I am french, and like the vast majority of my countrymen I suck at english. ( Ok, I have been watching "I suck at Starcraft 2" thanks to the SF2 thread )

Not fair, there are many times my opponent escaped me, and you can not see from people's game list who escaped first.

_________________
长考出臭棋.

escaper should just not allowed to play another game unless resign the previous one.

_________________
长考出臭棋.

For sure they are different regarding their meaning, but the pronunciation is quite the same. Similar problems exist with "it's" and "its", or with "there", "their" and "they're".

Not only non-native speakers are prone to these mistakes.

If we just take "you're", are you not aware when speaking that you are actually saying "you are"?

_________________
A good system naturally covers all corner cases without further effort.

I think these mistakes are are almost always made by native speakers. Not having to translate their thoughts when typing lets them enter 'acoustic' mode
I remember some other discussion where people actually defended their use of 'should of done something'.

_________________
To sig or not to sig, that is the question.

- Experience tells that in large majority of cases where the leaver isn't disconnected, he left the game in rage.

System's worthiness is no greater than its "weakest link", or biggest problem. Not features as such. Imagine a game that was fun to play, but it would occasionally crash to blue screen of death on some systems. Surely the company would make it their top priority to fix this issue. Now in KGS there's a feature to adjourn the game you're playing. Unfortunately it's being used maliciously to prevent loss record. This is causing distress and confusion to victims. Will the person be back? What will I do with this italics game on my record? Perhaps he makes a public accusation and gets ridiculed by veterans - like this very thread proves. At the very least the victim will call admins to explain the escaper policy for their hundredth time.

Won't all these consequences prove that escaping is at least a problem? Just saying "deal with it" will solve this one case, but new cases are being made daily, because the system is unintuitive. I'd call a plumber to fix the leak and have peace of mind. It's of course up to wms how big leak he considers this.

Do people ever adjourn rated games? Is the feature popular enough to justify the impact on the escaper problem?

Adjourning free and teaching games seems fine. Adjourning rated games seems like a possibility to cheat by giving yourself more thinking time. It seems perfectly reasonable not to allow this, or at least to require the opponent's permission.

My ideal solution - an [Adjourn] button, pressing which asks your opponent whether he's willing to adjourn the game. Closing the game or losing your connection doesn't adjourn the game, but activates the 5-minute timer, after which you forfeit. Is this how IGS works?

Your analogy to software in this case is poor, perhaps a better analogy would be "there is a game that is fun to play, however in a small percentage of time when it is played, there is a delay in recording the score obtained to the high scores list." There is no loss of program function due to escaping. Beyond that, this type of argument seems to me to still go back to the "it is a problem because it makes me angry." Instead to establish that it is a problem you need to lay out what criteria you want to use, and then determine whether or not a system can achieve the desired outcomes.

Personally, I've attempted to recall / estimate the criteria that were kept in mind when revising the KGS escaper system, and base a judgment using that. It may be that you disagree or think other criteria are important (which is perfectly reasonable), but I think in order to establish something as a problem, it would be nice to at least list proposed goal it is failing to achieve, or perhaps present a factor that was overlooked and argue for why it is important. Presenting a hypothetical anecdote to declare something a problem without a listing a realistic means of measuring or evaluating it makes it much more difficult to evaluate what might be the best way to solve said problem.

Now this was something I had meant to get back and respond better to, but I've been caught up in other things (=

While you don't come right out and say exactly what you are evaluating the systems on, you imply it pretty strongly by stating what you prefer...if I had to guess I would say you're looking for something like this:


1-All games should should be resolved, in a manner as quickly as is practical.
2-It is a player's responsibility to actively demonstrate willingness to continue the game, or else forfeit.
3-1 supercedes 2, except in cases of mutual agreement. (To allow for adjournment only if both agree)
4-In cases of dispute, the judgment of a disinterested 3rd party may be used to decide the result a game.


Anything I got wrong, or something else you care to add?

If this is close, I'd say it sounds pretty reasonable (personally, I'm not too keen on 4 but that's just a difference of opinion). I'd still be interested though in hearing more (from you or anyone else who cares to answer) about why you like these in particular (or if I got them wrong why you prefer the ones you do prefer).

- Leaver accounts can be abandoned too. However the point wasn't score, but unhandled termination of game where the other player is left wondering what just happened. Analogy with software crash is quite appropriate in this respect.

There are many solutions to this. One is the research on typical disconnecter behavior versus typical escaper behavior. Disconnecter loses connection; quitter presses the X button. Disconnecter wants to get back to finish the game; quitter doesn't. Adjourning a game makes sense only with both players' consent. Based on this information we can start proposing solutions to discourage escaping behavior while causing minimal harm to those with bad connections. I assert that the solution has to be in code. Else we'll continue to have these escaper threads no one likes, right?



- Those four tenets were all reasonable. I don't disagree. While that's true, it might be worthwhile to discuss why the current system creates so many complaints about escapers. Might escaping be the most discussed problem in English Game Room?

I'd like to help, but this topic is surrounded by negativity. Many people complain about the system without offering helpful feedback. I understand that it's a frustrating experience having to deal with escaper. Been there myself a couple of times.

Those criteria seem to correspond to my preferences pretty well. And like you, of the four listed, I like the first three the most. I like having #1 as a priority because, if all games are resolved, escapers don't exist anymore. The other criteria allow for a cushion that lets people to adjourn games, if desired.

I guess I would also add a fifth criterion, which says something like the following:
5-The system should allow for a way to deal with Internet disconnection.

However, this could also be covered by #1, depending on how "as quickly as is practical" is defined.

And again, the 4th rule is probably the one that I like the least. It is mainly useful for cases of disputed shapes on the board, or the players that used to play stupid moves against rapyuta, wasting everybody's time. But overall, if a system were to implement the other criteria, and not do the 4th one, I'd still think it was pretty ideal.

Edit: Oh, I think I forgot to answer your actual question: why are these particular criteria important to me? I think they're important to me because I think they're fair. Escapers are eliminated, people have the option to pause the game and resume later, and Internet disconnects are handled. I don't see any negative aspect to these criteria. It's probably because they're based on my preferences, but that's true of most of my opinions.

_________________
be immersed