Anti-escaping and Anti-stalling features

Ew no ill never fill my dame :crazy_face:

1 Like

I still agree with what I posted here, but I would like to add one more caveat to my comments on the Anti-Stalling Feature: only the player who passes x times in a row (I believe it’s 3, currently) ought have the option to end the game by server decision, the player who is still playing stones ought not be given that option

If this is already the case, then great! but I have not been able to ascertain whether or not it is

In the situation in which my opponent is trolling by refusing to pass even after all dame and teire are filled, I could click “end game by server decision” after passing a few times, and it wouldn’t matter that the opponent didn’t get the chance

But the situation I expect to encounter more often (though still rarely), is that I am playing a complete beginner. In such cases, I am usually happy to let them play well past the end of the game, because the rationale for why certain groups will be removed can be quite abstract at that level, and I hope that by playing a game all the way to its logical conclusion, the rationale for the shortcuts taken during scoring will make more sense to them (of course, if they pass I’ll pass right back and go to counting). If I in the course of doing this get an option to end the game by server decision, no problem, I can just decline the dialogue. But if my opponent gets that option, especially being new, they may assume that that’s just how Go games are supposed to end and click it (it seems we’ve had many people being confused about the popup and accepting it without realizing what it was doing), which would undercut all the time I had just put into giving them the valuable experience of playing out a game not until a 3k thought it was done, but until they thought it was done, and then seeing how the 1 and 2 stone groups (if any) which were left in my territory were removed during counting (since they passed, hopefully they’ve played to the point where they actually understand why they can’t save those stones, and aren’t just trusting me as the stronger player). They would be unwittingly cheating themselves of a valuable learning experience that they’ve already (whether they know it or not) invested some time into, and rendering my attempt at helping a new player wasted

5 Likes

The Anti-stalling feature did not trigger for
巧克力奶昔 vs. hoctaph :

I don’t know if this was

the threshold actually being ​ ​ ​ > 10.5 ​ ​ ​ rather than just ​ ​ ​ > 10 ​ , ​ ​ ​ whether directly or by rounding
or
resumption resetting the counter, even though from what anoek said it shouldn’t
or
some larger issue

.

Currently both players are given the option.

It’s a good point, and I’ve also noticed DDK players occasionally click this of their own accord when playing with me in this circumstance. (when I am happy to play on if they wish)

Edit : I have wondered whether that feature being used much more often will take away from their opportunity to learn why the end of the game works as such, too.

Relatedly, I had also been seeing it used at times in games where both DDKs seemed still confused about which groups were alive/dead or worth continuing to play from at the end.

5 Likes

Hopefully this is amended

2 Likes

@anoek @Eugene

It seems that a malicious staller is able to avoid this feature by bringing the game to the scoring phase in between the passes:

Perhaps the scoring phase should be ignored and the victim should be offered the option to end the game, as long as one of the players keeps passing and the estimated score does not change significantly.

7 Likes

Is this true? I just had a 19x19 game yesterday where my opponent and I went into scoring phase, i realized we had forgotten some borders, so exited scoring phase. The second set of passes did cause the anti-stalling button to show up.

(Which, in itself, seems like a bug since neither of us was trying to stall or trigger the anti-stalling feature)

2 Likes

Well, in my test above the button did not show up, despite White passing 7 times consecutively. Feel free to test it out yourself and see if you can reproduce it.

Given your report and @hoctaph’s report above, I’m wondering if there’s an issue with 9x9. At least that is the common feature

1 Like

Actually, could this just be the width * height / 2 rule? It looks like your test was under 40

1 Like

Isn’t it White + 48?

I thought it was that at first in this test, that’s why I redid the test with a sufficient margin.

Number of moves, not points margin. I believe that rule is to prevent triggering in the opening, when there is still a chance to catch up.

1 Like

Ahh…

Hmm, yeah… Then I think this rule should be relaxed for 9x9 games, because <40-move games are quite common on 9x9.

I’m not sure why number of moves matters in general. It should only trigger when passing does not significantly decrease your score, which will never happen in an open game where there are still useful moves to play.

5 Likes

Surely if clicking a button is unacceptable then we just shouldn’t show the button?

I mean in some cases sure there are ways to misuse buttons, misusing the resign to quickly and intentionally lower your rank - and we can’t not show the resign button.

Surely here though, if we don’t want someone using this feature at a certain time or instance, then we don’t show it. It’s going to happen well into the game anyway - and really the main abuse is to win a game you’re winning without playing it to the end?

I’d rather we just fixed the issue than warned users not to click a button we’re showing them that they’re “supposed” to be allowed to use to end the game.

1 Like

I feel like this is the difference between what people say they want and how it’s implemented.

I don’t think there’s any caveat as to how many points you’ve lost with the last 3 passes. Maybe you lost cumulatively 5 points, maybe you’ve lost cumulatively 30, but either way you’re winning by more than 10 with 99% certainty. So the number of moves is probably to make sure you can’t abuse
it too early on.

1 Like

Yeah, especially in handicap games

I think detecting end of game with all dame and teire filled is non-trivial, so I’m not sure how that logic could work. It seems like it has to be there, but we need a clear rule that it is only for use after all dame and teire have been filled

But that seems exploitable, so I’m not sure that’s wanted or necessary.

If this tool is really to stop people maliciously stalling the game, then they just do what they’re doing anyway except don’t fill all the dame.

It’s not that these players won’t get reported, it’s just that to the person that doesn’t want their time wasted, playing out a ranked game where someone is stalling, or waiting for a moderator to help. That’s the purpose no? It’s not about playing the game to its logical conclusion necessarily and only targeting a small subset of stalling where you stall only after you found out you lost in counting surely?

You fill it. If they don’t fill teire, exploit it. I don’t see the exploit

“You fill it” could be 20 moves, 50 moves who knows. 19x19 is a big board and when someone dies with a group, it’s not always close to having all dame and teire filled. Stalling is still possible.

Right?

1 Like