How about dumping all cheaters (regardless of what type of cheating) in a separate pool where they can play (and cheat) each other until hell freezes over?
2 Likes
We don’t do that?
2 Likes
The idea has been floated before 
FWIW, I do agree that the situation would be very much improved if moderators could change the score.
I don’t recall a compelling reason why this should not be possible, though I do recall there are complications like “what about tournaments and ladders” … etc
2 Likes
Well these are different perspective
3 Likes
I agree that OGS would be confusing if it allowed the Moderator to change the results.
So our appeal is not to allow Moderator to change the result.
In fact, even if the Moderator is allowed to change the result afterwards, it will give people a bad impression.
All I saw in the moment was that I lost.
I may play this one game today, and the next time I play Go may be a few months away.
You asked me to remember this matter for a few months, and wait until I come back next time to see if the result has been corrected?
Or I will continue to play in the future. Maybe I will play 20 games next, do I have to remember which game is cheating and needs to be corrected?
Moderator changing the result is the last resort that the system really can’t handle.
The best way to improve score cheating is to solve the problem from the root.
Only two very simple functions are required:
- After reaching a certain number of moves (for example, 350 times), the player can press the button for mandatory scoring.
- Players report score cheating to Moderator, and the game will be frozen. In order to avoid abuse, if the player fails to report in the live game, the opponent will be judged to win (opponent’s advantage) or cancel (our advantage), and be persuaded or warned.
I believe the above will absolutely solve 99% of score cheating. (In fact, 2 may not be used at all. If there is 1, it will definitely solve 99%, especially Blitz.)
Your first nine paragraphs are very confusing. It has always been widely agreed that the best solution would be for moderators to reverse the result. They already have and use the power to change the result by annulling a cheated game. Years ago, I was told that reversing a result was impractical for technical reasons, because the ripple effect of such a change, especially with many reversed games, one atop another, would gum up the operation of the ranking system (or something to that effect—I may not be expressing it quite right because I am not a techie). However, since the ranking system was subsequently changed to Glicko some years ago, it may be time to revisit this question. But only anoek can provide the answer, I think.
Your paragraphs about remembering are particularly confusing. Remember what? To what purpose? Here is how the report system works: You get score cheated, you make a report, a mod annuls (or ideally in the future, reverses) the game result, and the mod TMs to you that the game has been fixed. What is there to remember?
How does empowering mods to reverse a result give a bad impression? In truth, the OGS community would be elated.
This is contrary to the rules of go. Moreover, the OGS autoscore is not good enough to be a final authority on scoring the game. There are two recent threads with loads of examples of wrong scoring by the autoscore, including a persistent bug over the last two years. In addition, I have 20 or 30 more examples of bad scoring by the AS that I have never bothered to post.
This idea has already been discussed earlier in this thread and in one or more older threads. It was also discussed by the mods several years ago and rejected. The problem is the potential for abuse, which is not solved by your proposal. I don’t remember where I stood on this question years ago, but now I think it should probably be tried in the absence of anything else.
However, empowering mods to reverse the result—if it is technically possible—is a simpler, direct, definitive fix that would remove the urgency of score cheating and stalling reports that result in undeserved wins by the violator.
1 Like
The main problem of automatic scoring comes from the unclear judgment of life and death.
However, I don’t think there is anything ambiguous after 350 moves.
Unless it’s a beginner’s game, but this extreme situation definitely requires Moderator to step in, and no system will help.
What I am proposing now is that there is no way for the Moderator to come under the field and end the game.
If I make all of my opponent’s moves except Pass into Suicide, can I end the game?
No, he can still Resume, Pass, Resume, Pass.
How do I win? There is simply no way to win!
That’s not true. The system is still very flawed as amply documented in the threads I mentioned. Look them up (I don’t generally do research for other people). Also, as I stated and you ignored, there has been a bug in the system since August 2020, involving unscored territory.
Edit: this does not address the new paragraph you added.
350 moves is a lot, cheating, if any, occurs much earlier, before move 250. I don’t think people will have the patience to wait for 100 moves to get the scoring problem fixed.
BTW, I dont think Fox is a better environment with respect to scoring. Many times, after I “requested automatic counting” on Fox, my opponent refused counting and continued to stall the game by playing meaningless moves.
At Fox, close the border, fill the dame, and press the smart referee to end it. (only for 7d and below)
Above 8d, you need to seek Moderator or 350 moves. The problem is that Fox Moderator is getting there much, much faster than OGS. (never longer than 30 seconds)
Can OGS do it? Obviously not.
But OGS relies more on Moderator to solve problems on the fly than Fox, and that’s the problem.
OGS does not have any system that automatically scores points, even after 10,000 moves have been made in the game. Moderator processing speed is relatively slow, and the result cannot be changed after the fact.
Not true, if I understand you correctly. Moderators can and do decide a game in progress, including during the scoring phase.
The problem of perpetual restarts is a well-known form of stalling. I’ve seen two games in which cheaters restarted over 100 times. It should be reported. Currently, like score cheating, the game is lost if you don’t want to waste time fighting over it. Giving mods the power to reverse results would solve this too.
A friendly piece of advice for anyone reading this: don’t chat with the cheater at all, and don’t counter-cheat (some players try countering a cheat by claiming all of the cheater’s stones and territory). These actions only antagonize the cheat and give the cheat energy to continue the battle. A dogged resistance has a better chance of tiring out the cheat. I have seen many cheats back down. Also, when someone score cheats, just click the autoscore button to reset the original score.
oh my english…
I mean, Moderator is off the scene. I have no way to end the game.
In addition, I think another problem that causes score cheating is: the final result of the score is actually determined according to the final state (that is, the cheater).
If there is no way for the two parties to reach a consensus when the timer ends, and the state of the board is not automatically scored, then the scoring phase should stop, the game should be frozen, and handed over to the Moderator for decision. (However, this can also be done via reports)
In short, no matter what, scoring should not be done when it is in dispute!
1 Like
Now that is the first original idea I have seen here, and it has the advantage of not requiring a person to file a report while simultaneously battling a cheat. However, it does encounter another problem: escapers who do not accept the score. This happens quite a lot even when there is no dispute. The result would be many additional, unnecessary reports to the mods, so I doubt this would ever be adopted. However, perhaps it could be tweaked, so that the scoring would not be frozen if one player left the tab.
I’m going to bed, so I am done here for now.
1 Like
This is only required when life and death are changed.
If there is automatic scoring on the board, it will be automatically settled when the timer reaches zero (same as now).
In addition, in order to avoid the abuse of Resume, if two players still pass after Resume (that is, pass four times in a row), the Resume button will be disabled and only the scoring stage can be performed.
In this way, the problem of Resume, Pass, Resume, and Pass can be solved.
I would want 6 rather than 4:
If my opponent is scoring as alive my opponent’s dead stones, and
OGS has it be my turn after resumption, then I usually offer a pass in chat,
in case my opponent thinks that [who plays first after the resumption] matters.
I have so far not had anyone else indicate wanting me to pass in such a case,
but if a single pass-pass after resumption prevented further resumptions, then
me passing as offered in such a case would let my opponent stop me from carrying
out a capture even if I’m ahead by enough to actually play the relevant moves.
Enter the scoring stage after two passes, will the opponent pass enter the scoring stage again after Resume?
is that so? Well I really don’t know.
But at this time, Pass only has three consecutive passes, and both parties have not passed, so Resume should still be available.
Keep in mind that Resume is only disabled for four consecutive Passes, so I don’t think that’s a problem. So it should be 4, not 6. (Alright, Alright, if it’s AGA, then yes, 5 times.)
However, I think that after Resume, two players should pass consecutively before they can enter the scoring stage again. So yes, the current system is flawed and needs improvement.
The situation I was describing would be four consecutive passes, not three:
Before the initial scoring phase, I pass followed by my opponent passing.
That’s the pass-pass before the initial scoring.
My opponent scores as alive my opponent’s dead stones,
and we resume, with OGS having it be my turn.
I in chat offer to pass in case opponent thinks [who plays first after the resumption] matters.
This part has so far not happened to me, but I’m supposing
my opponent responds in chat that Yes, my opponent would like me
to pass so my opponent can effectively go first after resumption.
Furthermore, suppose I accordingly pass as described.
Now, under your four times in a row proposal, my opponent can pass
- thereby forming a pass-pass after initial scoring - to stop me from carrying out the capture,
even if I’m ahead by enough that I’d still be winning after playing the relevant moves.
Well, it looks like you want to play strictly by Japanese rules. (If you ask Resume, the opponent has the right to move first.)
However, OGS does not follow this rule (it will be moved first by the player who passes first).
In my opinion, for this situation, it should be up to OGS to correct their system (opponent moves first after Resume), instead of you manually passing to comply with Japanese rules.
They need to set the system so that the opponent moves first after Resume (if it is the Japanese rule), and the two players pass consecutively before scoring again. We are not obliged to do this by ourselves.
If there is a dispute because you moved first in Resume, the opponent can report it to the Moderator, and then you only need to argue for it. (Send change map to Moderator?)
You don’t need to prove to your opponent (because it is the system’s fault), but you should prove to the referee (Moderator) that you are right.
Well, it looks like you want to play strictly by Japanese rules.
Actually 
, the rules I’d want to play under would have,
territory scoring with an encore, where
to allow players to safely pass with an odd number of dame, resumptions don’t go
from the normal phase to the encore unless, as part of the previous resumption,
at least one player chose to have the next resumption go the encore
(Under correct play, this part makes no theoretical difference;
it just lets players safely end and score sooner.)
and
komi has a plus or minus epsilon part, and that part determines
who would win a cycle that (L) from this post of mine doesn’t resolve
and
who plays first in the encore if the encore is reached
.
(There are positions where the last part would matter, even assuming both that neither
player has passed yet and that both players play correctly from the position onward:
They involve miai values strictly between 0 and -1, or 0 and 1 under area scoring. example position )