This is a forum for members of the All India Correspondence Chess Federation.
Only members can post.
From the ICCF congress August 2013, which can be seen here, A new rule has been added.
"Table bases are introduced. 50-move rule suspended if TB is entered. Playing Rules committee along with the Service Director, under oversight of the EB will address the technical issues of implementation."
" Convekta Ltd (www.chessok.com) and ICCF have signed a cooperation agreement under which ICCF will have a free access to the endgame tablebases developed by Convekta Ltd and hosted on their website.
The 2013 ICCF Congress in Krakow (Poland) has decided that for all tournaments started after 1/1/2014, the players will be allowed to claim a win or a draw if the position can be resolved in a 6 men tablebase position.
And According to ICCF Webmanager Gino Figlio, It is limited upto 6 men now.
So I suggest as we are affiliated to ICCF , we should also follow the rules. 3,4,5 and 6 men Tablebase positions are easily available online at
So I suggest
1)using DTM (Distance to mate avoiding 50 move rule) and not DTZ (distance to zero using 50 move rule which is now irrelevant)in website.
2) If Arbiter doesnot want to implement in website then Players should have the right to claim Win draw loss (WDL) using DTM tablebases to arbiter whenever the game enters 6 men position without waiting for adjudication time.
Correspondence Chess Player/Computer Chess Enthusiast
I was talking not to follow DTZ50 using 50 move rule. We can use WDL + DTZ. Syzygy bases are the ones which considered the 50 move rule first in TB. But Nalimov bases are DTM metric which uses distance to mate. While Syzygybases use a variation of the DTZ metric (distance-to-zero, zero meaning any move that resets the 50-move counter). This special metric is one of the reasons that Syzygybases are more compact than Nalimov tablebases, while still storing all information needed for optimal play and in addition being able to take into account the 50-move rule.
My main aim is to follow ICCF rules.
Regarding educating those who dont know, It can be done easily throgh bulletin. But first we have to follow the rules and post the rules in server. That is how we can come out of stone age. Just because there are ignorant masses, it should in no way stop the ones who are not ignorant. Time is precious now a days. We need to follow steps to save time.
We should take account of all members opinions before deciding. We are not bound to follow ICCF rules - but we should take stock of them.
1. We want automation as much as possible. ICCF server is not that much automated and was not at all initially. Gradually they are also moving towards automation. By automation we mean no human intervention in games as far as is practicable. You will notice that ICCF has abolished special leave, we avoided it from the start, because it needs human intervention. My opinion is that server games should not be supervised (except to correct server problems). Manual effort of emails going back and forth defeats the purpose of the server. ICCF never automated a simple thing like the 50-move rule. We automated it from day 1.
2. FIDE rules of chess should be used. Since the 50-move rule is a basic law of chess it should not be altered. ICCF too is affiliated to/recognized by FIDE. We should not alter the basic rules of chess. The 50-move rule was introduced in OTB with a reason. That reason is equally applicable to CC. Positions which need more than 50 moves to mate (e.g. K+N+N vs k+p) were well known a century ago, long before the advent of computers, but the 50-move rule was allowed to stand.
3. Claim on table-base is a different issue than the 50-move rule. We can use tablebases that respect the 50-move rule. At present we admit the claim, but only after the adjudication date.
4. If you are playing against a weaker opponent and he loses a piece on the 3rd move (1.e4 e5 2.Nf3 Nf6 3.Nxe5) can you claim the game? Tablebases for ignorant players are a bit like that. If you know about it, and understand it, then it seems silly that we should have to play on. But if the opponent was not ignorant he would have resigned anyway. Likewise in the game above. Problem is that the opponent thinks its his right to play.
My opinion is that we should change the AICCF rules and allow a claim of tablebase (6-men) win or draw at any time (not only after crossing the adjudication date), but we should stick to the 50-move rule. We should look into the possibility of tabulating all possible 6-men positions and classifying most of them (if not all). For example the basic mates can easily be stored and even some others - so that a substantial amount of automation is possible and less human intervention will be required. It wont require 150GB (thats for storing also all the move trees, while we need only positions), sometimes only generic positions like "All KQ vs k", "All KBB vs k". As soon as such a position is reached the server should close the game without human intervention.
I believe that AICCF should start becoming modern in their outlook by accepting technological advancements. CC is a patient, time consuming activity requiring lot of effort and especially when engines are nearly at ELO of 3100.
I do not see any point in prolonging the game when tablebase has been reached. CC is totally different from OTB and as such following FIDE is not necessarily the right course.
Given all the above, my considered view is that AICCF must follow DTM policy (the 50 move rule is irrelevant for cc in my view).
Further I agree with Omprakash's view that "Time is precious now a days. We need to follow steps to save time."
The arguments that players are not aware of table bases and will be annoyed if tablebases are used are irrelevant in the context of cc (if one is unaware of tablebase then one does not stand any chance in cc and aiccf should not be catering to these segments.)
In conclusion AICCF must make the bold (!?) move to embrace tehcnology especially the tablebases (DTM variety).
1) I think your primary request is to permit adjudication on reaching 6 men positions, irrespective of 50 move rule or not.
2) Here is an example where two ICCF GMs continued to play 10 more moves after reaching a 6 men tablebase position: http://www.iccf-webchess.com/game?id=417889!
Yes, I agree that AICCF should also allow claims of win or draw on the basis of 6-men tablebases. It is not practical to store these large tablebases in the server site and hence automation should not be a goal. The problem with keeping the 50-move rule intact is that it would require to check how many moves have already been made without pawn moves and captures and the number of moves further needed (from the tablebase) to mate. So to do away with the 50-move rule is needed, as otherwise we would spending a lot of time in examining these claims. Perhaps that is why ICCF must have done so. Members who choose to remain unaware would simply play on. If both sides play on, then it wont be of our concern. We can announce this rule change in the February 2014 bulletin and include it in the Webserver Guide. We can fix a date and it can only come into effect for games that are STARTED after that date.
I had another curiosity (not relevant to the above). In the tablebases, if there is no limit to the number of moves, then how can you establish proof of a draw? Does the software (which created the tablebase) search on-and-on until the position repeats?
My curiosity question about establishing draws was in the context of how tablebases are computed. (Once computed, it is stored.) Tablebases are computed backwards. We start with all possible mate positions with 6-men or less. Then play moves backwards, to get positions. This backwards movement is very interesting. At each point, if we have less then 5-men, then we can "create" a piece by undoing a capture! Pawns move backwards and a queen or piece on the last rank can change into a pawn by undoing a promotion! If the 50-move rule is not applied, then we have to keep going without any limit until we have exhausted every possible position.
My first thought was that in every game the server could detect that a 6 or less-men position is reached and warn the players. (We could even alert the administrator). I was even thinking of requesting the online endgame tablebases sites to provide an automated query function in which any CC chess server (such as ours) can send a server-to-server query, submitting a FEN position, and get a reply of Win, Draw (or invalid 6-men position). This would be of use to ICCF and other CC servers as well.
However, looking at the complications involved, and the issues with players who would prefer to remain oblivious of table-bases, it is simpler to include a "claim table-base win/draw" button as soon as the piece count is 6 or less. This would generate an email to the administrator and an alert on the Administration Maintenance page (emails are not 100% reliable). The game should be continued while awaiting admin action. The administrator would end the game after verifying, and it would show on the server notices. By ensuring that a tablebase claim is possible only when the piece count is 6 or less, there will always be a result of win or draw, hence there wont be any need for the administrator to send emails to the concerned players. In case of any delay by the administrator, the players must continue as time-limit and silence rules would apply.
Dr. Chatterjee has talked to me and he is willing to implement adjudication of positions with less than 6 men for mate/draw once such a position has reached. But only one or two players continue to play even in clear mating or theoretically drawn positions. We should avoid human intervention as much as possible but in this case somebody has to adjudicate as soon as the less than 6 men position is reached and mate or draw is a certainty. If Dr. Chatterjee is willing to adjudicate, it is fine.
Absolutely, Om Prakashji, Dhanishji, Ambarji and Jayaramji are right in line with where we need to be:
1. We do need to move to accepting the table base once we reach that position.
2. It is not just an ICCF rule, which we should follow, but it is an extension of AICCF rules as it is. Already certain positons like K v K+N etc. are automatic draws on AICCF server. It is a waste of time , once we reach a drawn position to continue playing.
3. Yes, all the players need to be educated, which can be done through a clear post on the server login page or in AICCF publications.
I too agree with the view that we should use the table bases when there are 6 or lesser pieces left in the position. Not so sure about automating it, but probably an indication(email reminder to the adjudicator) should be good enough, if we are able to detect the number of pieces in the position easily.
Dear Anuj, Dear All,
It is easy for the server to count the pieces. (Presently the AICCF server detects mate, counts moves for the 50 move rule, and knows about all positions which are drawn due to insufficient material).
The AICCF site cannot store the endgame tablebase as it is too big.
We could ask the maintainers of tablebase sites if they would like to support server-to-server query and if they do that, AICCF server can itself query the tablebase and declare the result at once. This is 100% automation and ideal. But what if the tablebase site is down or develops a technical problem? So we can instead have the AICCF server maintainer (i.e. myself) carry out this task manually.
As soon as a game reaches 6 men, the server will know about it. Now we have two possibilities:
(a) Hard action: Stop the game and adjudicate.
Note that the game cannot reach 5 men or less as the game is already stopped at the stage of 6 men. As the game is stopped, any (small) delay on the part of the maintainer will not matter. There wont be any action needed by the players. The change in software is minimal and there is no visible change in the user interface.
(b) Soft action:
Give the player an option to continue the game or ask for adjudication. If he continues, then the opponent also needs to have the option to adjudicate or continue. The option has to remain available for the remaining duration of the game, allowing an adjudication request for 6 men or less. At any point if adjudication is asked for, the game would be stopped and any small delay on the part of the maintainer would not matter.
My preference is for option (a), but I need an opinion which option (a) or (b) to implement.
My preference is for Option a.
I also agree with Om Prakashji and Mohanji and go with option A.
Correspondence Chess complements OTB Chess. CC adds a lot of knowledge on positions because of extensive research on lines using computer engines and databases. The 50 move rule is relevant in OTB play because there is a genuine time constraint ( Many 9 round tournaments now complete in three days.). Correspondence Chess tests the limit of available Chess Knowledge, so if a position is resolved in 538 moves through an endgame database, that should be perfectly acceptable.
As no game should be allowed to continue for an indefinite period, some rules need to be introduced in chess. The logic behind the 50-move rule is to make a chess game definitely finite. We therefore observe there is an important parameter beside others, everyone should keep in mind to win a chess game, is the 'time-factor'. Let us examine a position where Black's lone king, White's King, Bishop and a Knight are on the four corner square and its white to move. OK, as per table-base, white can mate Black in 30 moves! But, if it were an OTB game and if White has only 10 seconds time left with no increment, is it possible for white to win the game? No Sir, he cannot. Now, suppose it were a CC game and White has only 1 hour time left. Imagine the situation for White who has to send 30 moves in 1 hour! I think this is also next to impossible. Both the players started the game with equal 'time' allotted to them initially. But despite getting a table-base winning position, White may not win due to the reason that he has consumed plenty of hours to reach the position. OK, I agree using computer in OTB Chess is cheating, but not in CC; but, does this statement stands enough to under rule all the basic factors responsible for a win in Chess e.g. space advantage, material advantage, winning advantage or even time pressure factor? There are so many examples in Chess where a player with huge advantage lost the game to his opponent having even only a pawn on the board due to 'Flag-down'. 'Time'-factor stands equally important in CC. , I could have no objection to be in the opinion of scrapping 50-move rule in CC, if we could withdraw 'time'-limit in CC first. As this is also not possible, We should have to wait long before introducing table-base adjudication ignoring 'time-factor'.
I beg to differ with your views. CC has to be treated differently from OTB (use of opening databases, engines, tablebases, books, the past methods of adjudication of games etc to name a few) and hence the analogy given does not stand on its merit.
I believe that AICCF should make bold the move to adopt technology and use the tablebases to decide when 6 man positions are reached (which is another form of adjudication by the way).
As said earlier my preference is for Option a. which should be implemented as soon as feasible.
Shri Gautam De has made his point and given his opinion. We know that Shri Sailesh Chandra, Shri Mohan Jayaraman and Shri Om Prakash are in favour of tablebase wins. We must be democratic and await opinions of others as well. The situation presented by Shri Gautam De is to highlight a point and need not be taken literally. I do see his point. As far as my personal opinion is concerned, I feel that we will have to accept tablebases as a reality.
I agree with option a too. Time factor is never a 'real' problem in CC. Either the person has lost interest and is not playing. Otherwise if he/she is winning or has big advantage, he will always have enough time on his/her side. We count in terms of days and not hours in CC. (even with real time introduced now). So option a would be preferable I think.
On 18th March 2014, AICCF took a progressive step and implemented tablebase adjuciation (6 men or less) with immediate effect. Some games on AICCF server have already been adjudicated like this. So we are ahead of ICCF as they will be applying this only to games starting after 1st April 2014.
I thank Mohan Jayaraman and Om Prakash for the strong push they gave in steering us in this direction and also all the members who have posted to the discussion in this thread.
Actually the fact is we are Privileged to have Dr Ambar jee as our chair of AICCF who atleast understands that we have to go according to time. I only initiated to save time of our members. I also thank to our Mohan bhai. It's not long we will have 8 men at our disposal. Thanks to Ambar jee. But we people are always ahead of ICCF. Our site is far better than ICCF.