on cheating.
To ensure a healthy competition, we know how important it is to fight all forms of cheating. Every day, lichess moderators and automated algorithms are kicking 50 bad players out. It's really working beautifully, and lichess may very well be the most secure site to play on!
But because some people will always try to find new ways to cheat, we're always working on improving our defenses. Recently we automated sandbagging and boosting detection, with great results. And now Clarkey is building a new version of our computer assistance detector, to alleviate the moderator team work by taking precise decisions very quickly. I can tell you it's a wonderful piece of code, and I feel very sorry for the engine users out there. 🙂
I guess lichess.org can afford to do that because they don't rely on players paying to play there. It never ceases to amaze me that cheaters on RHP are often subscribers, but the result of this is that RHP was cutting its own throat when it used to ban cheating subscribers for engine use.
Yün-men said: “See how vast and wide the world is! Why do you put on your seven-piece robe at the sound of the bell?”
Translation by Robert Aitken from the Wu-Men Kuan
Cheating is always a a touchy subject to breach. I'm especially sensitive to it in regards to online and correspondence chess. The range of engine assistance ranges from playing an engine with minor human input to oops I'm losing let me check this out. A couple points I'd like to throw in:
1. Defining "cheating" is difficult at best. From site to site, and venue, it changes with regards to what is allowed. And, how do individuals define cheating themselves?
2. Cheating is deeply personal. No one likes to lose, but likes the accolades of a high rating. But if you haven't earned it is an empty rating.
3. Cheating has become much easier with Smart phones and access to other programs. I like to play quick chess online but it seems like assisted games are much more prevalent than ever in the past. (Go to playchess.com and play some 15 minute games as a guest.)
4. The use of engines has changed the way chess has been played from the past. Consider: Go to your local chess club, break out your smart phone with any engine and try to find a game.
(It would be an interesting experiment but I don't want to get thumped! At least from the "old timers". Younger players would probably pull out a better phone and engine.)
Final Note: Cheating will always be around. We know its there but its to the point where you just have to say "It is what it is" and play on.
Originally posted by Rookpawn59I started "real" correspondence chess in 1972....(snail mail). It was completely OK to buy & study books on openings....wasn't that a form of cheating ? It was only prohibited to ask advice from another player. After the computers got good enough to beat master level players (like me) , I gave up postal chess. I had been a bonafide master for many years, and still am in ICCF. Personally, I think there are VERY few moves in a chess game that can be suspected as computer generated. So, a talented, able player is to be accused because he/she makes good moves ?? Darn near impossible to prove , unless you can get into someone else's head.
Cheating is always a a touchy subject to breach. I'm especially sensitive to it in regards to online and correspondence chess. The range of engine assistance ranges from playing an engine with minor human input to oops I'm losing let me check this out. A couple points I'd like to throw in:
1. Defining "cheating" is difficult at best. From site to site, a ...[text shortened]... e know its there but its to the point where you just have to say "It is what it is" and play on.
- Joined
- 21 Jun 06
- Moves
- 82236
Originally posted by woadmanThis subject has been up before. Conclusion then is that. Computer movers who always use computers on every move is VERY easy to detect. The computer plays chess differently than humans. I remember from due thread a now banned 2400ish guy as white played Bg5 in the Najdorf main line and than suddenly when book likes was out started to bring home his pieces. The computer did not believe in the attack and suddenly started to defend. All his games was like that. Typical for computers. Very untypical for humans.
I started "real" correspondence chess in 1972....
Have this site banned any players recently? Who was the player latest to be banned for using computers.
Originally posted by Rookpawn59I think lichess is serious about making cheating as difficult as is humanly possible. That is the point. I think it is pure folly to resign oneself to the idea that cheating will always be with us, it may be the case, but there is no need to accept it.
Cheating is always a a touchy subject to breach. I'm especially sensitive to it in regards to online and correspondence chess. The range of engine assistance ranges from playing an engine with minor human input to oops I'm losing let me check this out. A couple points I'd like to throw in:
1. Defining "cheating" is difficult at best. From site to site, a ...[text shortened]... e know its there but its to the point where you just have to say "It is what it is" and play on.
Originally posted by bikingvikingIt is supposed that computers have a chess "style", very tactical, greedy, not positional....
This subject has been up before. Conclusion then is that. Computer movers who always use computers on every move is VERY easy to detect. The computer plays chess differently than humans. I remember from due thread a now banned 2400ish guy as white played Bg5 in the Najdorf main line and than suddenly when book likes was out started to bring home his pi ...[text shortened]... s site banned any players recently? Who was the player latest to be banned for using computers.
-
So algorithms to detect the computer "thinking patterns" are very possible and very accurate.
- Joined
- 21 Jun 06
- Moves
- 82236
Originally posted by SchlecterThere are different algorithms for chess play. Randomization, search tree branching, different settings for everything. I looked some briefly at Komodo, an open source engine. There these things had to be set by the user by adding preferences. A number 1 to 10 I think. There was factory settings. These could be modified easily. I believe in open source. Chess engines are fun beings!
It is supposed that computers have a chess "style", very tactical, greedy, not positional....
-
So algorithms to detect the computer "thinking patterns" are very possible and very accurate.
http://komodochess.com/
https://chessprogramming.wikispaces.com/Komodo
Originally posted by SchlecterWell, there's a couple ways to go here. First, you can rant and rave and send nasty notes the other player. Feels good but does nothing to solve the problem and does little good.
so how we have to play in order to win against a computer chess program?
My approach is a bit more constructive. I try to keep my openings in a narrow tree and do my best to learn the theory, ideas and plans. I use to keep all this on index cards. (Little white cards with lines on them for you young dudes out there.) Now I use two programs that I found very useful. Chess Opening Wizard (COW) and Chess Position Trainer (CPT) I find both very useful in different ways. But basically they both focus my attention on what I should play and I can have options based on my studies.
But lets cut to the case of chess engines.
When I'm going playing through games and have an engine running in the background I can see where engines find moves that are not part of the opening theory I'm used to and differ from prescribed theory. After awhile you will notice patterns that turn up in your opening that computer engines seem to favor. I try to use this to my advantage. Here's what I try to do:
Study those moves or patterns and find a solution that fits into your style of play and your comfortable with. Study full games from your openings and well annotated games which describe plans and goals. Most of the time the engine users don't run the program for long and you may find a solution that isn't in there best interest in the long run.(ie beyond there horizon)
Learn and understand fundamental endgames! (It is difficult to beat a person running a 6-man endgame data base, but in quick games the big engine users sometimes don't have time to check and you can pull off a win.)
But when all is said and done. The strong engine user is going to win, unless you use a stronger engine on a more powerful system. It's hard to learn anything this way. Just watch some engine matches and see it you can understand whats going on. Most of the time its way out of my league.
From personal experience, it's hard to lose to a Cyborg when you've been whipping him pretty good and all of sudden he turns into a Super GM. But maybe he just got insightful! Use it as a training experience and go from there. Best regards, Tim
Originally posted by SchlecterFor the likes of us it is near impossible as one has to play with extreme precision. However some grandmasters are experts at beating computers. Twenty five years ago the thing to do was sacrifice a pawn as it messed up their evaluation functions. These days that won't work so well, but a similar principle is at play. They study the algorithm closely to understand the machines behaviour. Then head for positions that they know represent a weakness for the algorithm. One with a material imbalance of some sort where the algorithm thinks it is winning but where the problem with the position is beyond its move horizon. Cluttered positions where the branching factor is large, so one where there are a lot of available moves which the pruning algorithm can't prune, seem likely to me. But basically they play to the opponents weaknesses.
so how we have to play in order to win against a computer chess program?
Originally posted by moonbusThats interesting. Perhaps they have a reason? Correspondence chess is the art of analysis. I think its virtually impossible to stop correspondence players from consulting books. Its half the fun.
lichess ToS says that consulting databases and opening books is also cheating. However, these are accepted practise in correspondence chess.