Hello everyone! We’re a small team working on a 3D Go game called DeepBoardGo.
We’re curious to hear your thoughts, does playing Go in 3D add anything to your experience?
Thank you 
7 Likes
I guess the reason you want to do this would be because of the aesthetics?
Because even “2d” boards sometimes already do try to mimic the aesthetic of real boards with shadows, individual textures for stones, sounds - so I wonder if you should then be going even further to really stand out?
Like physics simulation of stone placement (lightly pushing other stones, dropping them into the bowl). Being able to play in some famous settings, maybe in naturally beautiful settings with weather?
Of course some of that would be a lot of work and kind of a gimmick. Right now it looks nice, but I’m not sure if it’s different enough for me to wish I had that experience over a web go board. Could you maybe share what kind of experience are you’re hoping to create for players that would make it different for them over existing 2d boards?
4 Likes
Thank you for the feedback, Siimphh,
We will consider these features, actually I do like all of them.
>>I guess the reason you want to do this would be because of the aesthetics?
Yes
I guess the experience I’m trying to recreate is like: having a real board (ideally a kaya board ), a nice set of stones, a good book, and about 30 quiet minutes to review a game by players you admire.
I used to do this, but in recent years I’ve found it harder to read game records directly from books/prints. I started using a computer instead, placing it next to my Go board, to review games or play against AI. I actually enjoyed that setup quite a lot.
That made me wonder whether perhaps 80% of that feeling could be achieved with a well-designed 3D Go board. That question is really the main motivation behind this project 
.
3 Likes
I found funny that 95% of the videos content is from the standard vertical view, which can be reached with the 2D we use.
I’m not saying yet that 3D project is useless, I was waiting for something different to see so I can judge for myself if it’s worthy or not?
1 Like
Thanks for the comment Groin, I agree that the top-down view is still the best, even in a 3D environment. What we’ve found is that a very slight angle in 3D can sometimes feel nicer.
We have a YouTube channel with a few videos showing the board from different viewpoints if you’re curious.
Again, totally agree that 99% of the time a traditional 2D Go board is more than good enough
1 Like
3D goban idea pops up every now and then for decades, and here is my reply on reddit r/baduk.
Hopefully, your implementation can bring something new.
2 Likes
I’m surprised the videos feature no stone-placing sound effects, or animations showing the stones hit the table. Both of those seem as important to representing the tactility of Go as showing a nicely-rendered board.
2 Likes
After first reading the thread’s heading I misunderstood…
I thought the “3D” Go idea meant a 3-dimensional game. Such that instead of a square Goban, it would be expanded into a cube.
As such, surrounding & capturing a group would require surrounding it on all sides in an x-y-z coordinate space.
The simplest example: a 6-point surround would be required to encapsulate a single stone.
This would be a bit of a mind bender (at least it would be for me). Kind of like 3D chess as seen on Star Trek and elsewhere. 
[How about working on that project instead of tweaking the visual aesthetics of 2D Go?]
2 Likes
govariants.com has the infra to explore alternative graphs, you’re welcome to contribute there.
I’ve heard that increasing liberties past 4 makes the game a little boring, but there are other 3D lattices that dont have this problem.
3 Likes
Thanks for the comments, guys.
@Counting_Zenist, thanks for the link! A lot of great info there!
@lazerwalker , thanks! yes these are must have features, we do have it but unfortunately not all the demo videos have shown that clearly.
@SomeGoGuy @benjito , thank you for the ideas.
It seems you’re also really interested in true 3D Go, where groups connect in the x, y, and z directions, or even higher dimensions, im interested in that as well, but I’ve never been brave enough to challenge myself with 3D or higher-dimensional Go yet. That’s definitely something we could look into implementing.
1 Like
Ah no problem, my comment was more directed towards @SomeGoGuy. A 3D interface for traditional Go is cool too!
1 Like
For the record, I would describe myself as being more “terrified” of 3-dimensional Go than “interested.”
1 Like
Introducing Luna, our lovely kitten, always supporting our project 
She seems to really like the smell of the kaya Go board 
Have a nice weekend~
5 Likes
Yes I would gladly watch other videos showing more on the 3D side, like the view playing game, or reviewing
2 Likes
Streaming/spectating may be where a very nice-looking 3d go board might be most appropriate. For super serious study, it might get in the way, but if you’re just lazily following a game, then a really pretty board would probably add to the value.
Another tangential idea I’m sure is on somewhere in the backlog already - funky boards and/or stones. If you have access to some material libraries, then maybe it would be straightforward to allow for metal or amber or even (why not?) wooden “stones”? On a marble board, or a board drawn on a piece of parchment or cloth or (why not?) a block of cheese? Well, maybe not cheese, but I wonder if there are some combinations that would be cool 
2 Likes
I do appreciate the aesthetics that 3D can bring. I was briefly a tester for a VR Go game, and I thought it was awesome (playing on a desolate planet is kind of incredible). Even though you aren’t targeting VR, I think 3D on pc can be a great experience too, and I would play it if it looked nice enough.
As constructive criticism, the lighting is at the moment pretty flat, giving it that “steam game” look. Because immersion is the core idea, I think you should work on shaders and the lighting model, as well as the design of the lighting and materials of the board.
I suggest:
- using ambient occlusion
- using the Burley or Oren-Nayar shader models, rather than the Lambert model your engine currently uses.
- bump mapping on the board and stones to give it a more 3d texture
I’m sure you are aware of others, but here are some that I’ve played for inspiration
https://ko-fi.com/post/Getting-ready-for-V2-K3K2DYBII
3 Likes
Thank you @siimphh 
Yes, that’s a really good point. For streaming or game reviews, to support that, we have plans for features like SGF editing, real-time win-rate evaluation, move recommendations, and variant suggestions, which should make reviewing more engaging.
At the moment, we have a built-in selection of Go boards and stone sets, covering classic Japanese, Korean, and Chinese styles, as well as metal, wooden, and plastic stones, along with multiple base board colours/textures. We also plan to support importing custom themes in the future.
Completely agree that even the most beautiful equipment can start to feel a bit repetitive after reviewing a few games.
1 Like
Hi @JohnnieDarko ,
Really appreciate you taking the time, and thanks a lot for the detailed and very constructive feedback.
Completely agree with your assessment. The lighting is still a bit flat, and you’re spot on that it can give off a more generic “game engine” look rather than a truly immersive one.
Right now we have some lighting features, such as directional and hemisphere lighting, but there’s a lot to imprve.
Ambient occlusion is on the roadmap, and you’re absolutely right that lighting is a central, if not the most important part of a Go simulation game.
Thanks again for sharing this. We’ll definitely keep iterating on lighting and materials. The apps and games you linked are also really inspiring, we’ll be learning from them :-)
2 Likes
If Swiss cheese, this may introduce a whole new kind of stone capture
2 Likes
