No need for estimates. Just do what any pro tournament would due if outside unforseen circumstances halt the play.

Wind back the clock, continue play from the position from where the disruption happened.


Both players have a referee/proctor with them to see that they play moves that aren't generated by a bot, so there's already a third party monitoring move input.
Shouldn't be too difficult to say if a outside influence (like lag or disconnect) caused the move not to be played. Specially if the move was entered with 10 seconds left.

Sorry, but estimates are needed. Maybe you could decide the game is won, maybe not. There has to be checked, if it´s really not the fault of the Player or not. Or to check, how the wishes of the Players are. Maybe Eric says, I don´t want to Play anymore and loses the game.

And it´s especially about estimating, cause you don´t have clear game like that every time. Sometimes you have not much time, you are in Byo Yomi. But it´s a real tactical situation. The time ran out, cause of a lag and it´s result is not ok, but simply start the Situation isn´t ok either.
In time measures maybe you didn´t find a winning move (a difficult Tesuji), but having more time, cause the Situation can´t be fixed immidiatly , gives you the possibility to find the move.

It´s about giving a Jury or a Referee the possibility to estimate. Sometimes your decision is right and it´s fair, but sometimes it´s not. Here is it about something Basic.

A lagging connection between KGS and the player is an outside factor that isn't something that is expected to happen often.

A referee on-site with the player can confirm this, and play can be continued from the position it happened, with the time that was available when it happened.

This is not in any way different from a real world tournament and outside factors. In those situations as well, both parties might get extra thinking time while the referees wind back the clock and restart the game. This is seen as an acceptable and working solution in the real world, why can't it be online? Specially since both have referees with them.

If you loose on time due to not playing on time, you loose on time.. that is how it works both online and in real live tournaments, however, if something outside your control affects the tournament and stops you from being able to make a move, you just return to the earlier position and restart play.

Has nothing to do with if the game is close or not, in early stages or not. If the game is stopped by outside forces, you just return back to the game when you can, from the position where the disruption happened.

The Thing that it doesn´t happen often, makes it necessary, that it´s estimted. Especially we have even more time, than in a tournament it´s fixed (and to honest in a tournament on a real board it wouldn´t happen that way, cause your move is on the board, there is only about, if the clock is working or not. I answer from a position I got, cause I am a apprentice in the public Service and I know some of the problems happening and I know solutions based on estimation.

Sometimes it´s unfavourable to have just one solution, believe, I know this for sure. It´s different having 2 minutes addited to the clock and maybe 3 minutes to clear the circumstances, than having ours before the problems are fixed. Both can have difficult position and time too. One has a winning Tesuji, that is hard to see. Give him ours he see it, or worse, he already talked to someone, don´t that is Happening something, cause is not fast to fix and than he hear the move that wins (that wouldn`t happen at a normal board).
This makes it definitely total different from your example and shows why an estimation is needed. In an estimation your way can be used, but also other variations, like a total rematch, according to the one, who decides.

we still do not know what the proctor said?

_________________
Humans live through their myths and only endure their realities.
I regard belief as a form of brain damage, the death of intelligence, the fracture of creativity, the atrophy of imagination.
RAW

Yes, I agree it's a handicap if one player has to play faster to compensate for their bad internet conditions and that's unfair. The question is, do you pragmatically accept that unfairness and adapt to maximize your chances of winning (I expect that if Mateusz kept 1 spare period as "lag insurance" and played fewer moves in the 5xth second then the deterioration in his play would have less effect than the decreased time-out chance), or idealistically continue as though those conditions don't exist and have an increased chance of losing? Considering things 'outside the board' is something pros need to do even in real-life events. For example in his commented games book Lee Sedol says that he had trouble eating oily and spicy Chinese food and was often playing his games in the Chinese league hungry, which he partly blamed for his bad results in the early years. So he took pot noodles and also adapted to eat the cuisine and his results improved.

A problem with not giving Eric the time win (and I say this as a slow player who has many painful losses on time, including in the 3 hours main and 10 stones in 10 minutes British title match) is that Andy could then turn around and say "If I knew the time limits weren't rigid then I wouldn't have played that awful time-suji blunder in my game in a winning position and I would have won". That argument does have some holes (he wasn't winning at the time of his blunder according to bots, it was a middlegame position with lots of game still to play rather than late endgame with a clear big lead) but does have some merit too.

i think they should first clarify what caused the timeout. i think it would help to adress the general problem with such situations apart from this specific occaision.

_________________
Humans live through their myths and only endure their realities.
I regard belief as a form of brain damage, the death of intelligence, the fracture of creativity, the atrophy of imagination.
RAW

This is the most difficult part of the whole ordeal - since it was a transient error in network transmission, there's no proof of what happened other than Mateusz's and the proctor's statements. Unless the proctor was paying direct attention to Mateusz's actions at the moment of the network timeout, they can't honestly say that Mateusz did submit an action in time. If he also has no directly verifiable proof of doing so, for example through a camera recording of himself playing, it will be near impossible to prove objectively that this was the case.

The situation would be easy to manage in an over the board event. If you have a faulty clock, then you ought to be able to stop the game and have it replaced. What are the rules here though? It seems inconceivable that they didn't consider what to do if one player is disconnected. The options are unfortunately a bit limited on the internet.

You could continue through a Skype-type application if you were near the end of the event.
That would mean that the Proctor would have to use a clock to count down 60 seconds byoyomi each move, whilst the two players used a demo board. Ugly, but not impractical.

You could also just restore the position with a 5 period byoyomi and let the players restart on SD.

Pity they didn't have the funds after all to send all 10 players on vacation in Iceland.

By the way - why isn't the Canadian Go Association a co-host?

_________________
North Lecale

no that makes it easy. either you trust the proctor or you do not. (there could also be the possibilty that the proctor did not catch it) assuming they trust the proctor (or why else would you have them?) this specific case should be easily solved.

ps. i am aware that this is not a solution for the general case, since it does open up loopholes. thats why i said its important to distinguish between the general problem and this specific case.

i am glad this happend actually. i hope aga and egf get together and make the best go server possible

_________________
Humans live through their myths and only endure their realities.
I regard belief as a form of brain damage, the death of intelligence, the fracture of creativity, the atrophy of imagination.
RAW

No proof of what happened?

Pardon my ignorance, by why doesn't the software record the time between receipt of the opponent's move from the server and the time the player's move is submitted? Net lag has always been with us, so I would have thought that for timed events the local time taken would be the proper record.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Because KGS was written 20 years ago by 1 guy as a hobby, rather than as a serious piece of software for business critical real time operations. And as Marcel says there's a question of trust between server and client so it's not trivial.

I would like to add a couple of facts to this discussion. First, Eric never resigned, that rumor simply needs to end.

Second, the people in the room with the players are NOT referees, they are proctors. They have two jobs - first, to observe the player and make certain no help of any kind AI or otherwise is happening, and to keep the communication open with the broadcast team. The proctor was NOT instructed to monitor any lag issues or to make any refereeing decisions outside of that mandate.

_________________
My days of not taking you seriously are certainly coming to a middle

well do you trust the proctor to be true to his word on his job? if yes, it doesn't matter if it was his job to monitor lag issues. there are 3 possibilities:

he saw mateusz time out because he forgot about time while reading: eric wins on time.
he saw mateusz try to play the move in time, but for what ever reason the move did not register: continue playing the game.
he can not in good faith confirm either of the cases. thats the tough case. as sad as it is, you would have to give eric the timewin, unless the parties agree upon another compromise.

notice he is not makeing a refereeing decision here: he is reporting to the tournament officials what happend and they make the decision. and i repeat again: it is important to swallow the pill in this special case and strictly seperate it from the general case for future handling of such matters.

_________________
Humans live through their myths and only endure their realities.
I regard belief as a form of brain damage, the death of intelligence, the fracture of creativity, the atrophy of imagination.
RAW

I'm not talking about KGS, I'm talking about the tournament organizers.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Oh, but making sure that the local timestamp is not forged would fall under the heading of making sure that no help of any kind is happening.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Truth be told, you could reduce the possibility of tampering with this mechanism, by introducing a procedure of delivering a trusted client onto a trusted system inside a virtual machine or something along these lines, to an "acceptable" margin; whatever "acceptable" means is another matter, but whatever words we would use to define it, one of them would be some form of "practical". However, this would require: a) a custom trusted client, b) a trusted source, c) a trusted system and a trusted machine (in case of a virtual machine, it's a matter of transmitting an image from the trusted source to the player, possibly along with the client). All these points add some technical difficulty into the mix, but when we're talking about a professional grade tournament, a bit of technical preparation beforehand won't hurt. Of course, this won't truly and clearly prevent timestamp cheating (without getting into details, the evil maid is always an inherent threat), but here we can go back to the pre-established definition of "acceptable".

If some new client software for KGS was devised now, it would surely draw from the experiences in network protocol design of popular gaming titles, old and new, which had and have to make sure not necessarily that everybody sees the same thing at once, but that eventually everybody will see the same thing, regardless of network quality, and the moments of inconsistencies will not lead to desynchronizations. After all, the server is the ultimate authority here and if inconsistencies do arise, it's much easier to enforce corrective or punitive action.

As a German I say "was für eine verfickte Scheiße, der Rotz geht einem echt nur auf den Sack", I won´t translate, but all can be sure, it doesn´t show any happiness.
I think every participant should have a camera in the on his monitor, keyboard and mouse, so all can see, if he did something or not.

Nethertheless is there a decision already? I didn´t hear anything, but first it was said we should look today. In the end, even if it´s decision against Mateusz, we could know, which pairing is next and we are still ahead in the EGF.

If go clients work in a different way this sort of problem would not happen.
IE if time was counted locally, starting from receive of opponent's move, and ending upon playing of your move. On this way, all lag and delay would not be counted on player's time.
Also adjusting byoyomi reading style would be useful.

I know that some might say that someone vozld use flaw in this system, and disconnect from internet in order to get more thinking time on opponent's turn, but it is easy to monitor such activity.