Life In 19x19

In Chess the Grandmaster Draw (while technically illegal) is very common. Two players agree to draw a game, sometimes to protect rating, sometimes to win a tournament, sometimes for other reasons. Yet in Go, where the opportunity presents itself (even komi / triple ko=draw ) this is completely frowned upon. Two Russians were basically disqualified for trying to do this at the London Open one year (one can presume that they thought there was absolutely nothing wrong with this).

What do you think about this?

I think chess players are a lot more used to draws in general, as 1/3 or so of games end up that way IIRC.

In Chess, in-game draws themselves are common, whereas they're quite rare in go, and one could even argue that it basically never happens: whole number komi is unusual without a "white wins ties" provision or similar, and triple-ko is technically "no result".

I suppose in Chess the idea might be that you and your opponent could just play enough drawish moves and you'd almost certainly draw. In go, it'd be possible in theory to somehow set up a triple-ko by agreement, but situations in which you could do so without making bad/losing moves along the way (and thus giving your opponent a chance to deviate) would seem rare, so any similar claim (i.e. "we could just play drawish moves if you insist that we keep playing, referee") wouldn't hold as much water.

Anyways, I put never, but really mean "basically never, but perhaps with exceptions for weird circumstances I haven't thought of." Possibly I should've checked the medical emergencies box. It's not particularly good for the game I think (one doesn't get to see as much go played, etc) so without a justification as in the previous paragraph, I'm going to have to say no.

Could someone please give me a good example of when this would be desirous? I assume there may be loopholes in some tournament rules that would let both players qualify for events or something if they tie?

Chew Terr wrote:Could someone please give me a good example of when this would be desirous? I assume there may be loopholes in some tournament rules that would let both players qualify for events or something if they tie?

Easy. It is round 9 of 10. Alf has 8 points, Nearest rival Bob has 7 points. If Alf has a draw, with his friend Wibblepaws, he is certain to pick up first prize. In chess tournaments you will see people throw out 4 or 5 moves and agree a draw in this setting.

I voted never.
With all modern rule-sets I'm familiar with there is either a non integer komi, or a rule that awards the win to white (or black) in case of jigo.

Maybe a game can be declared a draw if there is an argument and the referee is called. I think if both players agree a referee can call it a draw
(If someone knows this to be true or not let me know)

A triple ko is not a draw, but a "no result". there is a clear difference. for one, a draw in go affects your rank, a no result does not.

Javaness wrote:

Chew Terr wrote:Could someone please give me a good example of when this would be desirous? I assume there may be loopholes in some tournament rules that would let both players qualify for events or something if they tie?

Easy. It is round 9 of 10. Alf has 8 points, Nearest rival Bob has 7 points. If Alf has a draw, with his friend Wibblepaws, he is certain to pick up first prize. In chess tournaments you will see people throw out 4 or 5 moves and agree a draw in this setting.

Thanks for clarifying. This is the kind of thing I had imagined, but wasn't certain about. That said, good or bad, I wonder if this is much worse than if Wibblepaws just resigned to Alf? Even if they played the whole game, it's harder to question a lost game than a possibly intentional draw.

Chew Terr wrote:

Javaness wrote:

Chew Terr wrote:Could someone please give me a good example of when this would be desirous? I assume there may be loopholes in some tournament rules that would let both players qualify for events or something if they tie?

Easy. It is round 9 of 10. Alf has 8 points, Nearest rival Bob has 7 points. If Alf has a draw, with his friend Wibblepaws, he is certain to pick up first prize. In chess tournaments you will see people throw out 4 or 5 moves and agree a draw in this setting.

Thanks for clarifying. This is the kind of thing I had imagined, but wasn't certain about. That said, good or bad, I wonder if this is much worse than if Wibblepaws just resigned to Alf? Even if they played the whole game, it's harder to question a lost game than a possibly intentional draw.

I guess Wibblepaws gets a rating boost from drawing with Alf, who in this situation is likely to have a higher rating, and Alf gets the tournament win, so there's something in it for both of them.

I think that it should be allowed. After all, it is "legal" for a person to play poorly on purpose. Of course, it is more interesting if they actually play the game.

Kirby wrote:I think that it should be allowed. After all, it is "legal" for a person to play poorly on purpose. Of course, it is more interesting if they actually play the game.

That's an argument for why resigning should be allowed even when you're winning.

What do you make of the argument in the second paragraph in my first post above (#3 in this thread)?

Javaness wrote:

Chew Terr wrote:Could someone please give me a good example of when this would be desirous? I assume there may be loopholes in some tournament rules that would let both players qualify for events or something if they tie?

Easy. It is round 9 of 10. Alf has 8 points, Nearest rival Bob has 7 points. If Alf has a draw, with his friend Wibblepaws, he is certain to pick up first prize. In chess tournaments you will see people throw out 4 or 5 moves and agree a draw in this setting.

In fairness, this almost never happens in chess quite like you describe. Quite often, players in this position play openings that are known to be almost identically equal with little variation, and agree a draw perhaps on move 20 or so, but under 10 move draws are almost unheard of.

I missed your poll options, I really did, thanks for posting this one

I personally see no great value in draws - there are no real draws in Go, and this is fine by me. Without draws being a likely natural result on the board, any agreed draw would feel contrived. There are some big practical game differences. In a situation nearing the end where the game is really close, this is broadly as follows:

Chess

* 90% draw chance, 5% each player winning (or thereabouts)
* Very little either player can do to force the game in their favour
* Effectively almost unlimited remaining moves when both players know the outcome

Go

* 0% draw chance assuming non-integer komi, so ~50% chance of each player winning
* The endgame moves become vitally important, a single point or two can make all the difference
* The game will end, and the number of moves left is clearly obvious to both players

Practically speaking, the reason why agreeing draws in chess makes sense and doesn't in Go is close to self-evident I think.

Some tournaments still use integer komi (for Japanese rules) and I have had a jigo in one such event. Triple ko used to be a draw in BGA Japanese Rules , it was declared to be a draw in the UK under AGA rules once too.

prokofiev wrote:

Kirby wrote:I think that it should be allowed. After all, it is "legal" for a person to play poorly on purpose. Of course, it is more interesting if they actually play the game.

That's an argument for why resigning should be allowed even when you're winning.

What do you make of the argument in the second paragraph in my first post above (#3 in this thread)?

I don't disagree with the points you brought up in your second paragraph, really. It's just that, if people are trying to manipulate the result of a tournament as Javaness describes, they can already easily do so by mutual agreement (one player loses on purpose).

I'll admit that there is a slight added benefit for the "manipulating type" if mutually agreed upon draws are allowed - both players may benefit in the tournament result, whereas if one player resigns, only one player benefits.

Consider that point, maybe I've changed my mind a little bit... (Am I allowed to do that in an online discussion?)

I guess that there is not much added benefit to allowing for ties, and there is a slight negative consequence (in some cases, both players can mutually benefit). This doesn't take away from the fact that people can still manipulate a tournament result by mutual agreement, which, I guess, is another topic.

So I guess I would say that I am not strongly against allowing mutually agreed upon ties, since people can already manipulate tournament results, but I don't see a strong benefit to allowing the ties, so it would seem fine to disallow them.

Javaness wrote:Some tournaments still use integer komi (for Japanese rules) and I have had a jigo in one such event. Triple ko used to be a draw in BGA Japanese Rules , it was declared to be a draw in the UK under AGA rules once too.

Even in tournament rules where this is the case, the likelihood of a draw is still extremely low (< 5%). In chess, at the point offers are made, the chances are normally over 50%.

Kirby wrote:Consider that point, maybe I've changed my mind a little bit... (Am I allowed to do that in an online discussion?)

Not yet I think, but I hear that feature has been requested for Web 3.0.

Kirby wrote:I guess that there is not much added benefit to allowing for ties, and there is a slight negative consequence (in some cases, both players can mutually benefit). This doesn't take away from the fact that people can still manipulate a tournament result by mutual agreement, which, I guess, is another topic.

Fair enough. As you suggest, it's possible for them to e.g. trust each other that they'll keep going with the triple-ko setup instead of taking advantage of bad moves involved in the set up. I don't see any benefit in making such behavior any easier, though.