There was a thread a while back about alternative tilings for go (viewtopic.php?f=45&t=1163&hilit=hexagon&start=20).

I was thinking about this again, and I like the trihexagonal tiling:


Can you spot the snapback?

This has the nice property that each point has 4 liberties, and it's uniform. Also it seems like 6 stones make life in the corner:

I had to convince myself that your trihexagonal go board is not equivalent the our normal square go board. But of course it isn't. In normal go two neighbouring points don't share a common neighbour in trihexagonal go they do. I like your idea!

Interesting. When are we going to see first trihexagonalevich?

_________________
Signature space for rent

a bit off topic but i worked on this lattice (called the kagome lattice -from a japanese name for basket or something) during my physics day and the geometric properties are crazy ,and as a result the physics properties of the heisenberg model on it are crazy too (the ground state might have residual entropy at T=0 because it remains ennormously degenerated ). it would be very very different from a square lattice even if coordinance is 4 too: for example it is not bipartite ie you cannot divide it into 2 sublattices where no 2 summit on the same sublattice are niegihbors (this is trivial with a square lattice )

in some sense you always have bad shape: a 2 stone chain have only 5 lib instead of 6 in the square lattices. an empty triangle have 7 libs on the square, but a "full triangle" (3 stones of same color on same triangle) have 5 on the kagome.

making eyes in the center would be near impossible: you need at least 9 stones to enclose a single eye and be connected (vs 7 on the square lattice) and 12 stones to be alive in the center

on the other hand seki might be easier to get

_________________
In theory, there is no difference between theory and practice. In practice, there is.

I'd like to try it, but I'm not sure if there's an easy way to do the diagrams. For these I'm adding the stones manually in inkscape (svg editor). It's not too difficult, but it does require editing in another program.



What about this shape?

And you only need 4 stone to be alive in the acute corner (the straight line is enough) : it's a cute corner

In Esmeraldons first diagram can black connect his left group to his lower right group? I believe he can not.

How does that work Tryss? I can't see it.

_________________
When you think your opponent has made a mistake, that's when you should be the most cautious.

The second points from the corner are miai.

_________________
Signature space for rent

The two "1-2" points are miai for separating the corner space
into two eyes. So it's not Benson alive, but alive it is.

-John

Arguably, I would guess any single stone on an otherwise empty board is "alive" by that measure

I don't want to spoil the fun, but this idea has been implemented long ago in Medusa.
You can find the game at Sensei's and it is also featured in R. Wayne Schmittberger's 'New Rules for Classic Games' (John Wiley & Sons, Inc. New York; ISBN 0-471-53621-0).

_________________
cogito ergo erro

I don't know how to render it, but a spherical trihexagoal board would be pretty cool.

_________________
"When you are learning, it is not so important whether you are right or wrong." - Bill Spight

Playing go on other graphs is such a obvious idea that I believe any claims of anteriority are pretty pointless.

I hope factual information bearing on the subject isn't. Medusa isn't 'Go on another graph' either but a fairly exotic variant quietly orbiting the outer fringes of the Golar system. Before anyone accuses me of claiming it's a big deal, we never even made an applet for it at mindsports.

_________________
cogito ergo erro

Smallest (and easiest) way to do this would be an icosahedron,

Wikipedia wrote:

a regular polyhedron with 20 identical equilateral triangular faces, 30 edges and 12 vertices. It is one of the five Platonic solids.

But I think I wouldn’t want to play this … no corners, no edges, this ain’t something for me lowly DDK D’oh. Trihexagonal, not triangular <hits forehead against table>

_________________
My usual first chat line in an online game: “Hello, this is Tom in Germany”
Let me know whether I should add you to my Go circle/list on … Google+ · Facebook · Twitter (not much used these days)

This is interesting; I wonder if there's an advantage in one or the other visualization. It does seem that Medusa has some additional rule variations that make it something different than "go on a different topology", which is what I was interested in

Is anyone willing to try a trihexagonal malkovich?

Mark Berger, who's real name is Richard Kramberger, got the idea to play Go on the triple contacts of a hexgrid, like this:

Why wasn't this translation possible?
Mark called the game "Rosette" and published it in Games & Puzzles Magazine (issue 34).

The point being: it's not always possible to translate Go to just any grid without encountering difficulties. The above 'atari' problem also plays a role on the trihexagonal tiling that is discussed here, That's why Medusa isn't a 'translation of Go' but a game fitting the grid as best as I thought possible.

I never tried 'regular Go' on this grid, by the way:

That's because it came in the wake of Medusa and presented itself as a simplified version. It may be worth a try though.

_________________
cogito ergo erro

You could have revealed that fact six posts back when you said "I don't want to spoil the fun, but this idea has been implemented long ago in Medusa," implying that Medusa was the first. Your claim of anteriority isn't just pointless in this circumstance. It's misleading and grossly misguided.

_________________
finito ludos regula

I intentionally chose a grid with 4 edges per vertex, to keep the tactics somewhat analogous. However, check out this "ladder":