Time settings?

[MarkSteere]

I'm completely flabbergasted

I think a case can be made for strong feelings on the time setting issue - in particular, Gabriel's handling of the issue. People have long anticipated this grand, superior Go server. They've invested time, good will, hope, and $13,000. Now Gabriel has installed a universally unwanted feature, and he's totally unwilling to back down from it. He's doing semantic somersaults to not acknowledge that there even is a problem. At most, he'll ooch the time limit a little, which won't really accommodate the wishes of the budding membership. It's Gabriel's server and he's going to do it his way. Period.

Gegenzeit, you've essentially repeated the same argument that others have already made in this thread. Doesn't it bother you knowing that your sound rationale will be ignored simply because Gabriel has already made up his mind?

I would not mind the L19 mods being way more aggressive in dealing with it.

Crushing the discussion isn't the solution here. People are strongly disappointed in Gabriel's ignorance of their (including your) reasonable, intelligent argument - for no apparent reason other than to be in charge. I think people were justified in making their feelings known. I'm not even a Go player and I'm disgusted.

[jts]

I'm not even a Go player and I'm disgusted.

I think that's precisely the point. You may not have noticed, but we have a community here which is based, for better or worse, around a shared loved of go. Things may seem to get feisty or spirited from time to time, but even when we disagree with each other, we're on the same team, and that's a bond that helps us understand how to communicate even when our words get pointed.

As best I can tell, you've never played go and so have no stake in a debate about how time systems and rating systems interact; and your main interest is in insulting Gabriel because he won't create an online application where people can play a game variant you created. Perhaps if you had spent the last six months enriching our discussions about go here I would have a little more context for your words, but as it stands I agree with Gegenzeit... your disgust (I would say "venom") is rubbing me the wrong way.

[Kaya.gs]

As is today rated games have a "practical limit" of 5.5 hours.
When the game is longer than that, the game will have to be deferred to be rated, which means:

1) The players are not locked into playing that single game
2) The players can disconnect for any time period they wish.

You will be able to play rated games at any time you want. What is the problem? Who is being punished?

I don't really Undertand this post. Does this mean that you can play up to 5.5 hours? what does deffered mean?

Today as it is, there is a 30 minute main time limit. Byoyomi period limit is 120 secs/10 periods max.
That allows for a much slower game than 45 minutes. Granted, the distribution is a little different. I can increase but note that the longer the time setting, the more likely the ocurrence of stalling.

Deferred is one of the terms i like for "Turn Based Go". As i explain in that post, a deferred game can be played without the players being online at the same time. Much like topazg's OGS, or DGS. Of course, this is exceptional for long time settings like days, but in a live server, its very applicable to, for example, 2 day games.
The other main perk of deferred games is that because being online isnt a requirement, its not a problem to set up a few games at a time.

We want to do some cool and innovative thing with that, although there are more important things to do first, like improving the teaching tools.

The only issue with long main time settings is that if your opponent for example, has any kind of issue, good or ill-willed, you cannot play another game in the meantime. This means that you are locked from playing at all, potentially for as much main time as there is. This is my main concern regarding very long main time games, that demand players to be online and attentive to a game that might not finish(in go terms).

We can try to expand the time settings and wait for issues, and im up for that. I will increase it to 60 minutes for Beta 1.2.0. With that, the average max-time with Byo-yomi increases from 5.5~ to 6.5 hours.

The whole point is that the game is both fun and meaningful for both players.
A game interrupted like above is detrimental to both the rating system and the experience. Its not about being a nanny, but protecting against a very simple case that could not only ruin a single game, but prevent a player from playing at all while it happens.

Moreover id like to continue this discussion live on the server, so if you get online,speedchase, lets talk there. A chat is more fluent. Several people with accounts contributed to this, and it is known I'm very up for debate and discussions into finding whats best.

We have the largest number of feedback channels than all servers, with UserVoice, Tweaki, chats/live support, this forum and email, and we answer all of them promptly.
Let's not lose sight that the purpose is to become better, and provide the best. Lets keep the positive and forward-thinking attitude.

Regards.

[hyperpape]

However , and i note again, 30min 50/5 would yield a similar time game.

I don't think it's much of a defense. You're saying that 30 + 50/5 is basically the same as 45 + 40/5, but you're treating them differently. That's a paradigm case of an arbitrary rule.

[Mivo]

The whole point is that the game is both fun and meaningful for both players. A game interrupted like above is detrimental to both the rating system and the experience. Its not about being a nanny, but protecting against a very simple case that could not only ruin a single game, but prevent a player from playing at all while it happens.

As Nick mentioned, that's a proactive mindset rather than a reactive one. You're basically designing limitations around hypothetical scenarios that theoretically could happen, but may never actually happen, at least not in the frequency that you anticipate it. Even if it did happen to a degree where it actually becomes a problem for a substantial number of people, you could then put in the restrictions on the fly (starting with applying them only to newly created accounts, then expanding the measures dynamically as needed). Your ability to respond to issues quickly is the advantage of not being a large, heavy corporate institution.

I guess I'm just a supporter of a "less is more" approach when it comes to rules. The more rules, the less a community it becomes. The more homogenized, filtered, protected, engineered, and regimented, the less "real" it becomes. Less control is more choice, less intervention is more power and responsibility for each individual to handle what they can using the tools available (ability to report "griefers", set filters for acceptable time settings, etc.)

[MarkSteere]

As best I can tell, you've never played go

I've played Go a few times in my life, though I've never won. The last time I played I resigned partway into it because I simply couldn't stand it. Dare I say I was disgusted?

[lemmata]

From what I can see in the thread, other people who disagree with Kaya.gs are being cordial and reasonable in their discussion here. You seem to be the only one intent on fanning some nonexistent feelings of antipathy here. Why?

People have long anticipated this grand, superior Go server.

No. People have anticipated a new go server that can be accessed by a web browser, even a mobile one. People have anticipated a go server whose behavior is not affected by new versions of Java. People have anticipated a go server that can be accessed on iOS devices and Android devices without the need to develop a separate app for each. People have anticipated a go server with some other features that might not be available elsewhere, such as Fischer time. Very few people would have expected Kaya to be the perfect server, especially given that each person's idea of a perfect server is different. What reasonable people expect from Kaya is a go server with the set of features that have been promised on the Kaya.gs web site.

They've invested time, good will, hope, and $13,000.

No. People have donated $13,000 for the developers to use to develop the server according to the developers' vision. There is no formal investment in the sense of fiduciary duty here. Yes, I am aware that investment has more than one meaning. You are using the word here to give us the impression that the developers have some moral responsibility to develop the server in ways other than they have described on the donations page.

Now Gabriel has installed a universally unwanted feature,

No, we don't know if it is universally unwanted. I don't want it either, but Kaya has a couple hundred registered beta users and the number of people who have posted in this thread are...how many? We just don't know.

and he's totally unwilling to back down from it.

No.He has suggested a couple of compromises in this thread. There have been no indications that he will continue to stick with the current policy if it generates problems in the future. I do like topazg's suggestion from earlier though.

He's doing semantic somersaults to not acknowledge that there even is a problem.

Nope. First, he is just presenting his side of the argument in detail. Second, there is no real problem yet. The server is in beta and rated games on Kaya don't matter right now in the big picture. He outlined the theory behind his initial decision very carefully. I disagree with his premises, but he presented a coherent theory whose flaws won't become apparent until the server is further along in its development.

At most, he'll ooch the time limit a little, which won't really accommodate the wishes of the budding membership.

No. You don't really know that. We know that some members don't agree with this (myself included), but we don't know about what most of them think. There are a couple hundred. Also, the server is in closed beta. If you understand what that means, then you will know that the membership is not growing at the moment. Not many games are played on the server at the moment. Rated games with 45 minutes of main time is not so common even on KGS. While the arbitrary limit is annoying, I doubt that too many will be upset enough to quit because of it.

It's Gabriel's server and he's going to do it his way. Period.

Yup. Indeed. The developers own it. It's good to see that you recognize this, but you seem unable to draw other conclusions related to it.

Gegenzeit, you've essentially repeated the same argument that others have already made in this thread. Doesn't it bother you knowing that your sound rationale will be ignored simply because Gabriel has already made up his mind?

I hope you realize that it is possible for two opposing arguments to both have sound rationale when there are many unknowns.

I've played Go a few times in my life, though I've never won. The last time I played I resigned partway into it because I simply couldn't stand it. Dare I say I was disgusted?

What brings you to our neck of the woods? Actually, what makes you think that it is polite to spit on a game that we enjoy? You didn't say that you disliked the game. You said that it disgusted you...

You seem to be the only one intent on fanning some nonexistent feelings of antipathy here. Why?

Never mind.

[Xa17u]

I've played Go a few times in my life, though I've never won. The last time I played I resigned partway into it because I simply couldn't stand it. Dare I say I was disgusted?

By what? Not knowing perfect play the moment you took a stone?

[speedchase]

With the new 5.5 hour practical limit, wouldn't it be possible to remove the arbitrary limit all together?

[tchan001]

The way Mark Steere is painting the picture about kaya.gs on this thread is disgusting and only makes me want to avoid playing Redstone.

[topazg]

[mod]Please can we stick to the topic at hand, which AFAICS is Kaya's implementation of time setting limits and how they impact a game's rated status.[/mod]

===============

With the new 5.5 hour practical limit, wouldn't it be possible to remove the arbitrary limit all together?

This isn't a new limit at all by the looks of it, it's the fact the limit is geared around a 10x2:00 byo-yomi, which is insanely slow if people use all their time - the main time section becomes pretty much meangingless. It's certainly a much slower game than the one that was in the OP that wasn't allowed to be rated.

Gabriel, do you actually use a formula to hypothesise about the usage of byo-yomi, or is the cap exclusively limited to main time? If it's the latter, your implementation won't really even achieve what you're hoping it will for those that actually want to drag out the game.

Even 1 minute main time + 10x2:00 is a really really really slow game, but would your system suspect it was a blitz game done strangely?

[SpongeBob]

Two things have now become clear to me:

1. The issue of being locked from a new game until the present one is fineshed (i.e. the time has run out)

2. The fact most of us have not yet encountered the abuse problem is that when you play on KGS, it would make no sense to use the 'non-playing-strategy', because your opponent would probably just leave the game. He would get one escaper mark, but you would not get the win you were trying to get. Without the adjournment functionality the situation could very well be quite different. I always wondered why Tygem games are normally quick games and 20 mins seems to be already regarded as kind of a slow game. Maybe I now know the reason.

So with this, I can understand Gabriel's rationale and think it actually does make sense. However, instead of just increasing the 30 mins, I still think it would be better to display a confimation popup with a warning message, explaining the situation.

[topazg]

I think if the issue is a lack of adjournment, then the implementation of adjournment for real time games makes sense as a next step in this issue - it's a desirable feature anyway, and one I've used quite a few times on KGS.

Regarding escapers, I would like to suggest the following proposal - if people want to discuss it, I'll move it to a new thread for its own discussion:

1) If a player disappears, the game is automatically lost by timeout if he doesn't reconnect within 15 minutes.
2) This is prevented by the remaining player adjourning the game, at which point time is stopped.
3) The remaining player can "restart" the game at any point (regardless of whether the opponent is present, but it gives the players a chance to find a suitable time for both).

That way, people can't escape to avoid a loss, the remaining player isn't forced to fill up a list of unfinished games, and we deal with the time limit cap issue as the remaining player never has to wait longer than 15 minutes.

Remember 1) doesn't mean "hasn't moved for 15 minutes", it means "has left the board for 15 minutes". I don't think there's any problem spending 20 minutes on a single move in a difficult game.

[quantumf]

I'd like to offer the point of view that I find KGS's arbitrary and random time limits incredibly irritating. At any one moment, there might be a game at 5+100*.10 or 30+5*30 or 10+40/10 or 0+1*0.l0. These all require a fundamentally different time management strategy, which really rankles. I don't want to have to choose games based on whether a potential opponent is offering a normal game or an out-there game. I would prefer a standard time limit, that I can get used to, and after a few games, time management will become a background problem, leaving me to focus on the Go.

Note I am referring to rated games here, free games can be anything.

I would probably consider allowing three variations - fast, normal, slow, with some predefined time limits. They could be rated independently.

I do not regard this as nannying, I regard this as focusing on Go rather than meta-Go.

For example, at our club, rated games are always 30+25/10, we don't allow players to pick crazy variations from this - or if they do, they are not rated games.

[speedchase]

I'd like to offer the point of view that I find KGS's arbitrary and random time limits incredibly irritating. At any one moment, there might be a game at 5+100*.10 or 30+5*30 or 10+40/10 or 0+1*0.l0. These all require a fundamentally different time management strategy, which really rankles. I don't want to have to choose games based on whether a potential opponent is offering a normal game or an out-there game. I would prefer a standard time limit, that I can get used to, and after a few games, time management will become a background problem, leaving me to focus on the Go.

Note I am referring to rated games here, free games can be anything.

I would probably consider allowing three variations - fast, normal, slow, with some predefined time limits. They could be rated independently.

I do not regard this as nannying, I regard this as focusing on Go rather than meta-Go.

For example, at our club, rated games are always 30+25/10, we don't allow players to pick crazy variations from this - or if they do, they are not rated games.

You do realize that you entire argument is just "I want people to have to play me, on a time limit i like, so I don't want to let them choose a time limit they like.