Improving Go (Not Really)

My official position is that you can’t really “improve” Go. There might be something in there worth salvaging, but you can’t just tweak some rules and make Go better. That’s not because Go is so great, but because tweaking rules on an existing system like that tends to create vastly horrible results.
With that said, it might be an interesting intellectual exercise to sort of try to graft on the Clockwork Game Design concepts onto Go and see what you get.
Every day I throw down failed game design ideas. Today I thought I’d share one with you guys just to get a little game design conversation going. (With the election and everything, things have been a little slow on that front recently.)
Here’s as far as I got, just to get you guys started.
Some basic ideas for it:
  • 13×13 board, as a starting place. Would scale up or down as necessary. Maybe the board shouldn’t even be square, not sure.
  • Fog of war. Basically my idea was that you get a vision range of 2, but this doesn’t actually make any sense in practice for a few reasons. One is that you shouldn’t be able to place pieces on the perimeter (random), let alone across the board in some random fogged spot. And the second is that at some point (possibly 10-12 moves in) you’re going to just see the whole board – bye bye, hidden information. There may be solutions to these problems, but I don’t know. (I’ll come back to this at the end.)
  • At least 1 piece down already, probably more like 3-4 in a random, non-mirrored configuration (this is to avoid guessing what the opponent is doing in the fog).
  • Grey pieces are down in a mirrored configuration. Grey pieces turn your color when you put a piece next to them. Or maybe they do something different?

A few thoughts I had as I was giving up on this:

– Maybe this could be single player somehow? Like having to do with the grey pieces? Probably not.

Back to the fog of war and the problems with it: funny thing about this is, it doesn’t really work, and one of the reasons it doesn’t work is the “I can just lay down pieces willy-nilly whereever I want with no restrictions” element of Go to begin with.

Anyway, like I said, I make these kinds of failed little concepts all the time and since things have been slow around these parts recently, I thought I’d share this one totally non-working, bad idea with you.

How would you apply the Clockwork Game Design design methodology to Go? Just to review, here are some of the demands:
– No memorized openings/closings
– Some source of hidden information
– Ideally, something that looks like a core mechanism
– The game should be no longer than it needs to be (Go’s pretty long)
I’d love to hear your thoughts.
  • Matthew

    On the hidden information front, I thought about having “vision pieces” that you can play, that can’t be captured and stones can be placed on. The vision would be blocked by you or your opponent placing a stone in the vision area. The idea was that it would add a tension between playing for vision vs territorial control, as well as increasing the number of ways in which you can evaluate the power of different spaces (“I shouldn’t go there, because it would block my vision”). Ideally, being able to block vision would prevent you from gaining perfect information as the game progresses.

    I think that fog of war is very problematic in a game like Go, however, as the the pieces don’t move. So if you lose vision of an area, you haven’t, really, since you still know exactly where everything was when the vision was lost.

    I also think that being able to see every part of the board is more important in Go than in, say, chess. This is because, in Go, you can play in any empty space without limitation. In chess, you must decide on a strategy, and then invest moves to implement it. This means a player can work out what their opponent will do not by trying to guess their strategy based on the state of the game, but just by observing where the pieces are moving.

    In Go, however, there’s no “setup” is required to implement a strategy (since you can play anywhere at any time), so the only way you can predict the opponent’s strategy is to pretend to imagine what you would do in their position, which means having vision of the whole board becomes more important.

  • I should point people to the Dinofarm Games discussion on this post:

  • Max Hospadaruk

    Off the top of my head a game to look at imo would be Fantasy Flight’s Netrunner: Mainframe. Though by no means a perfect game, the core mechanism (surround and claim area with pieces) is very similar to go and I think there might be a few useful ideas in there to tease out?

    Also, kind of a tangent; I wonder what your opinion is of David Sirlin’s “chess 2” variant? Chess and GO share a lot of design challenges and in the short time my friends and I have been messing around with that ruleset we’ve been really impressed. Lessons learned?

    – throw the starting conditions out the window. the board should be drastically different in almost every game, whether that is by the addition of neutral pieces or by literally changing the shape of the board.

    -forced interaction: in chess 2 the midline rush rule does a lot of work to prevent material-focused long-game strategies from taking root; if you dick around for too long they may just move their king up and win. maybe, for a certain board size, if you secure a certain differential of area and the opponent cannot equalize on the next turn, you win outright. Not sure if this is even necessary as go can be more easily shortened than chess simply by reducing the board size.

    -I don’t know that the bidding system will work at all to provide hidden information here. In chess 2 it serves well as *just* the right amount of unpredictability that can still be planned around but prevents extreme the look-ahead contests that often suffocate chess games. Maybe a more experienced go player could tell me, is there some kind of generic disruption along the lines of “capturing piece also gets captured” that players could bid a limited currency against? Like, maybe, “captured player gets to place a fee stone if the win the bid” or something like this?

  • I’m very familiar with Chess 2. I’m not a big fan, and I think it represents the issues with trying to glom new rules onto an existing (especially super-old and cumbersome) system like Chess. The bidding system “works”, but it’s the kind of thing you could slap into any game and it would appear to work. I do like different starting conditions, but all the factions mean is that there’s more openings to memorize. I think there needs to be something more like Chess960. But even in that case, I just don’t think there’s much to salvage with Chess. It’s just a boring interaction multiplied by a million. If I were you, I would move on from trying to adapt these older systems and work on something entirely new.