Why is aborting considered to be a bad practice

@Brian-E said in #3:
> Indeed (#2)!
>
> However, aborting a game should be a really unusual measure to take because it is bad sportspersonship (sorry, I don't know a non-sexist word for this concept), and some players may experience lots of players aborting games against them even if you don't - people from certain "unpopular" countries for example. Therefore, I think a warning every time is completely appropriate.

sportspersonship is not a word no won will cancel you for saying sportsMANship.

@nadjarostowa said in #6:
> I think most of these are actually no valid reasons at all!
>
I think that some of them are, but for each reason you may or may not agree and that's perfectly fine. I think at least there's a grey area and it's not always easy to determine if there was a good reason for an abort. The first two are the ones that I encounter the most, the other ones are rare.
>
> If you ask for pairing you have to play whoever shows up.
>
I would agree, unless high lag affects the game substantially, which I think is the case with 1 0. (As far as I understand there is time compensation for lag, but a side effect is that it allows making long series of premoves with otherwise wouldn't be possible. This is something that is very hard to do perfectly right, I understand.). For longer time controls, this is not an issue, so I specifically mentioned 1 0. I have played a lot of 1 0 recently and it sometimes happens.
>
> He might have left, he not have arrived, he might just reload the page.
>
This case almost always happens when I complete a game, the opponent asks for a rematch, I accept within 2 seconds and then nothing happens (opponent's green dot also became grey). Then I have to wait for 30 seconds and most of the time, the opponent has indeed left. This case doesn't happen very often when seeking a new opponent of accepting from the lobby.
>
> Rating deviation is taken into account when adjusting the rating, so it should be a non-issue. The higher risk of cheater I'd say is marginal.
>
It's not about the rating, but the risk of cheaters, especially in rapid or classical time controls, when it is easier to cheat. Most of the times when I encountered cheaters, they were from relatively new accounts. So if it were an option, I would enable a filter of "having played at least X games".
>
> Wrong time control or variant is valid if you misclicked. :-)
>
This actually happened when I decided to accept a challenge from the lobby, but someone else accepted before me. Then the list of challenges in the lobby updates and another challenge is there where I wanted to click (may be a totally different time control or variant). This one is very rare though.
>
> Pretty unlikely that this happens exactly at the moment between starting the game and the first move. And how would you notice and still be able to abort the game?
>
Agreed that this one might be very rare.
>
> This usually isn't possible (it might be possible in the mobile client).
>
I once had a situation that my opponent offered a rematch, but I had to leave. When I returned half an hour later, I noticed that the rematch was still there, so I accepted. I'm not sure if it was a bug, but the opponent had left in any case. (Maybe I can try to reproduce it.)

> But I think the current automated system takes care of the issue reasonably well. It seems impossible to get it perfectly right. And aborting games should be highly discouraged.

Agreed. Aborting should be discouraged, so establishing warnings to habitual aborters and restricting them is totally fine with me.

I think my issue is more with the warning after every single abort, even one. When the system is threatening with a temporary ban after only one abort, then I wouldn't feel very welcome on this site.

So, most of the time I wait the 30 seconds, but it can be a bit annoying. Maybe an option is to allow an abort after 15 seconds (or something) without punishment?

@monty82 said in #12:
> So if it were an option, I would enable a filter of "having played at least X games".
You may have forgotten but you were also new once... How are those new users supposed to play those "X" games when noone wants to play them? They have to play their games against Mr. Someone Else, right?

@mkubecek said in #13:
> You may have forgotten but you were also new once... How are those new users supposed to play those "X" games when noone wants to play them? They have to play their games against Mr. Someone Else, right?
They would still be able to play other players that are new, until they reach the X games. This site has tens of thousands of players online. It's not that there's nobody left to play against in that situation.

I somewhat understand that with new players. But from the "server side view", letting them play against established users is the best way to get them an established rating, too.

Or, as you say, "When I am new and I am sidelined by the existing user base, I wouldn't feel very welcome on this site." ;-)

(On the longer time controls, if many users would filter newcomers out, that might lead to quite some waiting time. For bullet it's probably not an issue.)

As for the high latency users, I have never received that as a problem in bullet. In fact, I always thought it helped me. Maybe because I am too slow anyway, and their lag gives me time to get my premove in.

As with your 30-minute rematch - my guess is your opponent offered the rematch, waited some time, forgot about it but left the game (and rematch challenge) open, and then just left the computer.

@monty82 said in #14:
> They would still be able to play other players that are new, until they reach the X games.
How is it going to help them get an established rating? Games against other players whose rating is unreliable it not going to make theirs more reliable - or at least not enough.

When playing rapid or classical, I usually check the ratings of my opponent. When they have provisionary ratings, but established other ratings (for example in blitz or bullet) then it's usually fine (low probability of cheater).

However, if they only have played 20 games in total with a 100% win rate and the account has been created on the same day, then the probability of cheating is higher.

Especially when playing longer games, I don't want to waste time playing against a possible cheater. I would prefer an option to filter "high-risk players" out, but the second best option would be an "abort without punishment".

Can't help noticing that this still does not answer the question how is a new player going to get out of this (supposedly) "high risk" category.

I suspect that this induced cheating paranoia is even more harmful to the overall experience than cheating itself. And the practice of checking opponent's profile for "warning signs" (some of which are completely absurd, like "too big" difference in rating between different time controls) is something I personally find even worse than aborting games for stupid reasons. So don't expect me to accept it as a reason to be more benevolent towards serial aborters.

@mkubecek said in #18:
> Can't help noticing that this still does not answer the question how is a new player going to get out of this (supposedly) "high risk" category.
This question has already been answered. There are enough players left to play on this side. And if people think that it shouldn't be an option then they probably won't use it.

At my playing level (mostly between 2400 and 2500) I have encountered multiple cheaters on lichess, so unfortunately cheating is an issue. Maybe it is different at 1600-1800 levels.

Although the discussion is now focused on this one facet, maybe I have aborted two or three games for this reason in a year time. That is a totally different number than habitual aborters would have.

@monty82 said in #19:
> This question has already been answered. There are enough players left to play on this side. And if people think that it shouldn't be an option then they probably won't use it.
Another way to look at it: right now, aborting games is considered a bad practice and a warning is issued which discourages many users and limits those who are not discouraged enough. If lichess says it's OK and even provides a config option, a lot more people will be tempted to behave like this.

> At my playing level (mostly between 2400 and 2500) I have encountered multiple cheaters on lichess, so unfortunately cheating is an issue. Maybe it is different at 1600-1800 levels.
Can't say. Neither if those you encountered were really cheaters (and were cheating in a game against you - no, it's not nearly as easy to recognize as some people believe). All I can say is that I'm aware of two of my opponents who were marked for "ToS violation" (which does not necessarily mean cheating) later and I have absolutely no idea if they cheated in the games I played against them. In both cases I made enough mistakes to lose.

> Although the discussion is now focused on this one facet, maybe I have aborted two or three games for this reason in a year time. That is a totally different number than habitual aborters would have.
And also a frequency at which, an occasional warning is hardly an issue if your really believe you are right. However, at least half of the reasons you presented were about as questionable as this one, perhaps even more.

Maybe I'm biased but I'm rather seeing this from the other side. It already happened to me few times that people just aborted a game, some of them even after playing their first move, before I managed to reply (i.e. unlikely to be an accidentally started game with wrong time control). And after my latest rapid game (lichess.org/rsSvYijb5Rhl/), my opponent said "I'm never playing anyone with a ? Besides there rating again." which is funny, considering how bad I played.