What is the definition of a "move" in an Internet go tournament?
(a) For the player to identify a place that s/he would like to play, and click on that spot; or
(b) The selected intersection is received by the server.

Mateusz did not make the move on time, by the definition in (b). So what? It's not in his control, folks say. But many online tournaments don't care - I brought up poker; it can be the same for some online chess tournaments, too. It's not uncommon to adopt the policy of: "you play in this tournament understanding the risks that lag may occur; timeouts may happen. These exceptions to not change the result". This is a perfectly reasonable policy. Lag sucks, but it's a reality of online gaming.

But maybe we don't like that policy. Maybe we'd like to be more forgiving for the (a) scenario, above. Okay, sure. Then let's make a policy for it.

But we have no policy like that, and there is no automatic reason that it has to be that way, if you're trying to argue from a "legal" standpoint.

_________________
be immersed

According to the EGD, Zeijst lost against Dinerstein in Tampere - his only win against Dinerstein was at the 9th European Fujitsu cup (http://www.europeangodatabase.eu/EGD/To ... n=10337382).

_________________
be immersed

Maybe we have different definition of what means "punished", but its clear that his situation got worse after he told truth about lag in earlier game.I call it punishment, you can call it differently, but this is undeniable fact.

Well, you are again assuming sth that is not explicitely written. You wrote " Getting the result that is in accordance with the conditions of competition is not punishment. " .I dont see any condition of competition which are in accordance with ruling. You consitsensly ignoring fact that KGS was nothing more than venue (and that is actually explicitely written) exactly like some mall in center (having game of go in the center we dont follow center rules of playing go if we already accepted different rules). Rules under game was played were japanese rules, and under that ruling referee decision is completely off it (well, i assume it means 1989 rules, but i doubt that 1949 rules says anything about lag or clock, if 1989 dont say a thing).

2nd referee ruling was that it should be rematch (also debatable, but lets go on). But after learning that Surma got lag in earlier game he changed ruling to lost by time. This is point which is completely wrong (and must be because as you pointed lag is not mentioned in tournament policy, so change in ruling must be without accordance to it).If tournament policy would have lag component, then it could be true.

The difference between 1st and 2nd ruling also is very suspicious. How in all reality someone can be sure player is not cheating and not watch moves? Do Lui proctor observed monitor or not? Also assuming that Lui proctor could not corroborate with Lui claim with possible lag, also looks bad. We should assume that Lui was unlucky that hes proctor could not testify, and rule as common sense and spirit of game apply (for everybody it means different things, personally i would probably grant resume with byo period restored and oblige proctor to observe KGS lag).

Yes appeal court in many places can do it, however you seems to not understand that the earlier game was over, and its completely indepedent of next game. Whatever happend in earlier game should not have any impact on next.

Final decision of referee violated article 1 of japanese rules, and its really sad.

@Kirby
There is no definition on move in japanese rules, but considering preambule of japanese rules, we should assume that its when player click his mouse and not when server get it.You assume to get definition "a" by default because game was played on KGS, but as i say again (and i guess again i will be ignored) game was played under japanese rules, and not KGS japanese rules.

I'll try to summarize the positions in a simple way. The original authors can correct me if I get it wrong:
- AloneAgainstAll's point is that, by arguing against the first appeal, Surma's ruling became more unfavorable toward him. So it could be viewed as a "punishment".
- Bill's point is that the default ruling is a loss by time, since they were playing on KGS and that's how KGS works when your clock runs out. We can try to make an exception to KGS policy, since we have a referee, but in the end, the referee agreed with the default KGS ruling. From this perspective, Surma was not punished - it's just that the referee didn't make an exception to the default KGS ruling, with the arguments that were given.

_________________
be immersed

Can you point me to where Japanese rules explain the nuances between server side move selection and client side move selection?

_________________
be immersed

The point is, AloneAgainstAll, we are trying to play an Internet tournament on KGS with a Japanese ruleset (which as you've stated, is ambigous --> 89 Japanese rules? KGS Japanese rules? which Japanese rules?). So it's not exactly the same as an in-person tournament, since the venue is KGS and not in person. But on the other hand, maybe we want to try to get things closer to an in-person tournament than to an Internet tournament.

We are somewhat in a grey area, since it's a virtual tournament played online - there are implications that don't exist in in-person tournaments; yet, we'd like to emulate the in-person experience. The problem is, nobody defined the boundaries of what should happen in this grey area. In situations where the virtual tournament is different than the in-person tournament, should things more closely match typical online games? Or should they be closer to the rules of in-person games?

Nobody really said either way. We didn't have a rule about lag.

I maintain, then, that one possible policy is to treat lag as you would in other online games - the server decides when a move occurs, and arbitrates the time. I also accept that there are other possible policies that would be closer to an in-person tournament - but it'd be an exception, since you'd be introducing a rule about lag that didn't previously exist. There exist tournaments online, as I stated, where the policy is closer to the online format - timeouts are a reality, and you should lose when you run out, even in the presence of lag. So it's not unreasonable to accept the same here. We don't have to. But it's also not unreasonable.

_________________
be immersed

Haha, you know that they dont.They dont have even article about losing on time!Nothing about clock, or clock malfunctioning.

Thats why i said, we should assume, not that we must do it (beacuse rules explicite state it). Japanese 1989 rules says, that rules must be applied with spirit of good sense, while article 1 stricly says that game of go is contest of skill, and not inernet latency or any other external factor. Taking that into consideration, i stand with what i wrote earlier.

I wish that had been the case. If so, they would not have used KGS to keep time. My proposal is, in fact, to have time kept for each player locally. One easy way to do that is to keep time with a real clock.

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

Visualize whirled peas.

Everything with love. Stay safe.

This is the root of the problem. With no explicit rules common sense, reason and more general or broader rules apply. This is why I think (and I may not be alone, given the extent of public objection against the final ruling) that losing on time when I completely and perfectly finished my part of "making my move" - and this is even officially witnessed - is out of question. Even from a "legal" standpoint.

Scenario (b) would imply that regardless of byoyomi, the first player to lose connection loses. Or that he loses if his time runs out earlier than KGS detects the disconnection, which gets dangerously close to random.

I think we all want to make a decision in the spirit of good sense. There are just different opinions about what makes sense here. I don't like lagging out on KGS when I play rated games, either. But I still consider it a loss on time if I lag out, because the KGS timekeeping is the definition of the clock. Yeah, it's just one possible interpretation. But I don't think it's outside of the spirit of the game - it's just how I see the clock on KGS.

But it's ambiguous.

In the current case, I've changed my mind a few times about what I think should happen. Because it's ambiguous. If you look at my earlier posts on the subject, I was of the opinion that, if we can trust Mateusz's word, we should resume the game. Then after reading different opinions, I thought Eric should win by time, because that's what KGS says. Now I kind of think that, due to the ambiguity to what a clock is, etc., a rematch would have been best.

But we have a referee for the tournament. And their job is to arbitrate when things are ambiguous. So I guess it's best to accept that decision.

I think that this would be in the "spirit of good sense".

_________________
be immersed

I just don't understand how you can make a "legal" argument in the face of this ambiguity

_________________
be immersed

I think within quotes it goes.

"OK - fair enough!"

_________________
be immersed

I think you exaggerate organizers skill. I can bet 10k KGS bucks, that they did not thought about anything what can goes wrong. Lee wrote that they considered disconnection. Do you see any policy for disconnection written in tournament policy?



I completely agree with locally time kept. Also with proctors watching game we could be pretty sure that nobody is meddling with it.

If game would be played with Fischer time, possibility of such an incident would be less i think. Unfortunately KGS does not support Fischer time.

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.

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

Visualize whirled peas.

Everything with love. Stay safe.

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.

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

Visualize whirled peas.

Everything with love. Stay safe.

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.

_________________
be immersed

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...