Just to note something important on lag, and organizer's lack of knowledge:

- lag between server and player's computer is always present.
In most of cases, it is so small that it is impossible to notice. It must be there, due to laws of physics and equipment performance.
Sometimes, it could be few seconds. Mateusz had obviously experienced it, and he is used to it.
In his games, when it first occurred, he decided to play 10 seconds earlier, for safety reasons. Because he was so used to lag, he did not considered to report it to organizers.
Lag of 10+seconds is obviously something that strikes rare, and it is not something that anybody is used to.

Finally, comparing KGS clock to tournament clock is not something that should be done.
Referee's final decision was influenced by analogy that player should report if there is a problem with a clock.
But KGS clock is always "broken". In most of cases, it's delay is not possible to spot - but it is there. Sometimes it is few seconds, and sometimes could be very long. Organizers and referee have completely overlooked this.
Instead of is there lag (what referee thinks), there should be question what is acceptable level of lag?

For me, anything more than 1s (detectable by player) is too much.
Considering Mateusz' play, he was obviously used to it, and considered few seconds acceptable for play.
But 10+ seconds is not something that he could find acceptable.

Therefore, referee's decision that Mateusz' should have reported lag is flawed.
There is always lag. It's duration is what is important.

Yes, the organizers should never have used KGS to keep time.



You would have to ask him about that.



Right. It should not be used to keep time in a professional tournament. Never. Ever.



The organizers were clueless. They thought that even a lost connection posed no problem, because the player could simply reconnect. That would have been the case if time had been kept locally, not on KGS. Clueless!



Lag should not be an issue at all.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

Who cares? KGS is not the timekeeper. Use clocks that support Fischer time.

Physical clocks might not work if you have several games going at once locally, with the opponents who knows where. But this tournament has only one game going at any time. Local timekeeping is no problem.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

There is always lag. Certain ISPs and regions have larger amounts of lag than others. Certain days of the week show more lag than others, in aggregate. Patterns can be seen in network traffic based on a number of factors. I work at a network analytics company, and I've seen graphs that show this plainly.

The way I understand it, the referee's interpretation was that Mateusz knew that he was experiencing significant lag on a regular basis - enough to lose byo-yomi periods in games. Playing from a different location or via a different IP would certainly impact the likelihood of lag. If that were not possible, a discussion could still be had.

It's true that the referee's decision was without having the network data about Mateusz's connection, so the interpretation that the lag that caused Mateusz to lose the game was commonplace may indeed be flawed. The referee made some assumptions. But you don't have that network data, either, so you're making assumptions of the same nature.

_________________
it's be happy, not achieve happiness

It is good that you are able to understand something from the referee's statement.
Both the AGA and the EGF have a tradition of an appeals committee. Will they now institute one?
The AGA has overruled KGS server rules before. The EGF has never had the opportunity to. What will they decide there?
Will they decide to connect referee and proctors over skype?
Will they decide that the referee should watch the game?
Will they decide that a byoyomi period lost via lag can be reclaimed? (Both the AGA and EGF support immediate protests related to clock malfunctions in any one game)

_________________
North Lecale

There are so many discussions going on, in different places, turning in every direction, that it is really hard to distinguish between what are official facts, un-official but (somehow) confirmed facts and rumours.

IMHO, the main source of anger of so many Go fans comes from the fact (is that actually a fact?) that the referee did judge the way they expected - resuming the game - but then, due to a protest from the AGA-team, changed his mind.

When I was following the game and then heard that Mateusz actually made his move in time (confirmed by the proctor) I thought that we were lucky that this lag-issue happened in such a clear-cut situation, where it couldn't cause any damage - I was so wrong

That's why I would really like to know what arguments this protest contained, that made the referee change his mind?
So far, I only found things like "it was not the job of the proctor to control time" or "Mateusz should have complained earlier about lag-issues"
Anyway, all "arguments" I have found so far (fact or rumour?) appear to me like dodges used by lawyers to bail out their client, at whatever cost and by all (legally available) means. Of course, this is a lawyer's job. His job is not to be fair or to act sportmanlike, quite the contrary.

But I can't quite believe that such "things" were used by the AGA-team??
I really hope that I won't have to say once more: I was so wrong...

Saying that the referee “did judge the way they expected” depends on what you consider to be a decision of judgment.

Yes, it’s true that, according to discussions that happened behind closed doors, it appears that the referee had a change of heart. According to Hajin, this was due to added information that the referee received.

But the fact of the matter is, no official judgment was made until the end when the decision was announced to the public. If the referee now changed the judgment after the public announcement, that’d be a different story.

—

The stuff that happened before the public announcement was not made public, so if we want to speculate about the decision making process, we have to go off of rumors or what interested parties have told us.

In my view, there was a single ruling - the one that was officially announced. How they got there has been partially described, but we are really just gossiping about it since we aren’t part of the discussion.

_________________
it's be happy, not achieve happiness

I hope we get the chance to meet in person someday, Javaness - I think we’d be friends

_________________
it's be happy, not achieve happiness

It is that added information that I would like to know. This case is so clear-cut that I really can't see how there could be relevant information that was not known the moment the proctor confirmed that Mateusz made his move in time.

Ok. IMO, the issue is more polarizing than clear cut - almost like politics or religion. Yours is one perspective that many share. This thread shows evidence that it’s not the only perspective.

_________________
it's be happy, not achieve happiness

First, according to the conditions of contest, written or not, Mateusz did not make his move in time, since the official record of time was kept by KGS. (A lousy decision by the organizers, IMHO.) OC, the referee could have overruled the official record, and apparently did so at first. The referee has a good deal of discretion.

As for what the added information was, which should have been discovered before the initial ruling, I don't think has been revealed, but is easy to guess, given a principle that referees and TDs follow in a number of games and sports. The principle in question is that players have a responsibility to protect themselves.

For instance, suppose that I am playing a ko fight and the opponent takes the ko back without playing a ko threat. We have both been lost in thought and I play a threat and then take the ko back. At this point I realize what happened and call the TD. The TD rules, correctly, in accordance with that principle, that by continuing to play after my opponent took the ko back I tacitly condoned his infraction, and have no recourse.

In this case, Mateusz continued to play with serious net lag for a long time, without complaining to the referee or organizers or asking for relief. That is the new information that could have affected the ruling, as it can be construed that by doing so he accepted the KGS time record as authoritative, even with netlag. OC, one can argue that each instance of net lag was a new irregularity, but the referee need not accept that argument.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

Ah, so there are some unwritten conditions?

According to rules that are written (japanese rules), Mateusz did made a move (as was corroborated by his proctors), but KGS server failed to recognize that.Referee job was very simple, to question proctors and outrule KGS fail.I guess we will ping-pong that matter all the way, till some admin/mod will close this topic.

There might be another argument hidden in the post of Hajin Lee on facebook:
viewtopic.php?p=244202#p244202

edit:
Probably not. At least I deduced it from a post of Justin Teng:
viewtopic.php?p=244736#p244736
"I think the main conclusions that should be drawn from this whole situation are that detailed tournament rules should be in place from the beginning"

(It looks Justin Teng is well informed and has a direct connection to the AGA team. He was able to correct statements of Mateusz Surma:
viewtopic.php?p=244744#p244744 )

Could be. After a few moves it would be apparent to all that KGS was keeping the time. No need to write that condition down.



I don't know about that. It is apparent that he made a move, but it is also apparent that he did not make it in time to prevent his byoyomi period from expiring on KGS.



As far as I am concerned, I have no dog in this fight. But as a TD, I have a good idea what may have happened, and I can share that.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

So, what may have happened?

So..let me extrapolate on this idea...

Assume that someday I should happen to play in some event in which Kim is the referee, and I happen to play the same player twice ( maybe the second game is a playoff game ). And further assume that in the first game I take a ko without making a ko threat, and my opponent does not notice, and I get away with it. My opponent realizes that I have snookered him, and that a complaint would not change the result, so he does not complain.

Continuing this line of assumptions: in the playoff game, I try it again, and this time my opponent realizes it before he makes his move. He insists that I have made an illegal move. He calls the TD over to the board. Kim listens to both sides, and says 'You know by now that Joaz cheats, and you didn't complain last time, so his move is okay."
My opponent is disgusted, but he respects the decision and makes his ko threat, I defend, and he takes the ko. I again take back without making a threat. Again he calls over the TD. Again, Kim tells him that he missed his chance. So I win the ko, and the game.

It seems logical to me.

:-) Joaz Banbeck made a good point.

But let us analyze it a bit more:

Let us think about reasons how a referee may reject a complaint:
a) You didn't complain before, so you accepted status quo
b) You didn't complain before, otherwise we could provide treatment for the problem
c) You didn't complain in time, you ask to nullify one half of the game

Now let us stock up some examples
1) the example of taking ko too early, in same game, by Bill Spight
2) the example of taking ko too early, in another game, by Joaz Banbeck
3) the case Eric-Mateusz
4) my example that I will describe now

Imagine a classic real tournament. The organizers chose a small playing area, and so there is not enough room for tables and alleys between them. Alleys are extremely narrow. in order to walk in alleys, people must touch tables.
I am in byo-yomi, someone walks around and shoves to the table slightly. Stones shake themselves, but they stay at their intersections and I can play and press the clock in time.
Later someone walks again and shoves to the table more swiftly, all stones are moved, they must be rearranged and I cannot play within my byoyomi limit.

Now let us compare possible argumentations of a referee:

In situation 1) only c) may be applied
In situation 2) no argument may be applied. The Argument a) is invalid
Situations 3) and 4) are very similar. Myungwan applied b), but I think it has serious flaws: it was the organizers who chose KGS / a small playing area, they should expect problems with lags and people walking between tables. And players knew that organizers knew, and so it was natural that players would try to manage non-ideal circumstance by themselves, if possible

No, because making an illegal move is an infringement of the rules. Your opponent is responsible for protecting himself, but does not have the authority to change the rules. Neither does Kim. (Edit: Condoning an infringement of the rules by continuing to play does not make the next instance legal, because the rules remain the same.)

KGS is not doing anything wrong by keeping the time. Indeed, as the organizers set the tournament up (although I hope they change that before the next game) KGS was supposed to be keeping the time, regardless of netlag. The question is whether Surma accepted that condition, despite obvious and ongoing difficulties with netlag. Kim had the authority to overrule the loss on time, because of the netlag. But he had no obligation to do so.

In my opinion, had the netlag been a single incident, the principle of restoring equity would not have been in conflict with the responsibility of Surma to protect himself, because he brought the problem to the attention of the referee or organizers when it occurred. So I would have overruled the loss on time. Apparently that was the first ruling in this case, as well. But once it became clear that Surma had not brought the continuing netlag problem to the attention of either the organizers or the referee, the two principles came into conflict. OC, it was possible to regard each incident of netlag independently, but they were not really independent. Furthermore, the tournament had been set up (erroneously, IMO) without any provision for netlag. The referee's decision could go either way, but the default is not to overrule the official KGS time, and I suspect that a lot of TDs and referees would have declined to do so.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

Virtually all of my notes about this incident have addressed that question. Here is a recent one. viewtopic.php?p=244875#p244875

Kim consulted people with experience running online go tournaments. I suspect that he got advice to the effect that, by not complaining about the continuing netlag, Surma had tacitly accepted the KGS time as authoritative. The way the tournament had been set up, the KGS time was the official time.

_________________
There is one human race.
----------------------------------------------------

The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

MS does not suggest that he had continuing netlag. He states that had one incident before. You are aware of that already.

We could accept the judgement that KGS's clock is the absolute truth, but clearly nobody had previously decided that they were. There is no confirmation the previous netlag incident was observed y a third party, hence it can be considered impossible to objectively rule upon. (Just like a misclick would be.) Here you had an instance where netlag was observed by a third party, it could be ruled upon. The nature of that ruling will probably only become clear when the new rules appear for the remaining matches.

_________________
North Lecale