Just chiming in, in reply of two points, regarding this offence repeating, and false positives, so whichever post numbers these two were:
@ChessWhiz2031 said:
>if this repeats in more than one game
and
@AsDaGo said:
> Unless false positives happen frequently to the same person, I don't really see this as a problem.
This is exactly what happens! False positives may tend to happen over and over.
How does this happen? The user is not made aware, as rwsf put it informatively and in quite some depth, of the why or the how. The "offending user" just gets a standardised message. It would be beyond most to realise for example that they thought for 59 seconds while the secret time limit is 30 seconds for a particular time setting.
Not getting the specific information has two effects, IMHO. One, it is easier to feel unjustly accused, that the message comes out of left field. And second, the user may unwittingly do it again, because no useful information was given by the warning, just harsh words.
It is nice to see that issue 19320 was opened on Github, although it is likely from that discussion a better decision "behind the scenes" might be implemented for that particular case rather than solving the "lack of information leading users to not change behaviour" part.
@ronin3b said:
> This is exactly what happens! False positives may tend to happen over and over.
Do you have evidence for that, or are you just guessing?
> How does this happen? The user is not made aware, as rwsf put it informatively and in quite some depth, of the why or the how. The "offending user" just gets a standardised message. It would be beyond most to realise for example that they thought for 59 seconds while the secret time limit is 30 seconds for a particular time setting.
It's probably best if the information is not given, since that would make it easier for malicious users to cheat the system. Also, I'm pretty sure it's more complicated than just having a certain time limit for each time control.
> Not getting the specific information has two effects, IMHO. One, it is easier to feel unjustly accused, that the message comes out of left field. And second, the user may unwittingly do it again, because no useful information was given by the warning, just harsh words.
It very rarely happens to me, and I tend to be a slow player. So yes, it is possible that the user does it again by accident, but I think that will not happen enough for the false positives to be a problem. If you have evidence to the contrary, I'd certainly be interested to see it.
Also, having the information and trying to avoid it happening again is not ideal (even for an honest user), because it would prevent you from playing naturally.
Title: "threatening messages"? It is a friendly reminder! Perhaps the algorithm saw you using one minute to analyse the critical position in a 3+2 game, as stalling?
@AsDaGo said:
> I'm pretty sure it's more complicated than just having a certain time limit for each time control.
Why did I bother linking the source code if you are just going to ignore it and say "I'm pretty sure"...
@AsDaGo said:
> > This is exactly what happens! False positives may tend to happen over and over.
>
> Do you have evidence for that, or are you just guessing?
It happened many times in a row to my games in the past. Not recently, but still.
Already a casual player I just switched to other platforms.
I don't like your accusatory, ad-personam tone, by the way. Allow me to reply.
If you are going to have that tone, I'd recommend you check your sources first, as also Rw pointed out, as it seems the one "just guessing" is you.
This nitpicking of people's forum comments isn't going to do anything. I think the only way forward is to find some simple way for Lichess to show the warning but only when appropriate.
I created that GitHub issue to try to put forward OP's suggestion as positively as possible although I am deeply skeptical that a solution exists. But I've taken up lost causes before and on occasion found solutions.
Surely such an important issue cannot be resolved by us mere humans. Let's see how AIs would write that message:
“ATTENTION! We’re seeing some very bad behavior, folks - players who are totally losing and still wasting the other person’s time. Very unfair. Very sad.
If you do this, we’re going to take a BIG, BEAUTIFUL tariff from your future wins - that’s right, a chunk of your rating, straight off the top. We’ll keep it. Believe me.
Don’t stall. Don’t delay. Make your move or resign like a winner. The best players don’t waste time - they play fast, they play strong, and they respect their opponents.
Waste time, and you pay. It’s that simple."
Also, they suggest we write it in a golden font.
When I raised this issue on the lichess discord a least half a year ago, it was met with what I'd charitably characterize as contemptuous silence: https://discord.com/channels/280713822073913354/1403854774990016602
@ZugAddict said:
> When I raised this issue on the lichess discord a least half a year ago, it was met with what I'd charitably characterize as contemptuous silence: https://discord.com/channels/280713822073913354/1403854774990016602
its not up to us it is up to the other accounts, but hopefully @lichess will find a way to stop it.
Honestly, I think stuff like that should be calculated as a probability. I am sure more mathematically inclined folk here can suggest a formula that would define the probability of each level of bad behavior for a player based on past data. Instead of lazily presenting the message to people when playing - and also confusingly to their opponent, in the chat - they could get a notification letting them know their scores have reaches a threshold that demands attention.
After all, we are not trying to single out people who dared do the deed and burn them at the stake, we are trying to improve the whole Lichess community as a system. That requires probabilistic math to solve, determining the main contributors to bad sentiment and fixing each of them in the order of their negative impact.
This topic has been archived and can no longer be replied to.