ButtonGo/Discussion

Anonymous: Why should taking the button remove ko bans? The button will surely be taken after all useful board plays, and so to lift ko bans would change many strange ko situations under normal rules, such as the 5x5 example here. To think of taking the button as altering the board state does not really reflect the idea of the button in my opinion. I would think of it as just giving white a point for passing first (and the half a point thrown in with komi).

Bill: If there is a ko, taking the button will not take place before all useful board plays. And taking the button does prevent a repetition of the game, just like making a play on the board. It is just another move. And, like other moves, it removes ko bans.

As for lifting ko bans in strange situations under normal rules, some people, like Ing, would regard that as a good thing. And what is so normal about so-called normal rules? (Anyway, these are side issues. Button go bridges area and territory scoring in a simple way.)

The Count (anon. cont.): Ah, yes I guess their is no problem with becoming different to the normal rules. You prevent board repetitions regardless of whether you think of taking the button as a move or not though. Similarly, you could have a superko rule which forbids the third repetition of any board situation. This would also prevent endless repetition. So because it is a matter of choice whether a button-take lifts ko bans, I don't think it should be an implicit part of the button rule, although it can be an additional option. Would you agree?

Bill: Taking the button is simply a play, and is treated as such. Passing is different. The Ing rules treat it as a play, but other rules do not, and the traditional meaning is that it is not a play.

The Count: It is not required that taking the button is treated as a play. Do you disagree? Of course, you realise my personal preference would be that it is not treated as a play. I will lend my argument in favour. The taking of the button should not even be a physical move; it is acceptable to simply remember who passed first and give that player half a point, just as komi is given to the player taking the second move. Similarly, superko is implemented by remembering the previous board positions.

Bill: IMO, one of the virtues of button go is its simplicity. Giving 1/2 point to the first passer can introduce needless complexity. For instance, suppose that a pass does not lift a ko or superko ban and two consecutive passes end play. That seems normal, but you could run into a situation where a player should fill in territory (gaining nothing locally) instead of passing, because if she passed (gaining 1/2 point) the opponent could pass and end the game in an anomalous position. Gaining nothing instead of gaining something because of a rules quirk is definitely weird. That would give people something else to quibble about.

The Count: Who says that player deserves the half point? If this position is anomalous, it is because passing does not lift ko bans. Consider playing the same position under territory scoring. Then the player has no button to take, and so must play inside her own territory, also losing a point (total), before she takes the ko. If this position really is anomalous, we should make passes lift ko bans to sort out territory scoring at the same time.

Bill: Pardon me, but this seems to be the kind of quibbling that I am talking about. Does the player deserve the half point? Should a pass lift a ko ban? Should two passes end the game? What is anomalous? And so on. Better avoided, IMO.

The Count: I think I have not been clear. What I meant is essentially that you are quibbling! Were you not saying that button go is more virtuous than half-point-for-first-passer go because of a subjectively anomalous situation in the endgame involving passing and kos? My last post was meant to say that one could easily quibble against your point. As you know, I just want you to admit that half-point-for-first-passer also works.

Bill: I proposed button go as a form of go that combines what I consider to be favorable qualities of both area and territory scoring. (In fact, I hope that it, or something similar, will resolve the clash between the two as go internationalizes.) Giving a player 1/2 point for the first pass instead enmeshes the idea with all of the questions that have been raised about passing. From my point of view, that does not exactly defeat the purpose, but it muddies the waters. Better to keep the two ideas separate. Sure, half-point-for-first-passer is a viable rule. Ikeda proposed a very similar rule. But it does raise questions about passing. OTOH, you can add a button to any area rule set without reopening such questions. Taking the button is just another play.

Bill: FYI, it is not necessarily the case that the player would have to play inside her own territory under territory scoring. For instance, under Japanese rules.

The Count: Whoops. Yes, I didn't realise this about Japanese rules. I should have specified territory scoring, where only board plays can lift ko bans.

willemien: I was wondering is button go possible if I also want that optimal play leads to jigo

MrMormon: With the right komi, sure.