I once made an interminably boring document, full of legalese and clauses and fine-prints just for a forum game, and it was all in jest. So, at the risk of being interpreted as hostile (which I’m not, I swear), the answer would be “Not really much, no 
”
3 Likes
I’m wondering how one can even debate whether Go is better than chess in terms of game design without debating which ruleset of Go one uses and which one is best at some point…
I don’t consider this topic to have derailed yet, so I’ll do the same as CM and change the topic title to reflect my personal opinion on the state of this topic.
5 Likes
Minor note here:
If there is a single ko or double ko that is a mistake to keep fighting (i.e. there are much more urgent valuable moves outside of resolving the ko) but both players keep “stubbornly” maintaining the cycle, would you annul the game? No, that would be absurd. Why is it that at the magical number of 3 kos a game should be anulled? Purely convenience, not fairness.
Drawing the line at 3 isn’t entirely arbitrary, because a single ko and double ko usually can’t be stubbornly maintained under a simple ko rule. Both of them forcibly require the players to spend turns doing something other than repeating the stones that are part of cycle, thus usually progressing the game.
And really, that isn’t even the rule, it’s not “annul for triple ko”, rather it’s more just “annul for any cycle at all that simple ko doesn’t prevent and neither players wants to give up”. Triple ko is merely one of the multiple ways that can happen, but it’s not like the rules need to specially call out that case. So if you’ve chosen simple ko as the rule in the first place, then in fact there isn’t much arbitrariness at all beyond that.
Of course, the choice of simple ko itself could be considered arbitrary enough. 
4 Likes
I guess the difference is this:
No ko rule leads to infinite games, since any ko can be immediately recaptured, leading to a “stalemate” unless one of the players wishes to ignore the ko. All rulesets that I know of find this problematic.
Simple ko still may lead to infinite games, but these cases are relatively rare. With simple ko, one does not need to remember the game state further back than your previous move, thus this does not depend on the complete history of the game in a strong sense. From any position, you can determine whether a move is allowed, simply by asking what the last move was.
Superko makes infinite games impossible, but has the disadvantage that it depends on the entire history of the game. One can imagine a theoretical position where players play a game of Go, then suicide all their stones and start a fresh game of Go. However, in this “fresh” game of Go, no position from the previous game may be repeated. This is certainly something that humans struggle with, and in fact even computers will feel this limitation if you employ this “start anew” strategy often enough (there’s simply too many possible games). So, there are games in a theoretical sense for which we have no practical method of deciding whether the game follows the rules or not.
3 Likes
The coordinates of the last move aren’t enough; you also need to know whether it was a capture, right?
1 Like
Most split threads on OGS have been split for far less deviation than seen here. Indeed, the ruleset discussion could stand on its own, and neither it nor this thread would be worse off for the loss of the other—rather, they would both be improved by the loss of clutter.
1 Like
I consider a “move” to be the difference between a board state and its previous state, thus to include that information, yes.
On the board, if someone would drop by and ask what the last move was, I wouldn’t say “Black placed a stone here”, I’d say “Black captured a stone here by playing here”
3 Likes
Then feel free to discreetly flag the posts that you’d like to separate, instead of childishly changing topic titles to show your dissatisfaction.
When an OGS game under Japanese rules gets stuck in a long cycle, sooner or later one of the players will realize the game is not progressing and call a mod for help. So we get to see these cycles. But even if a player can win by abandoning the cycle and the mod is aware of that, mods can’t give playing advice to either player, so the mod will typically declare the game a draw and annul it.
The basic ko rule is enough to prevent cycles in single and double ko situation. No mod intervention is needed in those cases.
There is nothing magical about the number 3 for kos, except that it’s the simplest cycle not prevented by the basic ko rule.
Fighting a single ko or a double ko is perfectly legal under any rule set, because other moves (usually ko threats) intermit the recaptures, which breaks the cycle and progresses the game. You seem to confuse long ko fights with long cycles? Fighting a ko or a double ko for a long time is not a long cycle.
An example of an avoidable triple ko: Let’s say black is leading a game by 10 points and the game is almost finished. Only 3 half-point kos remain to be resolved. When white captures the 1st ko, black should just connect the 2nd ko, instead of recapturing the 3rd ko. Problem solved.
But not all players would realize this. A superko rule (in an online game) may help them to play the best move, because the mistake they want to make, would not be allowed by the server (perhaps they would call a mod to explain why).
I don’t see how that would work in practice with rational players who try to win. Superko gives an advantage to one of the players (by restricting their opponent’s options). Why would that player agree to anull the game, unless ofcourse they are still behind even with that advantage, but then why would the opponent agree to anull?
1 Like
You got a lot of nerve calling other people childish when you once used a sick-out-tongue emoji against me. What is that, about 2 years old?
My change of the topic title was to reflect accurately the content of the thread.
Typically, my post was strictly about the issue at hand, and your response descended into personal abuse.
Clearly I’m not the only one contributing to unnecessary “clutter” in this thread. Come on man…
Looks like we have established that the clutter is okay.
Much less of it would be occurring but for your typical abuse of me.
Can regular users ban moderators?
Neither of those two users is a moderator currently.
2 Likes
Not even of the forum rather than main site?
Correct. That’s why their names don’t have the shield like mine does. They’re both trust level 3 called “regular” which gives them some abilities like changing thread names, but it’s not a granted level it is handed out automatically by discourse based on interaction criteria (certain amount of reading, posting, etc.)
2 Likes
Do you have a proof of that?
1 Like
The proof would be something like:
- Both players have an optimal strategy because this is a perfect information game.
- The two perfect players can play optimally to scoring without komi, and there will be some result B+X (e.g. B+7).
- Playing with X as komi, if both players use an optimal strategy, the result will be a tie.
4 Likes
For myself Go is crazy good, chess is crazy boring. Every game is different if not vs a bot. Chess board feels so small and makes me feel stressful. I play go to feel better, why another reason else do I seek.
4 Likes
Yes:
Zermelo’s theorem states that for any finite perfect information 2-player game, either one player has a winning strategy, or both players have a drawing strategy. With a superko rule, Go is a finite game, since the number of board positions is finite, and no game can contain multiple of the same board position (situationally nor positionally).
In particular, for any komi, there is one player with a winning strategy, or both players have a drawing strategy. It is clear from the rules of Go, that if White has a winning strategy for a komi of X points, then White has a winning strategy for a komi of X + Y points for any positive Y, since White could use the same strategy. Similarly, if Black has a winning strategy for a komi of X, then Black also has a winning strategy for X - Y komi for any positive Y.
It is also clear from the rules of Go that if White has a winning strategy for X komi, then White has a winning strategy for any Y strictly larger than ⌊X⌋ (= the largest integer less than or equal to X): every game has an integer score difference. Hence if X is such that White has a winning strategy for all games with komi larger than X and White does not have a strategy for komi less than or equal to X, then X must be an integer.
Finally, it is also clear that Black has a winning strategy for a komi of -362: Black simply passes continuously until White runs out of possible moves, and Black wins by komi. Similarly, White has a winning strategy for +362 komi.
Therefore, there must be a least integer for which White has a winning strategy, let’s call it X. That means that regardless of Black’s moves, White can win with at least 1 point more territory than Black. Now suppose that Black has a winning strategy for X-1, and let White play according to their winning strategy for X komi, then White can end the game with at least 0 points more territory than Black, in other words, White can either win the game or tie the game, which contradicts the assumption that Black has a winning strategy for X-1 komi. Therefore, since X was the least integer for which White has a winnings strategy, it follows that Black and White can force a tie for a komi of X-1.
Similar reasoning shows that a komi of X-2 will result in a winning strategy for Black, since Black can use the tying strategy for the komi of X-1 to end the game with at least 1 point more than White.
10 Likes