What? Conrad? Hyperbole? Never!
It used to be that the serial timeouts would show as losses but silently not affect ratings, right? I think the change to showing them as annulled dramatically increased the number of complaints.
Rage driven development
a thing that leads to many a questionable feature.
FWIW, when I implemented this (my “rookie” contribution) the problem I was trying to address was pretty orthogonal, but the conversation went something like this:
Me: “why did OGS just make me lose my game?? I just went to browse the site while my opponent was thinking?”
“Don’t worry, your game was annuled”
Me: “No it wasn’t what are you talking about?”
“If you scroll all the way to the bottom left of your profile, you can see it was annulled”
Me: “well can you fix it? I shouldn’t lose a game by Disconnection during my opponent’s turn!”
“OGS is open source, want to fix it yourself?”
Me: “Only the front end is open source. How can I prevent the server from ending my game?”
“I’m sure you can think of something”
And this is the solution I thought of ![]()
Touche’. Still, it’s a driver for progress ![]()
Repetition and hyperbole are devices of rhetoric, a component of style. See Marc Antony’s speech in Julius Caesar, or Churchill’s “We shall fight on the beaches” speech.
I admit that my experience is probably different than yours. I don’t really care about the rating points, anyway ranks fluctuate by ±1 and the most important thing is that they reflect the players’ strength. On the other hand I do find timeouts annoying (although not to the point that I’d consider to quit playing correspondence games): if I start to watch an interesting movie I’d like to see the end, even if I can guess the likely outcome of the plot. Changing the serial timeout rule may not affect much the frequency of timeouts.
There is a third way: the one advocated by @richyfourtytwo . Do what you have to do to protect the rank system, but don’t call it a win.
–
Ian
After reading most of this thread, I feel like the following, though it may be obvious / trite, needs to be said:
Correspondence games are as important as live games, they’re not some attic for weird / crazy players.
So for example, already the fact that @Conrad_Melville keeps mentioning, that a timeout in a correspondence game isn’t reportable – while presumably leaving a live game is – rubs me the wrong way. It really is the same rude behavior either way, as far as I can see.
–
Ian
Funny thing to say, since this is primarily a correspondence server and we’re fighting hard to make it a place also enjoyable for live players.
The situation is quite different.
For live games the player is sat specifically at the computer, likely for a specific limited window, and if their opponent goes AFK they have to sit and wait for them.
For corr players, they check a few games as they have free time throughout the day/week/month and then go about their day. If an opponent times out they may grieve the unplayed potential of the game, but their time is not taken from them in the same way.
AIUI, these are still ”wins" and you should still move up the ladder, so all you “lost” was a rating adjustment, am I understanding the situation correctly? Or was this not your experience?
Edit: sorry if this was already answered, I didn’t really follow the thread, just chiming in ![]()
I was the challenged one, so all that happened from a ladder perspective was nothing ![]()
But yeah it’s the lack of rating points that, and as genbeart said the feeling of rude.
It’s the same rude behaviour, as he said. The impact on the person at the other end is less, but it’s still rude.
That said, it’s not the feeling of “that was rude” its being robbed of “at last, now maybe rank up, ooo goody… ooo, I hope so … ooo surely they can’t get out of this … … … bzzzt ‘annulled’”.
Personally, I don’t want that again.
I have found this discussion really helpful even if it’s probably not been what @GreenAsJade wanted exactly!
We have established that it’s not a matter of cheating or maliciousness on the whole but rather of rudeness - people not following “expectations” - and then unexpectedness leading to a sour taste - I thought I was going to win and get full credit for that but didn’t.
Some people feel that corr time out is completely equivalent to live time out and others (including me) see it as @BHydden set out - there is not the same level of wasting people’s time as for live games.
I would like to note that we often don’t enforce etiquette (my understanding is that it’s poor etiquette to do so in fact). For example, some people feel that playing a lost game to scoring is rude and etiquette is that you should resign a hopeless position. But we don’t sanction people who play to scoring as long as their play is reasonable. (I realise the server decision thing can be a challenge to this but that is intended to operate to deal with unreasonableness and the discussions show how hard it is to draw a line)
I think that we have a system to deal with the technical difficulties of serial timeout in Corr but the system introduces a human impact. This is tougher to resolve but I would suggest that the “solution” (only a partial one) is to make the system feel as fair as possible - e.g. logging in interrupts the sequence - and being clear about etiquette and the good etiquette of accepting others rudeness with grace.
We can’t control others behaviour but we can control our own reactions!
I didn’t read the whole thread, but I like this a lot.
We do, actually. We enforce “not escaping in live games”.
But don’t we do that because it’s a kind of harassment or obstruction due to preventing people actually playing go?
Timing out of corresponded doesn’t do that
Laying aside whether we police etiquette, I think it’s worth reviewing the methods we use to address bad behavior.
Usually bad behavior is met with:
- A report
- A warning (or a few warnings)
- Suspension
- Banishment
Mods feel free to add to that list, but my main point is that we don’t do this as part of the moderation process:
- Deduct ratings points
Why? Because the rating system is meant to measure skill, not character.
So I think we can agree that serial-serial-timeout is bad manners, and bad for the site. But we should not be looking for ways to get their games rated, but rather use the moderation tools as they are used in every other case of bad behavior. (And hopefully correct the behavior or remove the bad apples)
No, it is not a matter of rudeness in the sense that you frame it (rudeness merely as rudeness). It’s a matter of habitual, presumably mostly unconscious, actions (escaping) that happen to be rude, but are much worse because they are destroying games.
The main complaint is not the waste of time but the annulling of people’s wins and probable wins. Wasting their time is true, but of secondary importance.
This subject has nothing to do with enforcing etiquette; it is about stopping actions that are unnecessarily destroying games.
It should be noted that the old arguments comparing escapes in live versus correspondence games do not apply to serial timeouts. It should be obvious by now that no one anticipated the bad effects of allowing timeouts in correspondence games while the serial timeout rule is in effect. Indeed, some people still seem not to understand those effects.
The trouble is that this is completely impractical. The main potential source of information, reports, largely do not exist for correspondence games, because timeouts are not reportable for those games.
Not impractical, and I had no trouble filing a report: R111
If by “not reportable” you mean “not something mods will address”, at least one current mod has weighed in about the possibility to ask users not to abuse the system:
You missed my point, as did shinuito a while back. You were able to file that report because you knew about the case due to unusual circumstances. Usually this isn’t possible. The mods will never hear about the vast majority of timeouts that trigger the rule. This is already true of score-cheated games, which get reported only about one-third of the time at best. (This comes from my experience of reporting cheated games, whereby most cheaters had at least two unreported cheated games in their history. Some mods think the report-rate is much worse.)
Since reporting something like score cheating is so rare to begin with, and correspondence timeout is not officially reportable, the mods will never hear about the large majority of cases.