 bnolan
Chess rating: 2009
LCF 1865
 
United States
Give chess goodie
Chess goodies: 3 |  Sat Mar 24 2012 1:04PM | MsgID: 15300628
How I use the openings database: I decide on my move, then I look up the position in the database and see if anyone has tried that move before, and what were the results. If that move has been played with reasonable results, I make the move. If it is not in the database, or my side always loses, I do not make the move. I try instead to figure out why it is not played, and try to come up with a better move.
How I might use an endgame tablebase: I might have a position with eight total pieces on the board. I am considering exchanging down to a six piece endgame. I could check the tablebase to see if the six piece endgame is a win or a draw. If it is a win, I make the move. If it is a draw, I look for a better move.
In neither case do I let the computer or anybody else suggest a move for me. In both cases, I could legally do the same thing with a book.
|
Taotaomonas
Chess rating: 1724
Australia
Give chess goodie
| Thu Mar 8 2012 1:30PM | edited: 1:35:32 | MsgID: 15254206
I have come to realise since joining the site I will never have the desire or the the required patience to make a good CC player. I have however come to appreciate and have respect for those who do because it is a completely different game to OTB that demands a whole other set of skills.
The question really for me has never really been what and what not is considered fair research. But more perhaps how and where you differentiate between a very good chess player and a very good researcher.
I assume that many top OTB players naturally graduate toward CC over time. They will likely already have good knowledge and grasp of all aspects of the game, specialise in certain openings and gambits, recognise drawn or lost positions mid game, etc, etc. I envisage these players studying and playing through completed games, researching where they went wrong or didn't, possibly even using engines and other interactive software to analyse options for future games.
These type of players primarily like me share a love of chess itself and are simply looking to constantly improve and enjoy the game.
There are obviously other players who run chess engines during the game, research options again through other interactive software. The site offers a certain amount of assistance itself through data base, analyse tools, etc. I can see little fun or point myself in using all of the above combined. But does this mean it is wrong!? Or is it just someone getting as much pleasure sourcing the best move as others do playing the game!
A table base is something I had never come across. Following the link in the first post I found to the untrained eye a facility that allows someone to basically input an end game position of 6 pieces or less into an engine that will theoretically run through every possible outcome.
I cannot see therefore how that is playing chess! if you were to consult such a tool during a game. Is it fair to research a previous game through as with any other system? I cant see why not!
I would add that for someone like me who cant necessarily always see a lost or drawn game. It would be a very useful and simple tool to learn about close end game positions. Using it during a game though would seem pointless. I might as well just ask someone else to make my moves for me!
|
SteveCS
Chess rating: 2380
LCF 127 Fide approx. 1885
United Kingdom
Give chess goodie
Chess goodies: 9 | Thu Mar 8 2012 9:59AM | edited: 10:00:49 | MsgID: 15253804
[Quote from: "Theseus"]| [Quote from: "bnolanThe point of my original question, however, is not should we use tablebases, but rather can we? Is there an official stance? Or even a broad consensus? |
I think this has been raised before (I've been on thsi site a while) and that the consensus was, "No". At the time, I knew even less about tablebases than I do now - which is not very much - so I contributed little. And, in truth, I doubt I can contribute much now.
I think there's a world of difference between researching strategies and lines of play, which is not only encouraged in CC but is probably essential to good play, and using a device to generate moves (e.g. computer generated moves) - tablebases undoubtedly fall into the latter category.
I'm fascinated by endgame play. I enjoy Karsten Muller's excellent series of articles on ChessBase. And, of course, I seek improvement; but through study and practice, through an understanding of the concepts and principles involved. I would not wish to follow a set of moves like an automaton.
Can we?
Well, I suppose it can't be stopped or detected.
Should we?
Most certainly not; I have no wish to use or play against such a resource. |
One might suggest that in many ways we already do (follow a set of moves like an automaton) - everyone knows the trivial won endings such as KRvK, KQvK, KBBvK. We also already know how to recognise if we can force a draw against a single pawn. These are deterministic procedures built on combinations of well-known endgame concepts (opposition, triangulation, zugzwang, co-ordinate squares, the zwischenzug, etc) - get it wrong and the outcome will change.
Actually, I agree with everything you say. I am also fascinated by the endgame, and I find my desire to understand these often beautiful ideas trancends any desire to just 'know' the outcome. For example, most people 'know' that KBBvK is won but not all of these will know why it is won such that they can force the win from any KBBvK position* - tablebases do not provide this knowledge and so contribute little to human understanding of the game, which is a fundamental premise of CC.
Given that the ability to passively read the tablebase is now available I can see no reason why they should not be used as reference material, but if doing so replaces the need or desire for understanding which then stops people researching and exploring for themselves then this will be a sad day.
/S
* except of course where the opponent has the move and can capture a bishop in which case you need to know why this is then drawn - just knowing that 'it is' is not enough.
|
Theseus
Chess rating: 2237
United Kingdom
Give chess goodie
Chess goodies: 3 | Thu Mar 8 2012 5:52AM | edited: 5:56:07 | MsgID: 15253452
Originally posted by: "bnolanThe point of my original question, however, is not should we use tablebases, but rather can we? Is there an official stance? Or even a broad consensus? |
[/quote
I think this has been raised before (I've been on thsi site a while) and that the consensus was, "No". At the time, I knew even less about tablebases than I do now - which is not very much - so I contributed little. And, in truth, I doubt I can contribute much now.
I think there's a world of difference between researching strategies and lines of play, which is not only encouraged in CC but is probably essential to good play, and using a device to generate moves (e.g. computer generated moves) - tablebases undoubtedly fall into the latter category.
I'm fascinated by endgame play. I enjoy Karsten Muller's excellent series of articles on ChessBase. And, of course, I seek improvement; but through study and practice, through an understanding of the concepts and principles involved. I would not wish to follow a set of moves like an automaton.
Can we?
Well, I suppose it can't be stopped or detected.
Should we?
Most certainly not; I have no wish to use or play against such a resource.
|
whitecat25
Chess rating: 2062
England
Give chess goodie
| Thu Mar 8 2012 2:23AM | MsgID: 15253281
I recently took a decision that I would not consult books for my correspondence games, although it is allowed. They only confused me, and with 40 years of otb experience, I should know by now what is playable and what is not.
I can however see the benefit of checking a potential endgame against a database, because against reasonable opposition there is no point in playing on in a drawn
endgame such as R & B v R. As I play a lot of games, I tend to offer the draw as soon as it it becomes trivial, as in bishops of opps a pawn up - that way, I can concentrate on my winning opportunities.
If I think that an opponent is using an engine to play their moves I will just resign the game - I can play against Fritz any time! On Chessworld, I have never found this to be a problem. Long may it continue!
|
bnolan
Chess rating: 2009
LCF 1865
United States
Give chess goodie
Chess goodies: 3 | Wed Mar 7 2012 11:32PM | edited: 11:33:09 | MsgID: 15253032
Originally posted by: "Theseus"
It may be a really stupid question, so please forgive my ignorance but: What benefit would this bring?
...
And what do we achieve - there's no competition, no decisions, no analysis, and because we've now actually stopped playing chess, absolutely no fun.
So,what's the point? |
That's a very good question. I think I would seldom be tempted to use a tablebase. I rather enjoy trying to figure out endgames on my own. Perhaps if I were offered a draw in a position with 6 or fewer pieces on the board, I might check the tablebase to see if it is a theoretical win or draw. But then I would likely continue to play on finding the moves myself.
The point of my original question, however, is not should we use tablebases, but rather can we? Is there an official stance? Or even a broad consensus?
|
Theseus
Chess rating: 2237
United Kingdom
Give chess goodie
Chess goodies: 3 | Wed Mar 7 2012 8:56PM | MsgID: 15252577
Originally posted by: "bnolan"
So my premise is that this is endgame analysis, performed by a computer, and published in an electronic database prior to the start of your current correspondence game, and should therefore be allowed.
|
It may be a really stupid question, so please forgive my ignorance but: What benefit would this bring?
All of the other resources that you mention require one to make a decision, a judgement on the validity of what's been previously researched. An opening that will guarantee a win doesn't exist; previous games, even those of the great, may be imperfect, and even computer checked lines may be unreliable (the prefect chess-playing computer programme has not yet arrived).
However, tablebases are different(so I'm led to believe); all possible variations have been played to conclusion. So, in a game, you set your tablebase to work, I set mine running (I don't have one and don't know how to use one but for the purpose of argument ...)and off we go.
Tablebase versus tablebase.
And what do we achieve - there's no competition, no decisions, no analysis, and because we've now actually stopped playing chess, absolutely no fun.
So,what's the point?
|
KristalC
Chess rating: 2393
United States
Give chess goodie
| Wed Mar 7 2012 4:19PM | MsgID: 15251762
In this link: http://www.k4it.de/index.php?topic=egtb&lang=en which was already mentioned in the thread I entered a 7 piece endgame position and I didn't get any analysis only the statement that there is "no information available" so it looks like
it is possible to passively analyze using tablebase information without an engine.
|
|
Play ... Latest Forum Posts > Chess Forums > Chess - General discussion
The case for tablebase use
