I’ve been developing a Go variant that uses two or more boards simultaneously, designed to model multi-front strategic dynamics (originally inspired by geopolitical thinking, but playable as a pure abstract game).
Core rules:
Two or more boards played simultaneously by two players — suggested size is 7×7, but boards can be of different sizes, which naturally creates asymmetry between fronts
Each turn, a player places one stone on any board of their choice
Standard Go rules apply on each board independently
Capture transfer: when a player captures stones, they face a choice — keep them for end-game scoring, or immediately redeploy them as their own color on any other board
Transferred stones do not count toward the capturing player’s score
Scoring: sum of territory across all boards — highest total wins. No komi when boards are equal in number and size. In asymmetric configurations (different board sizes or number), komi should be calculated accordingly.
Design intent:
The choice of which board to play on is already a strategic decision — allocating attention is itself a move. The capture-or-transfer dilemma creates interrelation between boards: an advantage on one front becomes a resource that can be projected elsewhere, at the cost of a scoring opportunity.
Important note:
This variant has not been playtested yet. I’m posting here to find people willing to try it and help identify whether the mechanics hold up — particularly whether capture transfer creates interesting decisions or degenerates into runaway dynamics.
Questions for the community:
Is there an existing variant this resembles that I should know about?
Is there a way to implement this on OGS or in another place?
I think it is an interesting idea. The transfer mechanic is a big question for me. Are the stones immediately placed on the other board or is it deferred to the next turn? Can the stones that are transferred be used to capture as well? Is this is chain-reaction the runaway dynamics you mention? If I capture two stones, does that mean I can used them to capture a two-eyed group?
Another variation could be that the stones don’t change color when they are transferred and they can be used to reduce your opponents territory or even commit suicide.
Thanks for the questions, you’ve identified exactly the right pressure points.
On the first point, yes, stones are immediately placed on the other board(s)
Your second suggestion is creative, and thinking about it more, it could work with one constraint: transferred stones cannot be placed inside the eyes of an already-living group. Groups with two eyes are protected from sabotage; everything else is fair game.
So two variants could emerge:
Version A: captured stones change color when transferred, reinforcement
Version B: captured stones keep their original color, but cannot target living groups, sabotage with limits
Both feel coherent with the original design intent…
Shouldn’t be too hard to implement it at govariants.com, two or more disconnected boards for one game are already possible, additional moves after captures just require to change whose turn it is (unless you want one move that consists of multiple stones, still possible). Deciding whether to use captured stones immediately or keep them as points would be somewhat new, there’s no mechanic for that yet AFAIK.
If you need help with implementing it there, ask away.
I feel like the capture vs redeploy question is going to lean heavily toward redeploy. The earlier you use stones, the more they are worth. (For example, think about the difference between 9 stone handicap vs 9 points komi)
@Jon_Ko Thanks a lot, I’ll check the link and read through the information. It’s encouraging to know the multi-board structure is already supported, and the keep-or-transfer mechanic being new is actually interesting, it might be the most distinctive part of the variant (if it works)
@benjito thank you! You’re probably right. Early stones are worth more, and rational play may consistently favor redeployment. Only play will tell this and other possible implications
Actually no, I wasn’t inspired by them — I’ll look them up, thanks for the reference.
The starting point was geopolitical interpretation. I’ve long used Go as a lens for reading international dynamics — it maps surprisingly well onto strategic competition between states. But one part always felt incomplete: how different domains like technology, trade, defense and diplomacy interact in a relationship between two or more countries. A country doesn’t play on one board — it plays on several interconnected ones simultaneously, and an advantage in one domain can be projected into another.
That’s the idea behind the capture transfer mechanic. And it’s also why I imagined multiple boards — possibly of different sizes. A larger board for a domain that carries more weight, a smaller one for a less critical front. The asymmetry between boards would already encode something meaningful before the first stone is placed. But already with two symmetrical boards I cannot forsee the possible implications.
The rest followed from trying to keep it as close to standard Go as possible — because Go is already so elegant that it handles most of the complexity on its own
With area scoring (e.g. Chinese rules) redeploying should always be more worthwhile, unless all boards are full and the only option is suicide.
I think komi is still necessary for equal chances. Having one stone on one board is better than no stone on any board (that’s why black makes the move instead of passing). In ideal play, white can always just catch up, never get ahead without komi.
Quick rules question: Are chain reactions allowed? I.e. capture on one board → redeploy on another board and capture something by doing so → redeploy etc. … all in one turn?
Territory was specified in OP. Area could still be interesting, but as you pointed out the choice goes away.
I’m wary of rulesets which talk about eyes.
a) Traditional rulesets never mention eyes, though it’s key to the game.
b) Eyes are usually not so clear-cut.
Do we only count 1-space eye? Two-space eyes? Does group with one large space that we would usually deem alive count as two eyes? Seki?
One idea might be to nerf the redeployed stones such that they must have liberties when you place them (even if they are able to capture). Or perhaps that these stones aren’t allowed to capture at all.
@Jon_Ko : You’re right on the komi point. Even with equal boards, black’s first move is an advantage — that doesn’t disappear with multiple boards, it just distributes. Komi remains necessary. The exact value would need to be determined through play.
On area scoring: good observation. With area scoring the keep-or-transfer choice likely collapses — transferring is almost always better. Territory scoring preserves the dilemma, which is probably why it feels more interesting here.
@martin3141 : Chain reactions — good question. My instinct is to allow them. Adding a rule to prevent them feels like a patch, and on 7×7 the opportunity for long chains is naturally limited by board size. But I’d like to see it happen in practice before deciding. But next point could solve the point.
@benjito: the critique on eyes is well founded — “two eyes” is an emergent concept, not a formal rule, and using it as a ruleset element creates exactly the ambiguities you describe. Of your two alternatives, I find the second more elegant: transferred stones cannot capture. It creates a clean distinction between normal stones and transferred stones — one class builds and captures, the other only occupies. It adds a small layer of complexity to the ruleset, but it remains manageable and is probably worth the clarity it brings
I share the excitement. My hope for this variant has always been the same: few rules, genuine depth — the kind of elegance Go itself embodies. Chain reactions could be part of that, or they could break it. I’m not sure, and I’m aware I’m not the most objective judge of my own idea. That’s exactly why feedback and play matter more than my intuitions at this point. Let the game take the shape that actually works
But area scoring and territory scoring tend to provide similar results in most cases, thus I think the scoring method doesn’t matter much, playing stones is more valuable until there are only dame left.
That’s a fair point for single-board play, where area and territory scoring converge in most games. But in a multi-board context I think the difference becomes more significant. With area scoring, every transferred stone has guaranteed value as long as there’s space — which could make redeployment almost always rational. With territory scoring, a transferred stone only matters if it contributes to actual influence — an isolated stone in enemy territory scores nothing. That asymmetry might be enough to keep the dilemma alive. Though I agree it’s hard to know without playing both versions.
I’ll try to set up the game on govariants.com this week — I’ll explore the platform first and if I manage to get it running and anyone is interested in a test game, I’d be happy to try it out.
