I think abolishing the 50 move rule when a tablebase position is reached makes very very few practical difference.
Can you give an example of an actual correspondence game that was drawn because of the 50 move rule but would have been won without the rule? I've never seen one

That there are big differences in rules about how, where and when the moves are played, seem to me normal between correspondence and OTB. It is another thing if we start to change the basic rules of the board itself even if as small as a 50 moves rule.

For pure correspondence players (so not active in OTB-play) this change will likely not be a problem. However assume an active OTB player playing correspondence to learn something about his petline. At some moment the player can achieve a TB which wins but will also need to break the 50 moves rule. Seems to me a pretty annoying dilemma. Or you avoid the certain correspondence win to find out more about the correct OTB evaluation, risking not to win the game or you go for the immediate correspondence win but you are afterwards unsatisfied as you weren't able to fully test your petline.

TalJechin wrote on 09/18/13 at 12:11:34:

I red on a german forum: http://f3.webmart.de/f.cfm?id=2035593&r=threadview&t=3995685&pg=1 that there are a lot of discussions still about which TBs will be accepted and which not.
Can we have 100% guarantee that the TBs don't contain errors?
Which TBs will be accepted? Only official ones available for the public or are own generated TBs also good as proof?
This last item can be important for defining the result.

Assume you have a specific 8 piece TB generated which proofs a win (e.g. via the finalgentool). However you need to break the 50 move-rule to achieve a win. For TBs accepted by ICCF you are allowed to break this rule so the win in such scenario depends if your TB will be accepted or not.

TalJechin wrote on 09/18/13 at 11:41:33:


I have the feeling that with touching the basic rules of what is a win/draw/loss a new important line has been crossed. The door is open for more changes to the basic rules. In fact with introducing in the last congress separate ratings for 960Chess, ICCF is already showing that they are not afraid of making further steps.

Today I read on http://www.iccf-webchess.com/Message.aspx?message=557 that they decided to cancel the 50 moves rule when a tablebase is entered. To be able to do that they even had to change the statutes which before didn't permit to deviate from fide.

To understand the impact, i show the position below which i covered in my latest blogarticle: http://schaken-brabo.blogspot.be/2013/09/schaakintuitie-deel-2.html


(white to move)

If you use the finalgentool (which can be found on http://www.mtu-media.com/finalgen/home_ing.php) then you will notice that black needs 85 moves to win.
Now it isn't fully clear to me if this is feasible without breaking the 50 move rule. I found variations with suboptimal white moves which went beyond the 50 move rule but it is as well possible that suboptimal black moves could still do the job without breaking the 50 move rules (so winning in more than 85 moves in total). 

From above it is clear that it isn't always simple to define if a tablebasewin can be done without breaking the 50 move rule. As iccf wants to give players the possibility to make claims based on tablebaseverdicts, likely the decision is made not to take the 50 move rule anymore into account.

The benefit of this decision is that players don't have to play out a tablebaseposition anymore for many extra months (years?) to get the expected result validated. The disadvantage of course is that there is no clear link anymore between the fide and iccf basic rules.

The decision doesn't surprise me. It is a logical step compared with the previous iccf decision to hide their databases for non-iccf members (so mainly fide-players). I've the impression that ICCF has chosen to distantiate themselves clearly from fide. Unclear to me if this is a good or bad thing.