Only Chess
07 Jan 08
09 Jan 08
I understand what Dragon Fire says.
Some day, perhaps tomorrow, my opponent will chose the Sicilian defence. Does that mean that I can't analyze some variation with an engine the 10th move of Sicilian because perhaps that will happen in a game in progess in the future?
Yes, we respect TOS, but between black and white there are gray areas too. I believe that people knows the difference between cheating and not cheating, that we obey the spirit of TOS and what it is meant to be interpreted.
I know that Dragon Fire knows what he is talking about. I think that (almost) every one agrees with him. Common sense rules.
09 Jan 08
Dragon Fire: "As it turns out despite having subsequently played, since doing this analysis, over 60 games where my position at move 4 turned up I have yet to play any where the position at move 12 recurs."
But note how specific in scope this illustration is. Dragon Fire has merely stated that, in a highly restricted instance, a particular reoccurrence has not occurred. This does not address the question of more general probabilities of occurrence; and furthermore, there is nothing preventing anyone from deliberately choosing a highly unusual instance in order to falsely argue a more general point. It is a common enough tactic in debates, and even has a name: the Fallacy of the Biased Sample.
09 Jan 08
2 edits
Originally posted by FabianFnasWhat if the tenth move is two moves away from a game in progress? Where do you draw the line, and why, and how do you enforce such arbitrary choices on the general population of players? The terms of service say nothing about exceptions, precisely because the administrators seem to be the ones demonstrating "common sense" here!
I understand what Dragon Fire says.
Some day, perhaps tomorrow, my opponent will chose the Sicilian defence. Does that mean that I can't analyze some variation with an engine the 10th move of Sicilian because perhaps that will happen in a game in progess in the future?
Yes, we respect TOS, but between black and white there are gray areas too. I beli ...[text shortened]... what he is talking about. I think that (almost) every one agrees with him. Common sense rules.
Or, what if the tenth move follows a predictable book form -- Sicilians have been known to do that -- so that chances are good of getting a particular position with a strong or well-booked opponent whose initial moves have already indicated a particular line or variation? (And when, furthermore, one might be playing numerous games with a particular opening.)
09 Jan 08
1 edit
Originally posted by Mark AdkinsYour argument suggests that DF is deliberately trying to provide false evidence whilst I believe he is genuinely asking a question to seek the opinion of others. If you can accept my proposition (that it is a genunine question in search of clarity) I'd like to know what is your opinion on the subject he raises. I find the accusatory note in your posts is distracting me from absorbing your other comments. Likewise I think DF will benefit by avoiding locking horns with you on these notes. Although I can understand him wishing to defend his position given that he is very generous with his time making positive contributions to these forums....I have one of his posts pinned to my wall.
Dragon Fire: [b]"As it turns out despite having subsequently played, since doing this analysis, over 60 games where my position at move 4 turned up I have yet to play any where the position at move 12 recurs."
But note how specific in scope this illustration is. Dragon Fire has merely stated that, in a highly restricted instance, a particular reo ...[text shortened]... t is a common enough tactic in debates, and even has a name: the Fallacy of the Biased Sample.[/b]
Assume for a moment if you can that DF is a long standing friend and chess companion of yours (as he is to others in real life) and you are enjoying his hospitality at his home and he pours you some of his finest armagnac as you feel the warm glow from the fading embers of the log fire after he just beat you in a game but took time to complement you on the strength of your play and he raises the question: you know I was wondering about...
Back to my thoughts on the subject:
I believe there is a grey area here. The rules of the site cannot enter into this because it is too subtle and it's not possible mitigate against every possibility and create rules that are both reasonable and deal with the complex nature of DF's question.
My argument or response, if you like, to DF's question is that the nature of research DF is talking about would only be undertaken by a genuine player and a genuine player will have a good sense of when any analysis is becoming likely to affect a game in progress and would stop and wait until that game had played out.
As I see it there are two requirements to create the scenario of possible wrong doing that DF is referring to.
1) You are studying and analysing with an engine, either master games or your own completed games.
2) That you have the ability to recall a specific position from one of your games in progress that is somewhere near the end of the book line.
Surely if you are doing these then you are a genuine player and not a cheat.
OK as I write I notice a whole in my logic. You could be like an athlete who takes steroids....working hard AND using the engine in this grey area, researching positions that might occur a few moves ahead in the hope of gaining an edge. And this is cheating. But it is deliberately doing it to gain an advantage that is cheating.
To avoid cheating you have to use your judgment and if ever you feel your legitimate engine use might affect the outcome of a game in progress then stop.
If the rules did give something more specific on this issue they would be hard to communicate to everyone as it requires a level of chess knowledge to explain it. There are players here who don't know what an engine, a book line or a table base is and the rules have to be comprehensible by everyone.
if you do have enough chess knowledge to understand the question then you have enough chess knowledge to avoid cheating...if it feels like cheating then it probably is.
09 Jan 08
Originally posted by Mark Adkins2 moves away is obviously far too close - you are being ridiculous now and know that is something I have never advocated. I initially suggested analysis that was maybe some 12 moves away. Subsequently I have searched my games, database and analysis to find a good practical example and found one that was 8 moves away from the standard Morra Gambit position at move 4 that I had analysed using an engine only to find that such a position had never recurred in all the games I have played here. I have looked through all my other games and found no practical examples where I have analysed a completed game that was closer to that 8 moves. My analysis has identified a number of instances where I did not like the line played, could not find improvements I liked and which I have steadfastly refused to play subsequently. I presume you think using an engine to identify lines I don't want to play and then never playing them is also cheating even though I am avoiding them at move 2 or 3.
What if the tenth move is two moves away from a game in progress? Where do you draw the line, and why, and how do you enforce such arbitrary choices on the general population of players? The terms of service say nothing about exceptions, precisely because the administrators seem to be the ones demonstrating "common sense" here!
Or, what if the tenth ...[text shortened]... tion? (And when, furthermore, one might be playing numerous games with a particular opening.)
I did analyse some obscure opening lines for 1. g4 in considerable detail in preparation for the Grob tournament. In particular I analysed extensively 1. ... Nh6 as a response but this analysis was at a time when I had no games in progress that had commenced 1. g4 ... Nh6. I used that analysis extensively in subsequent play but was out of all my prepared lines in all my games by about move 7.
This has actually got off the subject now. I was anxious in hearing opinions on doing end game research using books and tablebases on theroretical positions that may (but probably won't) arise some considerable distance away from the current position in a game. For example analysing K&P vs K,B&P endings at a time when a game exists with, for example, K,Q,B and 4 pawns vs K,Q and 5 pawns. Such a game is maybe 50 moves away from the type of position one is proposing to analyse. If you find such analysis unacceptable what about when it is K,Q,2R,B and 4 pawns vs K,Q,2R and 5 pawns - still unacceptable well how about when you have K,Q,2R,2B,2N and 6 pawns vs K,Q,2R,B,2N and 7 pawns. You see where I am coming from? A line must be drawn somewhere but where? Purists among you may suggest there should be no such analysis at all but that is not realistic. The simple answer may be that if it could influence a current game it should be avoided but I would suggest that any analysis of any game that commenced 1. e4 could influence some time in the future the games in which you are now commencing 1. e4 and analysis of any end game position could influence your play in any current game.
I would suggest that analysis using engines or tablebases of any position that cannot be forced from the current position in any of your games and that would not be reached by "best play" is probably reasonable preparation and as such should be allowable.
TOS 3 b states "While a game is in progress you may not refer to chess engines, chess computers or be assisted by a third party. Endgame tablebases may not be consulted during play but you may reference books, databases consisting of previously played games between human players, and other pre-existing research materials and that implies that you should not analyse anything with an engine. I would suggest the intention of this paragraph is intended to mean with reference to the current position in the current game(s) or any position that could be reached with reasonable play but not to restrict research in an abstract way on a theroretical position which existing games may never reach.
Maybe the TOS need clarification if they are so rigid they prevent reasonable research and analysis of past games. I am not suggesting they be rewritten, certainly not by me, but interpretation with a healthy dose of common sense which I fear you are not displaying is perhaps required. As individuals we each need to decide where to draw the line. Sail too close to the wind and a deserved ban will be forthcoming (2 moves is definately too close to the wind. I am not and never have advocated that).
09 Jan 08
1 edit
Originally posted by MahoutI beg to differ: it is using an engine with respect to ongoing games that is cheating. The gain is illicit, whether deliberate or inadvertent: and as mentioned before, an experienced player can hardly claim to be ignorant of the possibility of such gain with respect to existing games.
. . . OK as I write I notice a whole in my logic. You could be like an athlete who takes steroids....working hard AND using the engine in this grey area, researching positions that might occur a few moves ahead in the hope of gaining an edge. And this is cheating. But it is deliberately doing it to gain an advantage that is cheating.
I myself have had instances when I would have liked to know about a tactical exchange that had already occurred and could not reoccur in any ongoing game, but I scrupulously avoided any engine use because I knew that the position which had passed, though no longer identical to the current game position or even possible as a future position in that or any other ongoing game, still possessed certain features relevant to an ongoing game (being derived from an earlier position in the game); and in using an engine I might get ideas about the importance of certain spaces, files, ranks, diagonals, holes, or other positional or tactical features, that might still apply, but which I had not obtained on my own; and because I FORMALLY AGREED to adhere to the terms of service in regard to the prohibition on engine use during ongoing games.
So you see, it is not even a question as to whether a position analyzed by an engine will occur *exactly* in a game, but merely whether in using an engine to analyze lines that COULD occur in an ongoing game, one is getting ideas of play from the engine. If one uses an engine, and the analysis is carried out during an ongoing game to which the analysis COULD become applicable, and one gets ideas about how to play the game -- ideas which one did not have prior to using the engine and obtained through engine use -- then that is most assuredly cheating. The fact that a player is otherwise strong or weak is utterly irrelevant also.
As for imbibing brandy with Dragon Fire, that is extremely unlikely. I perceive (rightly or wrongly) a certain machiavellian jesuitry in his strain of argumentation, and I would not wish to taste the cup of the Borgias.
10 Jan 08
Originally posted by Mark AdkinsEngine use of a past position in an existing game is not permissible.
I beg to differ: it is using an engine with respect to ongoing games that is cheating. The gain is illicit, whether deliberate or inadvertent: and as mentioned before, an experienced player can hardly claim to be ignorant of the possibility of such gain with respect to existing games.
I myself have had instances when I would have liked to know abou ...[text shortened]... n jesuitry in his strain of argumentation, and I would not wish to taste the cup of the Borgias.
You are welcome to come visit, have a Southern Comfort or Bud (I don't drink brandy) and play a few games of chess any time you are in sunny England.
10 Jan 08
2 edits
Using an engine or tablebase to research an ongoing game is cheating. It doesn't matter what distance (+ or -) exists between the position being analyzed and the game's current position. The legitimate time for engines and tablebases is after the game is over, so you can learn from your (and your opponent's) mistakes.
Any engine generated analysis used in a game, should have been known to you prior to the commencement of the game. If not, you are being assisted by an engine, no matter how directly or indirectly you came by that knowledge.
Once out of the opening book, the chances that a human being will remember the engine analysis of a specific position that may occur among the almost infinite number of possible positions, is astronomical. Prior engine research is an entirely pointless exercise for the average player. Again, the benefit of engines is in hindsight instruction.
Using an opening database that contains games played by engines is also not permitted. And, of course, following a long line in a unique game previously played by an engine is just asking for trouble.
If you are knowingly using an engine or tablebase's influence in a game just ask yourself whether you possessed this knowledge prior to the commencement of the game. If the answer is no, you are being unfairly assisted, and in violation of 3(b).
10 Jan 08
Originally posted by Mark AdkinsIt is not all black and white, you know, there is also a gray area. TOS can't give us specific rules for all cases. That's why God gave us common sense.
What if the tenth move is two moves away from a game in progress? Where do you draw the line, and why, and how do you enforce such arbitrary choices on the general population of players?
10 Jan 08
Originally posted by Mark AdkinsI perceive (rightly or wrongly) a certain machiavellian jesuitry in his strain of argumentation, and I would not wish to taste the cup of the Borgias.
I beg to differ: it is using an engine with respect to ongoing games that is cheating. The gain is illicit, whether deliberate or inadvertent: and as mentioned before, an experienced player can hardly claim to be ignorant of the possibility of such gain with respect to existing games.
I myself have had instances when I would have liked to know abou ...[text shortened]... n jesuitry in his strain of argumentation, and I would not wish to taste the cup of the Borgias.
Whilst you're text carries a degree of verisimilitude I believe you're over imagining the level of iniquity in our suburbs and perhaps even projecting.
10 Jan 08
2 edits
Instead of creating a new thread, when it so well fit in this one, I ask my question here:
What about a member A gives member B a hint of how B should proceed in a game? B has not asked for a specific help about his game, rather reported a bug whereas A says that B is in check and actually can make a particular move.
Has B committed a violation of TOS 3b? I say no because he didn't ask for a good move. He didn't even say what game it was.
Has A committed a violation of TOS 3b? I say no because A is not the one that was assisted, by the rule of TOS 3b.
TOS 3b is as follows: "While a game is in progress you may not refer to chess engines, chess computers or be assisted by a third party. Endgame tablebases may not be consulted during play but you may reference books, databases consisting of previously played games between human players, and other pre-existing research materials."
I would want to include in the 3b text that it is not permitted to assist others either.
I.e. It is not permitted to get assistance, nor give assistance.
Comments, anyone?
10 Jan 08
2 edits
Originally posted by FabianFnasA may not discuss an in-progress game if his/her comments would assist or influence either player.
Instead of creating a new thread, when it so well fit in this one, I ask my question here:
What about a member A gives member B a hint of how B should proceed in a game? B has not asked for a specific help about his game, rather reported a bug whereas A says that B is in check and actually can make a particular move.
Has B committed a violation of T I.e. [b]It is not permitted to get assistance, nor give assistance.
Comments, anyone?[/b]
B should ignore A's comments. Rather like a jury should ignore inadmissable evidence. That doesn't mean that B cannot make the move that A suggests, only that he should make his move solely on the basis of his own efforts.
B should send feedback to the site admins or game moderators.
Giving advice (unsolicited or otherwise) in an in-progress game can be prosecuted beyond the confines of 3(b).
10 Jan 08
1 edit
Originally posted by GatecrasherYour post is absolutely correct and I agree with it 100% but I fear it is over simplistic.
Using an engine or tablebase to research an ongoing game is cheating. It doesn't matter what distance (+ or -) exists between the position being analyzed and the game's current position. The legitimate time for engines and tablebases is after the game is over, so you can learn from your (and your opponent's) mistakes.
Any engine generated analysis us of the game. If the answer is no, you are being unfairly assisted, and in violation of 3(b).
Maybe I am being too dogmatic here because the reality of the situation is that my analysis of past games has rarely benefited me in practice except in so far as it has helped me decide that I do not like a particular book line and I should avoid it in any subsequent game. I gave the previous example because it was the best I could find from all my games here but the reality is that in maybe 6 games I have reached a position where my book stated "white (black) stands better" and proceeded to lose because I couldn't capitalise on it. In all those cases I have subsequently used an engine (and discussed the games with other chess players) to help determine where and why I went wrong and made notes of this analysis in my (paper) books but, so far, I have had no cause to refer to it as none of my subsequent games have gone down those lines.
The only times my research has been beneficial has been in the very early stages in researching a thematic opening and in investigating common early deviations but even then not for very many moves. All this research has been done prior to the commencement of any current game.
But, however unlikely, the situation does remain that I could be innocently analysing a just completed game and that as a result of that analysis I discover my mistake and find improvements. Some moves later an existing game that I did not have in mind at the time of my analysis goes down that line so I play my improvement. Have I cheated? Certainly not deliberately as the analysis was never done with a current game in mind.
... so now we move to the end game ...
End game research beyond browsing my books and ensuring I know and understand the most common endings I have rarely done. I do not even have a tablebase so its use by me at present is purely hypothetical. You are however a stronger player than I, so I assume you will understand the need to do research if you want to improve your game; not only in the openings but also by analysing endings that you feel may be beneficial to you. There is no point in doing this research if you do not hope to use it one day and although in practice much of it will never be used it still hopefully helps you improve.
Lets us assume for the moment that the point of this study and analysis is to add to your pool of knowledge generally but has come about because a current game has provoked your interest but that the position being studied is entirely theoretical and your current game cannot reasonably be expected to go down that line. Such analysis is surely not only reasonable but normal for players such as you and is, or at least should be, entirely acceptable. If such analysis is intended to directly assist you in a line that you can reasonably be expected to reach in a current game with the intention of helping you decide what to do later in that game then such analysis (with engines or tablebases) is clearly unacceptable. The problem arises when the intention is not to directly assist in a current game but to expand your pool of knowledge in a area down which a current game could possibly go by using books (a legitimate tool) and a tablebase (not legitimate if used to assist in a game) to study a certain type of ending. In the example I gave I wanted to study K & P vs K, B and P at a stage where one of my current games still had the Qs and loads of pawns on the board but where I was a B for P down so it could, in theory, have some possible relevance and I was concerned enough to wonder if this would be a reasonable thing to do. As it turns out I did not do any research other than in Averbachs end game books and the game didn't get even remotely close to such an ending so the matter remains a totally theoretical one. Studying such endings is in my opinion totally reasonable but using a tablebase to play such endings is a breach of the TOS.
Using an endgame database (as opposed to tablebase) would also be reasonable study (for those in doubt can I state that many people refer to end game tablebases as a database and the fundamental difference is not what they are called but what they are. A tablebase is a list of all possible moves and the result that best play will give. A database is a collection of past played endgames. There is no question that the latter is reasonable research.
No one has actually expressed a view on my example of outside assistance where I go to my local club, ask how to play and then I practice K,B & N endings with local players until I get it move perfect when I have such an ending in progress here. Such a situation could potentially be quite common but you are adding to your theory and general knowledge rather than seeking direct assistance in your game.
Applying knowledge gained from that study to a game is surely reasonable if you use that knowledge rather than computer (tablebase) generated moves and I suspect that is actually the fundamental issue. You can surely use a tablebase to assist in your study as long as you do not use it to make your moves and you can apply any knowledge gained in that process at a later stage in a current game. It is knowledge that you are using not moves decided by a computer.
You did not have that knowledge at the start of the game, certainly but when you are playing dozens of games over months and sometimes years there will be a lot of knowledge you did not have at the start that subsequent research and analysis adds to.
I think I personally know where the line is and won't deliberately cross it (hopefully I won't inadvertently cross it either) but I felt this discussion was worth having and the points worthy of airing.
Many weaker players will perhaps not understand what we are talking about nor why. In their case the situation is clear cut as they are rarely doing this research to improve in this way. In my case however my objective is to improve OTB and this all becomes more relevant. At the moment my play here is 150 points above my OTB grade and I am hoping my OTB grade will catch up over the next few years and I will reverse my long decline. That improvement is a direct result in devoting more time to the opening and more detailed analysis of tactics. The stronger one gets the more legitimate research needs to be done.
10 Jan 08
1 edit
Originally posted by Dragon FireI recently finished a game Rook Each, Kings, Me with a pawn.
We already know that researching opening books and databases is not cheating.
Some time ago we had a discussion about using an engine to analayse not only your past games but positions you may reach in existing games if they followed a book line to a point where the book ended (maybe a dozen unforced moves from any game you are playing at present with ...[text shortened]... uld not now load up the tablebase.
Is this legitimate research or does it infringe the rules?
I did not research this ending because it IS cheating. If I wanted to know about it, I should have done it before the game, or after.
I'll research it now to see how I did.
Do you feel this research will help you in the game in progress? If yes, it is outside help. If no, why bother doing it now.
P-
Before you ask about "Book and Data" as help, we are allowed to do this. If you are searching Table or Engine that is completely different as the rules for correspondence chess go.
10 Jan 08
Originally posted by PhlabibitI agree 100% with what you are saying in this instance. I am talking about research somewhat further removed than that.
I recently finished a game Rook Each, Kings, Me with a pawn.
I did not research this ending because it IS cheating. If I wanted to know about it, I should have done it before the game, or after.
I'll research it now to see how I did.
Do you feel this research will help you in the game in progress? If yes, it is outside help. If no, why bother d ...[text shortened]... rching Table or Engine that is completely different as the rules for correspondence chess go.
Although I would just mention that I felt uncomfortable enough about the situation that prompted me to start this thread NOT to do any research, other than in end game books in this instance anyway. In fact I eventually found all I needed to in end game books.
As I said earlier with dozens of games on the go it is possible that any innocent research could directly or indirectly help you with a game in progress.
I think that most of us probably do understand the difference and maybe I am having a measure of difficulty explaining the sort of situation I am envisaging but I agree with you in principle that if the "research will help (with) ... the game in progress" it is contary to TOS.