New Correspondence Timeout Rule?

One time out = one loss, very simple. However, if this is not a straight forward change technically, don’t worry about it and the development time is better spent somewhere else, because rank is not that important in my opinion, and one win/loss is statistically insignificant to one player.

To add, I also check if a player has timed out recently and try not to play such players.

3 Likes

Problem with corr games is that sometimes life happens, and you just lose a bunch of them in a row by timeout. This happens to even the best of us, it’s unfortunate, but not unheard of, and certainly not usually on purpose. Losing even 5-6 ranked games by timeout would drop a players rank by few stones, while losing 20 would absolutely decimate the rating of a player.
Would you want to play a ranked game with an opponent, knowing that he/she is 10 stones stronger than you, but you still need to give him/her handicap just because the system is penalising them for some unfortunate lifesituation? Or would you feel like you yourself would be punished on previous timeouts of your opponent?

Getting the T is not only a protective measure for the player who loses a bunch of games on timeout due some random life events, but it is also to protect everyone else who will play with him/her in the future, and in a large-scale it’s there to prevent the rank volatility of the whole player pool of OGS.

3 Likes
4 Likes

Corr games have no difference than any other games. When your clock runs out, you lose. It has nothing to do with your life. Losing a game is not even a real punishment in life. Why over analyze it and make things unnecessarily complicated?!

Playing an accidentally game with someone 10 stones stronger than me, so what?! I will resign within first 50 moves and probably learn a thing or two.

Do I just open the topic of not resigning because … lol

4 Likes

Yes, as @txwolf points out it’s not a life-related problem. If you timeout you lose, and I didn’t say anything about rank-effects. Just that players that can’t finish a game - loose. Always. It’s about respecting the opponent. And the player who caused it can keep their rank, that’s not the point! It’s the frequent abuse using timeout as a normal finishing strategy that I’m after. And a player with multiple timeout issues during the last month is not someone I want to play with simply because they are wasting my time. Again, use vacation or resign. Simple.
Please don’t overcomplicate and make it a rank-affecting problem.

1 Like

But that is precisely what this thread is about. At least that’s my impression when I read the first sentence in the first post.

2 Likes

Yes, technically @richyfourtytwo, if you insist on missing my point. I’m just trying to explain that the bigger problem is that players abuse the system deliberately. But for a Go server to take into account players life-world problems that’s accidentally behind the timeout, I think is taking things a bit too far. And the rank-effect caused by a timeout doesn’t address the actual problem. And I’m just suggesting a simple solution: informing players of the timeout tendencies of others, so that the risk of wasting other players time is minimized. I for one do not want to waste time on games where my opponent have a track record of just ignoring games on a regular basis. I don’t care why they do it. If you or anyone is ok with playing an opponent who has let more than say five games finish w/ timeout you can choose to. (And yes, maybe a discussion for another thread, but come on…)
So _KoBa, no, of course the integrity of the ranking should be as good as possible, but I want to be able to stay away from players who can’t finish a game by points or resignation. Not endless waiting and anti-climax.
@AdamR, any thoughts?

2 Likes

I read the embedded thread and understood a little more about the rule: to reduce fluctuation of the ranking system.

  1. Is there evidence to support this rule improves the system? Will you twist again when users’ behaviors change accordingly? Simple is better in my book whether in life or software development.
  2. Does it worth to improve an abstract system at the expense of user experiences? I say no. The accuracy of the ranking system we don’t see and feel that. Users’ frustration is real. Why do we make our users, maybe just a minority of them, happy first? It does not hurt anybody else and it is fair in my view. More importantly, why do you developers, who have done so many wonderful things to provide us this platform to enjoy Go, step up voluntarily to take the blame? I don’t really get it.

Again, I don’t care one way or the other. That’s the end of my opinion. :sunglasses:

3 Likes

I will leave this line of argument as well. I’ve addressed the real issue as best as I can, and I don’t have the capacity to explain it further. Maybe I’m the only one suffering from this.
Good night everyone.

1 Like

Yes. Before the current timeout rule was in place, it was common to see players whom ranks were way down to 30k because a string of timeouts, regardless how strong they were. I guess there is no point saying that it did mess up the rankings a lot.

I get your point, but those correpondence games can (and will) go on months, or even over a year, and no one can guarantee that they won’t get suddenly hospitalised, lose their internet service provider, or anything like that for the next year.

The main point being, ranks and ratings are only tools for finding equal pairings for games, they are not points of respect nor indicator of reputation. Having rating points demolished due a string of accidental timeouts won’t serve this purpose at all, if anything it only distorts the possibility of finding equal match-ups in the future.

1 Like

Their ranks dropped because they deserve it, no? And that is what system is doing what is supposed to do.

What if I get sick tomorrow and request to cancel all my games without affecting rank drop, will you process it?

Before this rule in place, it is simple, nobody has a good reason to be unhappy about it. Now a few people is unhappy, rightly so in my opinion. The site’s customer satisfaction level on this feature drops a little, not much. And you guys think you did a smart thing and are willing to spend time answering this same question again and again. lol

Bye, if I come back reply this topic again, ban me. lol

3 Likes

Yea, it’s probably best to just lock this thread. No point rehashing arguments we’ve all heard before.

5 Likes

To anyone still wishing to debate / better familiarize yourself with the issue, please read at least comment #136 in the Escaping by Timout thread, where I tried to summarize the majority of comments.

As not to be rude to a direct question, let me also answer to @Meegus, I assume the question was about the idea of displaying player’s monthly timeouts. I am not on a dev team and the decision or implementation is thus not on me. Strictly personally I am not really a fan of cluttering the (already complicated) interface with other non-essential data. I think if anyone wants to quickly check a last month’s worth of games for timeouts it is not that many clicks away.

But I am also quite oldschool in these regards of “public shaming” (notice the quotation marks). I know for example on wbaduk (I think it was) you can literraly rate the player you played with (which is kind of similar in idea). I can see the thought behind it, but personally do not like it.
I believe that a player should either feel completely welcome here (even after some minor transgretions) or not be here in the first place (when unable to stick to our rules of conduct).

3 Likes