Only Chess
07 Jan 08
12 Jan 08
Originally posted by Dragon FireYou don't have a bank of supercomputers. While it is certainly true that engine analysis of the sort I suggested could be done before a game, in order to avoid the scheme I outlined, a player would need to take a sabbatical -- cease all his games, or at least, all of those games for which such analysis might apply. So, not only would you not be able to play for an indeterminate period, you would also not be able to test the ideas gleaned through engine analysis on other players -- and that would certainly cause problems, because in order to avoid a breach of the terms of service, you'd need to stop most or all of your games, do a whole mess of engine analysis, play a few games testing these ideas, stop all games again (returning to the analytical mode) and so forth. And that would slow the process of "extending openings" down to the merest crawl.
My how you ramble on about a supposed "loophole" when I am [b]not and never have suggested using such analysis in your current game but rather have asked the simple question that, to be safe and not infringe the TOS how far removed from a current game is reasonable.
You are proposing something that very clearly is a breach and you attribute ...[text shortened]... ll my past games (if you aren't at it already) and see what sort of match you come up with.[/b]
12 Jan 08
Originally posted by tomtom232All you have to do, my dear fellow, is stop reading the thread, or avoid it to begin with. It's rather like television viewing: if you don't want to see a program called "Is this cheating?" (the very question suggests that the topic is debatable and may perhaps be debated) then don't tune in.
They should have a "Chess Debate" forum that way when an argument like this breaks out it can be moved there where only the people who are interested in arguing have to see the thread.
13 Jan 08
3 edits
Originally posted by Mark AdkinsIs this not this exactly the (legitimate) process of trying to improve?
You don't have a bank of supercomputers. While it is certainly true that engine analysis of the sort I suggested could be done before a game, in order to avoid the scheme I outlined, a player would need to take a sabbatical -- cease all his games, or at least, all of those games for which such analysis might apply. So, not only would you not be able to so forth. And that would slow the process of "extending openings" down to the merest crawl.
1. You lose a game so you analyse it (using an engine if necessary to help you. After all if is was just a silly blunder you made the analysis is hardly necessary), find where you went wrong and discover an improvement;
2. You use that improvement in a subsequent game (that had not even started when you did your analysis);
3. Of course after 2 or 3 moves you find you are out of your analysis and (heaven forbid) you make another mistake and repeat the process;
4. This process takes years so (of course) your improvement is slow with many plateaus as you move up in ability.
This process will apply as much to the end game as to the opening.
Now I think every reasonably strong and / or committed player does this and I also think such analysis is totally within the spirit of the TOS provided it is not done with a current game in mind.
Its the phrase not done with a current game in mind that causes the problem and gives rise to the question we are debating. Specifically how far removed from a current game should we be before it becomes reasonable to do this type of anaylsis? Assume if you can, for a moment, that the intention is not to gave any advantage in a current game but that it is simply to add to your pool of knowledge generally.
In correspondence chess you might have 50 games in progress, many of which may be in the Sicilian or Ruy Lopez and you have just completed and lost a Ruy Lopez. Should you analyse this at a position x moves away from a current game? The simple answer is that this is probably perfectly acceptable if it does not in any way affect your choice of moves in any of your current games.
The dilemma arises as we cannot unlearn something so if we do this analysis with nothing untoward in mind and a current game reaches that identical position we will undoubtedly (whether consciously or unconciously) use that analysis in our current game. Therefore, the question - "How far away from a current position is such analysis reasonable?" (for it to be reasonably safe not to influence existing games).
I accept that the extreme answer is if you want to analyse a game that commenced 1.e4 make sure you have no games in progress that have just commenced 1. e4 or gone down the same line as the game you propose analysing. I think this is somewhat extreme as it prevents quite legitimate analysis not only of my games here but games I might be playing OTB when such OTB games might actually be in adjourned positions where (engine) analysis is not only allowed but is expected.
I think you should also try and accept that this question is posed not to try and bend the rules, change the rules or manipulate "loopholes" in the rules but is a legitimate attempt to ensure that someone does not inadvertantly fall foul of the rules. Anyone who wanted to break the rules would do that and never ask the question (wouldn't they?).
13 Jan 08
Originally posted by Mark AdkinsI don't read it...but its annoying when it keeps popping up on top of the forum. But now I have got myself caught because I will not be able to stop myself from seeing the responses to my posts and then I will have to respond to those responses...etc.
All you have to do, my dear fellow, is stop reading the thread, or avoid it to begin with. It's rather like television viewing: if you don't want to see a program called "Is this cheating?" (the very question suggests that the topic is debatable and may perhaps be debated) then don't tune in.
13 Jan 08
Originally posted by Dragon FireHow could they prove that you were researching certain positions with a current game in mind?
Is this not this exactly the (legitimate) process of trying to improve?
1. You lose a game so you analyse it (using an engine if necessary to help you. After all if is was just a silly blunder you made the analysis is hardly necessary), find where you went wrong and discover an improvement;
2. You use that improvement in a subsequent game (that had n ...[text shortened]... ne who wanted to break the rules would do that and never ask the question (wouldn't they?).
13 Jan 08
1 edit
Originally posted by Mark AdkinsLet me give you another example of what I believe to be a quite reasonable situation.
You don't have a bank of supercomputers. While it is certainly true that engine analysis of the sort I suggested could be done before a game, in order to avoid the scheme I outlined, a player would need to take a sabbatical -- cease all his games, or at least, all of those games for which such analysis might apply. So, not only would you not be able to so forth. And that would slow the process of "extending openings" down to the merest crawl.
You know you are scheduled to play player X in the next round of a tournament and that the games will commence in a few weeks time.
Now if player X is graded say 500 points below or above you you may not need to worry too much as you will win / lose fairly easily and it really is not worth the effort but if he is rated within 100 points (say) and you want to win then a bit of research can help your case a lot.
So you research his past games and find that he struggles more against 1.e4 than against other openings. You then find he prefers the response 1. .. c6. Now you load up your trustly engine and DB and start researching his lines building up the data in the process, exploring deviations and finding ways to improve on the lines his opponents have previously played. Eventually of course you do find improvements and determine what you are going to play. This has of course taken you quite a long time and it is just for a single game so it is not something you do that often but you do do it for key opponents (GMs do it all the time). You save your research and in due course use it in your game.
Now such research is not only legitimate but is quite normal the stronger you get especially when you are playing opponents who are rated closely or above you.
But again you have deliberately analysed the Caro Cann using an engine with games that have not yet even started in mind but the fact remains you could at the same time have some Caro Canns in progress against other players. At no time did you intend using your analysis in one of those games but if you found something out in your research that was useful you couldn't unlearn it. The likelihood of this is possibly remote but however small it exists so is this legitimate preparation now going too far. I don't believe anyone would be able to ignore what they had just learnt in another game if it came up.
So you see I am raising these matters in relation to legitimate research. I gave you an example earlier where a position I analysed had in 100s of games never come up again despite being only 8 moves away from a standard position which gives some indication of how unlikely this may actually be to benefit existing games.
13 Jan 08
Originally posted by tomtom232I don't think this is a question about proof and being caught. It's more a case of finding some common ground on the best way to avoid any risk of legitimate engine analysis helping a game in progress. Of course you could eschew engine use altogether but that would be throwing the baby out with the bathwater as they say.
How could they prove that you were researching certain positions with a current game in mind?
13 Jan 08
I'm going to reverse, slightly, what I posted earlier about analyzing past games not seeming to conform to the TOS. On this I think the TOS is very vaguely and poorly worded. Clearly, and there is no ambiguity on this at all, using an engine to analyze a game in progress is absolutely prohibited. However I don't think it violates the TOS to analyze a past game, whether one has games ongoing now or not, provided one exercises a little common sense. Namely, any past game being analyzed should not be similar to an ongoing game.
Clearly this means that if you are on move 11 of Sicilian Sheveningen then you absolutely should not be doing any other engine analysis of games that featured that line. However if you want to analyze the Yugoslav attack in your completed Dragon game, I don't see the problem. The games are too dissimilar for engine analysis of the past Dragon game to have any bearing whatsoever on the current game, despite their both being Sicilians they are apples and oranges.
The only question in my mind, for one to stay within the spirit (and one interpretation of the vaguely worded) TOS is how much of a margin between an ongoing game and a current game is too close? That's where judgment and scruples come in. Some people have them, some people don't. Those that don't are going to cheat anyway, and they aren't going to be bothered by such nuance because they're the ones who are already using engines to analyze their current games.
Bottom line is that those who want to cheat know they are cheating, however they rationalize it, and a discussion such as this is irrelevant to them because this involves questions of integrity and honesty that they just don't have. For the rest of us, we know there is a line that cannot be crossed. The simplest way to stay on the right side of it is to trust in your own intuition, integrity, and common sense. If you want the points and wins you've chaulked up to remain yours and yours alone, then should the question arise in your mind as to whether a past game you want to analyze is too similar to a present game then just set aside the analysis for a later date.
13 Jan 08
Originally posted by scandiumI would have to agree with this. The TOS says,
On this I think the TOS is very vaguely and poorly worded.
"While a game is in progress you may not refer to chess engines, chess computers or be assisted by a third party."
Well, I've had at least one game in progress here ever since I signed up, which is over six months ago. If I take the TOS literally, I should not use an engine at all for any purpose during that time, even to analyze completed games. The TOS does not say anything like, "You may use engines in ways that have no effect on any of your games in progress," although most people take that to be implied.
13 Jan 08
Originally posted by incandenzaI think that phrase "while a game is in progress you may not refer to chess engines, chess computers or be assisted by a third party" is ambiguous. Certainly one could interpret it to mean that you could not use an engine at all, for any other games no matter how dissimilar. I also think you could just as accurately interpret it to mean that you may not refer to chess engines or 3rd parties for assistance with your current game only, while it is in progress. English wasn't my major in university, but it seems "a game", singular, and "in progress", meaning your current game, is the subject that the remainder of the sentence is predicated on. Strictly speaking, it doesn't have to be worded any differently to mean just that. Unfortunately, the wording of it is can also be interpreted more abstractly to, as you say, mean any game so long as there's a game in progress.
I would have to agree with this. The TOS says,
"While a game is in progress you may not refer to chess engines, chess computers or be assisted by a third party."
Well, I've had at least one game in progress here ever since I signed up, which is over six months ago. If I take the TOS literally, I should not use an engine at all for any purpose duri ...[text shortened]... no effect on any of your games in progress," although most people take that to be implied.
As you point out, you can easily have a game, as you do, in progress for 6 months or even more. To demand that you refrain from analyzing any past games with an engine during that 6 month period is just insane, and I really doubt the TOS is meant to be interpreted that way. In fact, look at the last part of the sentence, the part about a 3rd party. Since that is grouped in the same phrase of the sentence as "chess engines, chess computers..." with an "and" joining them, then if one takes such an interpretation of the TOS then as long as you have any active games ongoing on RHP you could not discuss any completed games with 3rd parties either. Clearly that is not the case as its done openly and routinely on this very forum. I think that should put that interpretation to rest.
13 Jan 08
Originally posted by scandiumSo, vigorous debate has brought out an important nuance. That is well argued.
. . . In fact, look at the last part of the sentence [3b, TOS], the part about a 3rd party. Since that is grouped in the same phrase of the sentence as "chess engines, chess computers..." with an "and" joining them, then if one takes such an interpretation of the TOS then as long as you have any active games ongoing on RHP you could not discuss any complete ...[text shortened]... penly and routinely on this very forum. I think that should put that interpretation to rest.
Now the only question remaining is why Dragon Fire engaged in this elaborate piece of miching mallecho (if such it was). 😀
14 Jan 08
1 edit
Originally posted by Mark AdkinsWhich brings us full circle to the dilemma. How far removed from the "game in progress" is it reasonable to analyse (a) another completed game which at 1 time may have had the exact same position as your current GIP and (b) how reasonable is it to analyse a theoretical position that you GIP will in all probability never go down?
So, vigorous debate has brought out an important nuance. That is well argued.
Now the only question remaining is why Dragon Fire engaged in this elaborate piece of miching mallecho (if such it was). 😀
The simple answer seems to be it is OK as long as you use your common sense, the intention was not to gain an advantage in a current GIP and it does not in any way impact upon your current GIP because if it does it is cheating. The danger will always remain that, however remote, you analyse something innocently and it does come up and it is then impossible to ignore what you have learnt.
14 Jan 08
2 edits
Originally posted by Mark AdkinsI think the answer to this is clear earlier in this thread but I will restate it in simple terms.
So, vigorous debate has brought out an important nuance. That is well argued.
Now the only question remaining is why Dragon Fire engaged in this elaborate piece of miching mallecho (if such it was). 😀
1. I had a GIP that had Q & 4Ps (me) vs Q,B & 3P;
2. My best hope is to try and obtain a draw;
3. 1 way (perhaps the best way) of drawing is to try and get a perpetual check. On an open board with the Qs still on this will always be the best drawing chance;
4. Another way is to try and exchange off into a drawn B & P endgame;
5. Now I know that if I can exchange Qs and give up my 4Ps for his 3Ps he cannot mate with a sole B so I have my draw;
6. But there are drawing chances also with B & P vs P so this was an ending I felt drawn towards researching;
7. The question then arose in my mind, "Would this be a reasonable thing to do?".
I could have just done it of course and never posed this question. In which case no one would be any the wiser even if I had managed to convert into such a drawn ending. In the end I did not do so as I had enough information in my end game books and felt this area was grey enough to be avoided.
Now the important thing is (and I was pretty convinced of this when I posed the question) that this particular game could not be forced and would not transpose into a drawn B & P vs P ending unless my opponent made a whole series of unforced errors but, however small the chance there was still a remote chance and, therefore would the analysis be a reasonable thing to do? I am analysing a position that is wholly theroretical where there is an incredibly small chance that a GIP could reach it. I was interested in analysing that position because a current game attracted me to it but not with an intention to incur an advantage in that GIP.