Redstone at Board Game Arena

If I had to think about it abstractly, the redstones don’t belong to anybody so they shouldn’t necessarily prefer whose stones they capture and don’t capture.

In normal Go, the “partisan” capture makes suicidal moves useful (and legal for captures when they aren’t legal in the ruleset) and allows the capture of groups with one eye so it seems to make sense and add a lot to the game there.

I guess the current way simplifies the capturing mechanism when you just think “stones without liberties are removed” rather than “stones not of the player colour are removed, unless they’re the only stones to be removed (self capture)”.

1 Like

@ArsenLapin1 I don’t remember from last week. Much less ten years ago :sweat_smile:

For better or worse, that’s how Redstone is. I’m not changing it now. I like @shinuito deduction though. Simplicity is always a design imperative.

Btw, I invite you to try Oust. It’s considered my best game. It’s very thoughtful. Players spend considerable time on their moves from start to finish. It’s for intelligent players like you guys. Not for the largest group who like to click buttons and see pretty colors. I play a lot of Cephalopod because it’s mildly entertaining and there are a lot of players. But it’s not in the same league as Oust

1 Like

I may have a clue as to why I chose the obliteration capture rule for Redstone in a departure from Go-like captures. In 2006, I designed Cephalopod to have a capture-from-within mechanism. A consequence of this was a massive churn rate (i.e., ratio of turns to board cells) of about 6 to 1. In 2010, I designed Rive specifically to maximize the churn rate in a two color stones territory game. Rive’s churn rate is also massive, especially if you measure it by number of moves (which include a placement and multiple possible removal choices) per turn per board size. I like a high churn rate because it gives you more game for a given board size. It makes the game richer. In 2012, when I designed Redstone, churn rate was undoubtedly a consideration. Obliteration capture opens up a bigger space, the filling of which is then a game within a game.

As I’ve said, I know nothing about Go play, and I’m sure there are compelling arguments (that I wouldn’t understand) why Redstone should have been more Go-like. But I think I’m fairly satisfied with Redstone’s current design. I’ve reached out to some abstract game experts for their thoughts. I’ll keep you posted… :slightly_smiling_face:


Christian Freeling got back to me pretty quick…

“The nice thing about your games is that they are your games. I’ve thought about the ‘why’ of course but my conclusion is that it would make a different game, but not necessarily a better one. In fact it would make it more ‘mainstream’ from an inventor’s point of view. And who wants to be mainstream?”

In linked article:

Of course, ko no longer exists, which was Mark’s main objective with the game.

Ko is one of the best features, not bugs, of Go. It brings a tremendous interest, excitement, depth, and beauty to the game. Beginners are often intimidated by it, but once they get a bit stronger they begin to appreciate it.


This is a forum for Go players, and nobody here cares what a non-Go player is into or not into. But, that being said, I am not into ko - bigly.

This from my BGA Beta forum Redstone announcement…

“Redstone is naturally finite. That was my imperative in designing it. I know. Go is finite because it has superko. Any game can incorporate superko and call itself finite. Like, “You can never repeat a position because you’re not allowed to repeat a position.” Brilliant. That’s why I said “naturally” finite. I’ve heard that “in practice” finitude is never really an issue in Go. For me, it’s not about “in practice.” It’s about architecture. I don’t design non-finite games.”

Sorry if this is kicking the hornet’s nest.


OK, I can appreciate that objective of naturally finite rules with no bodge like superko. Personally I’ll take the bodge that I’ve never needed in tens of thousands of games for the richer gameplay it brings.

1 Like

This from Kerry Handscomb, editor of Abstract Games magazine and advanced kyu Go player…

“It might seem strange to Go players that you don’t capture enemy stones first, opening up liberties for your own stones. Captured stones score in Go, of course. Maybe that’s where it’s coming from. It makes logical sense to me to remove all stones whose liberties are taken with a red stone placement. They don’t score. The red stone placement results in the instant capture of all groups without liberties. Capturing enemy stones first seems arbitrary. I haven’t looked at how it would change tactics. However, my first intuition is that you made the right choice ten years ago.”

I’m fine with obliteration capture. I appreciate the input and feedback, but… case closed.


Very interesting, thanks for sharing!

Well, Redstone is extremely similar to go in many respects. Which means, when go players play a game of redstone, their intuition and experience in the game of go will undoubtedly help them. And where redstone differs from go, “adjusting” our way of thinking for the differences can be either fun or frustrating.

  • The fact that the capturing stone is going to be neutral and immortal, instead of our colour, is really fun. I enjoy exploring the consequences of that rule.
  • The obliteration rule, on the other hand, feels extremely arbitrary. I don’t particularly feel that one rule would be better than the other, and exploring the “new possibilities” opened by that rule isn’t particularly interesting to me.

This reminds me of a story about a film adaptation of “the three musketeers” which I watched when I was a child. The film took many many liberties from the books. Then at one point, one character who was a duke in the books was referred to as a count in the film. One of my older friends exclaimed: “this is outrageous! how dare they make him a count instead of a duke?”. Somehow, all the major deviations from the book were forgivable and in fact enjoyable; but changing the title of one character felt arbitrary and wrong.


Lol about the duke. Well, no game is perfect. It starts with a coin toss, and Player 2 starts from a different position than Player 1. But I try to make them as robust as I can, luck and asymmetry notwithstanding.

1 Like

I’m not sure one is arguing that the game is better without ko, or that the game without ko is somehow richer. Sometimes simplifying a game can be interesting, like playing capture Go, which also doesn’t have ko - but I don’t think anyone is claiming it’s out and out better.

I agree with your post other than that the obliteration rule seems arbitrary, as I’ve said it feels more natural given that the redstones are neutral. Most rules are arbitrary lets be honest. The Japanese rules of Go are filled with many arbitrary rules and judgements just to make territory scoring work well and fit with how the game has been played for centuries.

Chess is really a big collection of arbitrary rules that have been used to fine tune the game over time and make it interesting (one might argue), e.g. pawns can move two spaces on the first move (to speed things up), en passant, castling only when the pieces have not moved originally and are not in check etc.

It makes sense to me to object to the obliteration rule, sure, if one doesn’t want to go

as of course that’s where learned life and death shapes won’t help necessarily. One can’t transfer all ones Go knowledge across since one can’t say “I know this is a ko” since there’s no ko, or “I know this shape is dead once I do this placement/throw-in” since now an immortal redstone might give an eye where it wouldn’t normally.

I do find it interesting and I am curious about some new tactics, but that’s not saying I prefer it to Go, or that it’s necessarily more complex (if we want that to be a measure of interesting) than Go.

If anything I find the simplicity of the goal (capture all enemy stones), and the slightly simpler nature of the rules (we don’t have to explain territory, seki or neutral area, ko and or super ko etc) somewhat interesting and maybe easier to get into than normal Go.

One might imagine a Go variant like redstone where again the goal is to capture all enemy stones, but doesn’t use redstone for captures. I don’t think it would be exactly like Go with group tax or stone scoring since one will still want to play like

All said though, I think the endgame (non-go like part) can be a bit drawn out, more so than it would be with something like stone scoring. I think that can be a bit of a drawback. On the other hand one could also imagine just playing it quicker than the rest of the game in simpler situations.

1 Like

This is essentially just no-pass go (in both cases the first player that has to “suicide” a group (fill the second-to-last eye) might as well resign in practice - or is this not true of redstone?).

Hopefully I’ll get around to making the post about no-pass go in Weird and wonderful consequences of simple rules soon. The territory-filling stage works almost identically in redstone, I think (the diagram I gave above is a no-pass example that “happens” to work the same way in redstone).


Well that’s precisely what I was questioning above

In some cases one might imagine that when you suicide a group or reduce it to one eye and the opponent captures, that you can again play inside with the redstone and make another alive group.

I will at some point see if I can come up with a silly but working example, but I think I need 3-4 groups.

1 Like

Even if the captured group is such that it is possible to live inside again afterwards, usually the opponent will be able to make more moves than you in the new empty area (meaning that you will again be the first to run out of moves). It would be very impressive if you can somehow construct a counterexample to this!

(just living inside isn’t hard, just make the captured group absurdly big, say a 17x17 dumpling up against the corner)

1 Like

Ok I have a simple but very unrealistic example I think. (I’m not sure it’s what you’re hoping for?)

It’s black to play, and black needs to kill one of their two groups. So black chooses to play J9

White either needs to kill one of their own groups or capture the black corner group but if they do black follows with H9 and makes two eyes again

Now white needs to kill one of their groups.


Now to make it into a real example I could probably make the groups bigger but still with only two eyes (since inside moves would eventually be filled in a real game) but with a more reasonable amount of redstone to separate some of the groups say.

I’m sure it could be difficult to define real, but certainly the above example is a bit unrealistic

1 Like

That does work, yes! I believe it’s not too hard to construct a similar example for no-pass go using a non-standard board (i. e. if you allow arbitrary graphs), but it feels like on a grid board it might be impossible.


Here’s a “real” position with a lot of redstone, since it’s the continuation of the game I played with @ArsenLapin1 where I got very lucky with some of the redstone capturing rules.

So one can imagine groups being separated by redstone, but of course, the space left behind by captures is much more.

It did leave room later to make another living group in the bottom left for Black, but I also took the opportunity to join up my groups.

I can imagine being able to join groups being a plus in some circumstances, unless there’s a special case where sacrificing half gives you a chance to recreate a living group “in gote”, but beneficially :slight_smile:

1 Like

I think I can imagine ways to not have the groups touching also by playing suicidal/self atari like sacrifices in order to get a redstone played instead. It might be very dependent on the board though.

On the first line, imagine removing G9 and F9 in the last diagram and have black play a jump/sacrifice of F9 - separates the groups but gives white an extra eye.

1 Like

So there could even be a collection of Nakade at Sensei's Library type tricks where you kill your own stones in certain shapes but get to live again in gote.

Imagine something like where you can make a bulky five (top right) killing your own stones and then playing the vital point

J7 probably is better than J8 or H7 (locally so as not to gain sente immediately - globally no idea), but black could then play J8 and white captures their own stone with H7 (below first image). But then we get a familiar sacrifice to live again in “gote”