24 Jul 08
What should we all look out for when trying to find the difference between an "engine" made game tilting move and one made by an opponent possessing skill?
If there is a move made which is anywhere from questionably to obviously "engine" given how do we go through the process of reporting it?
If that is possible, what will happen once we report it?
24 Jul 08
Originally posted by nihilismori think youre going to need a lot of chess skill to start.,
What should we all look out for when trying to find the difference between an "engine" made game tilting move and one made by an opponent possessing skill?
If there is a move made which is anywhere from questionably to obviously "engine" given how do we go through the process of reporting it?
If that is possible, what will happen once we report it?
and a single move or 2 (or 6) doesnt mean theyre cheating, everyone plays out of their level now an then, for example if i think i have a shot in a tournament i may set it up on a real board and work at it for a long time...
that leaves the only way being to (once a game is completed) see the % of 1st choice engine moves. its a VERY controversial subject right now.. think assisted suicide times 10 😉
24 Jul 08
1 edit
Originally posted by coquetteOne move will not suffice. Contrary to popular opinion, there is no such thing as a typical computer move. The basic methodology has been discussed several times already in here so I won't go into details except to say that more games with more non-book moves = more significance.
Every time some player beats me with a clever move I've spotted an engine user!
Edit: Something to cheer no1marauder up. It seems that one game (if it is long enough and has the right sort of moves) may suffice. http://www.cse.buffalo.edu/~regan/chess/fidelity/
24 Jul 08
1 edit
Originally posted by KeplerContrary to popular opinion, there is no such thing as a typical computer move.
One move will not suffice. Contrary to popular opinion, there is no such thing as a typical computer move. The basic methodology has been discussed several times already in here so I won't go into details except to say that more games with more non-book moves = more significance.
Edit: Something to cheer no1marauder up. It seems that one game (if it is lon ...[text shortened]... and has the right sort of moves) may suffice. http://www.cse.buffalo.edu/~regan/chess/fidelity/
There are some kind of moves preferred by engines which strong human players hardly would choose. One such kind of move can only arise suspicion, but if these moves are more then it may be used as at least part of evidence.
24 Jul 08
Originally posted by nihilismorI think it is something that comes with experience. Someone who plays a lot of competitive chess over many years and uses computers for training will have a fairly good idea...a sense that the opponent is "playing like Fritz". I would not be able to tell. There are certain factors about the way computers work that make them very powerful at say calculation up to a horizon..which may be say 10 moves ahead. A strong human player will be able to think strategically and appreciate things (not of course perfectly calculating) that will come about beyond the horizon of the computer.
What should we all look out for when trying to find the difference between an "engine" made game tilting move and one made by an opponent possessing skill?
If there is a move made which is anywhere from questionably to obviously "engine" given how do we go through the process of reporting it?
If that is possible, what will happen once we report it?
Gary Kasparov hinted that IBM had cheated in one of the Deep Blue games against him when he made a "hand of God" statement - a reference to a world cup football cheating incident - after the match.
I simply don't think about it but I expect that at some point in the next few years I'll reach a level of understanding where it will become more obvious to me.
There's an excellent new book called "Streetfighting Chess" and the author has an interesting take on the place of computers in a chess players life....without demeaning what computers do...when I lay my hands on it I'll pull out a quote.
- Joined
- 16 Dec 04
- Moves
- 56692
24 Jul 08
Another way to spot them is to get them to talk in the game, ask them about the position or for advice on where you went wrong. I got GauravV to give me some of his 'pearls of wisdom' in a couple of games, needless to say he wrote some generic bs and showed no real knowledge of how the game had been played.
As for spotting singular moves, there are some obvious candidates like playing on in an obviously drawn position just because the engine states a slight advantage. Other than that I imagine it to be very tricky, especially for the simple minded like myself and others without very good knowledge of chess.
24 Jul 08
Originally posted by nihilismorI think you are a bit paranoid especially at 1200. As a class D player if you played class C or even someone 50 points higher than you, that would look out of your depth but the move played to a 1400 might be weak. This post illustrates the cheating paranoia found on this site.
What should we all look out for when trying to find the difference between an "engine" made game tilting move and one made by an opponent possessing skill?
If there is a move made which is anywhere from questionably to obviously "engine" given how do we go through the process of reporting it?
If that is possible, what will happen once we report it?
24 Jul 08
Originally posted by GatecrasherI feel a little skeptical about this though, because there could be some other reason why the strong player might want to end the game quickly.
One move can suffice.
Like when a 2000+ rated player sacrifices material to force a perpetual check on a 1000 rated player early in the game, because the Fritz eval is 0.0, which is just slightly better than the next best move...
24 Jul 08
Originally posted by GatecrasherDid you notice the guy who asked the question is rated 1200? How would you properly readdress the question with regard to the OP's situation?
One move can suffice.
Like when a 2000+ rated player sacrifices material to force a perpetual check on a 1000 rated player early in the game, because the Fritz eval is 0.0, which is just slightly better than the next best move...
24 Jul 08
Originally posted by GatecrasherI recall a discussion about this about a year ago. Someone had managed to imprison their knight, a move that "no human would make". It was then discovered that Fischer had made exactly the same blunder. I know that Fischer was a controversial figure, but I am convinced he was human.
One move can suffice.
Like when a 2000+ rated player sacrifices material to force a perpetual check on a 1000 rated player early in the game, because the Fritz eval is 0.0, which is just slightly better than the next best move...
24 Jul 08
Originally posted by MahoutNow that you bring that up... I was always wondering. How could the computer have cheated? Used another computer? Used a human move when the whole point was that the computer was supposed to be better than any human? I don't get it.
Gary Kasparov hinted that IBM had cheated in one of the Deep Blue games against him when he made a "hand of God" statement - a reference to a world cup football cheating incident - after the match.