The biggest problem for me on this site is standard non-paying, lower rated members, and some paying, falsely rating-elevated payers, using Stockfish, or some derivative thereof, and skewing ratings and diminishing everything important down to a base deception.
As I say in my profile I know within 15 moves who is up to it and I block them. Two recently, one ongoing blatantly using AI and this has been and is a persistent problem.
I have complained to the admin on numerous occasions that the accuracy is absurdly high with these people and named names. I have got to the stage where I have told them I shall not be renewing my paid status in August since they seem to do nothing about it or even acknowledge the message.
I cannot beat AI plain and simple nor should I need to be able to do so.
As long as this is allowed to continue and not even be acknowledged and looked at then it makes a farce of this, and other, chess sites.
- Joined
- 20 May 16
- Moves
- 41676
@MLSChess
As I say in my profile I know within 15 moves who is up to it and I block them. Two recently, one ongoing blatantly using AI and this has been and is a persistent problem.
How do you know, I’ve wondered about this many times with very low players who all of a sudden play extremely well and stats don’t match.
@mike69
Hello Mike69
I have the benefit of having a son who writes software for a living and he has devised a package based on Computer Aggregated Precision Score (or CAPS) that evaluates the strength of play for any chess player, or the quality of moves of any chess game.
Unlike standard ratings, the CAPS system applies a score from 0% to 100%. The higher the percentage reflecting the accuracy as would be 'played' by the top AI systems.
So a sample of say 3 games analysed would show up a pattern and clearly rating 'straight liners' up in short periods are always worth looking at and usually found to have ridiculously accurate scores of 90+%. So in excess of 9/10 moves are exactly what Stockfish or the likes would do.
I'm lucky on self analysis if I achieve 70% and I'm sure most of us are in the same boat.
There are some who think they are being clever by battling much higher rating players and asking for a draw early on or repeatedly. These play for a rating hike with much higher rated players and don't actually want to win but always play on to victory over those lower or level rated players and within 200 higher. It's a very transparent ruse at camouflaging the deception.
@mike69 saidHow would you know within 15 moves in a running game, unless you were using AI yourself?
@MLSChess
As I say in my profile I know within 15 moves who is up to it and I block them. Two recently, one ongoing blatantly using AI and this has been and is a persistent problem.
How do you know, I’ve wondered about this many times with very low players who all of a sudden play extremely well and stats don’t match.
@mlschess said15 moves is still within book lines for many common openings. You must assume that players here are using online databases (chessgames.com, for example) or MCO or similar published references for the first 15 moves before you accuse anyone of engine cheating.
The biggest problem for me on this site is standard non-paying, lower rated members, and some paying, falsely rating-elevated payers, using Stockfish, or some derivative thereof, and skewing ratings and diminishing everything important down to a base deception.
As I say in my profile I know within 15 moves who is up to it and I block them. Two recently, one ongoing blatan ...[text shortened]... ue and not even be acknowledged and looked at then it makes a farce of this, and other, chess sites.
@divegeester As above @divegeester
Unacknowledged and no follow ups.
I stand back from the conversation. I am very interested to see the replies although I'm bound to say, so far, one or two seem rather more interested in fatuous comments than addressing how this can be stopped.
The software my son has devised is in the process of a patent application so I cannot really divulge too much as to how it works but it is going to be a great asset to those of us who want people to use their skill and minds and not AI
@mlschess saidIn your OP, you said that two recent games and one on-going game looked suspicious.
@divegeester As above @divegeester
Unacknowledged and no follow ups.
I stand back from the conversation. I am very interested to see the replies although I'm bound to say, so far, one or two seem rather more interested in fatuous comments than addressing how this can be stopped.
The software my son has devised is in the process of a patent application so I cannot re ...[text shortened]... is going to be a great asset to those of us who want people to use their skill and minds and not AI
Three games (against three different players?) is hardly a sound basis for making accusations. To catch an engine cheater, you would need to analyse more than three games from the same player, and later than 15 moves into the game.
Bear in mind that pretty much every player these over 2200 trains against engines. So it should not be surprising that their games tend to look machine-like; that is the only way to beat a machine, after all.
Software which can compare game moves with several engines would be a welcome tool; I wish your son success in his venture. I understand your frustration, but until such time as reliable engine-detection software is up and running and tested, decorum is advised when making accusations on a public forum.
