone.

Perhaps to break this imbalance, one can formulate the following additional rule:

  • A new stone may only be placed in the first four moves of the game, or if the number of unencumbered stones is smaller than the number of groups with multiple stones.

So, you can only add new (unencumbered) stones to the board if you have comparatively little of them to begin with.

Additionally, this adds more interesting strategical considerations, since connecting two groups lowers the number of groups on the board, so it may result in not being allowed to place new stones for a while, because you have to shuffle a single stone all the way across the board to connect with another one.

I think the complexity of a game should not be an influence for whether you should design the game. Games are supposed to be fun, and while Go is perhaps exceptionally simple in its rules, there exist simpler games that can be less fun (tic-tac-toe), but more notably there are a myriad of games that are far more complex than “Go with complex rules about slithering”, yet also widely popular. If Magic: The Gathering can be popular, then complexity is certainly not an obstacle.

1 Like

Here’s some diagrams now that I’m not on my phone:

The bottom left one being by far the most likely to show up in a game :slight_smile:


Various hanezeki-like positions would also be forced to collapse.




Also, here’s a position that you shared earlier:

I think under the “Subvariant B” rules, White would be forced to collapse this seki by playing at F4.

1 Like

All those examples are great. I am a beginner, so if nothing else, this post has exposed me to more knowledge of seki positions than I had before :grinning:


It’s funny because “Subvariant B” is in my mind “how you should play Go if there’s a scoring dispute”

1 Like

I would suggest a movement inspired by moves in the game Hive. A stone can “rotate around” the chain it’s a part of.

For instance:


  • The stone marked △ can move to all intersections marked × by “rotating around” the black chain;
  • The stone marked □ cannot move, because moving it would break the connectivity of the black chain.


  • Now the stone marked △ can move without breaking the connectivity. It can move to any of the intersections marked ×.


  • A stone can only rotate around friendly stones, not around enemy stones. In this diagram, stone □ can move to every intersection marked ×, but cannot move to ○ because the path is blocked by white stones.


  • A stone also cannot jump over the edge of the board. Here the stone marked △ can move to any of the three intersections marked ×, but cannot move to any of the two intersections marked ○ because the path is blocked by a white stone and the edge of the board.


  • A stone also cannot squeeze through a diagonal, nor rotate around another group, friendly or enemy. Here the black stone marked △ can move to any of the three intersections marked ×, but cannot move to any of the two intersections marked ○ because the path is blocked by another black group on one side, and another white group on the other side.
1 Like

There is a thread buried somewhere in this forum where @antonTobi explored the variant “no-pass go” in detail to highlight the differences with go as we know it: Weird and wonderful consequences of simple rules

The variant redstone also has no passing, and the endgame gets pretty weird. I think I would enjoy a variant halfway between go and redstone, which could be called “restone with passing”.