The example given was to show what the system actually does in that given position with White to move and Black's flag falls i.e. it's a draw, while in reality Black is lost. It's hypothetical as many of these positions are not likely to turn up in play. Yet they still affect the laws governing chess.
Draws Declared By Remaining Mating Material Rather Than Possibility Of Checkmate Is Illogical
Sort:
Jun 14, 2023
0
#23
These are very good points. Have you approached the site directly about them?
Yes, I have a similar thread in the site feedback section and directly in the help and support club. I get disagreement that it's "not worth programming", yet they went through the trouble of putting in stats pages for king of the hill and 3 check chess LOL
Jun 14, 2023
0
#26
I think it's relevant. I will raise this as an issue through Arbiting contacts I have.
oldwoof wrote:
I think it's relevant. I will raise this as an issue through Arbiting contacts I have.
Thanks! That's awesome! Feel free to Show them the positions in this thread if you want!
Jun 15, 2023
0
#28
oldwoof wrote:
These are very good points. Have you approached the site directly about them?
There was a long post 5 or 6 years ago where the insufficient material rules were discussed and staff were involved and decided on the implementation that the site has today, with one minor change I believe in the intervening years. While it's possible they might revisit it, it's not highly likely (unless they were already planning on changing it)
Jun 15, 2023
0
#29
The help articles even specifically call out the site is different than FIDE
Jun 15, 2023
0
#30
Thanks for the time and trouble in explaining. I have had a response from a couple of International Arbiters. They know about this as well and it's accepted that all reputable sites and otb are different in this respect.
What I don't understand is why this site wouldn't decide on the most fundamental rule for this when programming it, yet they put in stats pages for 3+ chess and king of the hill variants? I also didn't know USCF considers 2 knights insufficient mating material. Because it is easier to blunder into than KNKB for example.
Jun 16, 2023
0
#32
wait
in what sort of logic system is 2 ns insuff?
theres a mating position...
I know, it is illogical. Since when is the forcibility of mate the determining factor of whether it's sufficient material or not? I know that's not in the game of chess! Why not just auto-declare draws when my king is in front of his rook pawn since he can't force a win LOL?
Jun 16, 2023
0
#34
i dont really get n v b not being there but whatever
and dead positions are honestly a bit hard to code
but theres literally a mate position
its like r v r
mate is possible there
but for that one it isnt a draw
but here it is?
and in both, there are certain ones where its forced; and both times its part of a simplification sequence (2r v r then the single r takes one of the 2 rs and it is going to be taken back next move, then a queen trade in qnn v q around the edge (so its forced))
these two are functionally the exact same and its easy to code for both
just remove the bit where 2 ns are considered insufficient, and make it sufficient (it is, even on an empty board)
and dont even mention the nn v p positions where its like forced mate in 20, is that also called a draw here?
No position should be declared a draw where checkmate is possible in any way for the opponent. If you have 9 queens and 2 rooks and you are still afraid of losing on time to his 1 knight, then sacrifice for the knight and you have nothing to worry about. 2 knights vs a pawn could be a forced mate in 19 moves such as below:
Jun 16, 2023
0
#36
EndgameEnthusiast2357 wrote:
What I don't understand is why this site wouldn't decide on the most fundamental rule for this when programming it, yet they put in stats pages for 3+ chess and king of the hill variants? I also didn't know USCF considers 2 knights insufficient mating material. Because it is easier to blunder into that KNKB for example.
They did decided on the rules to use, long before there were stats pages for those features. I can't ever say they won't revisit it, but it seems very unlikely.
Hii
Martin_Stahl wrote:
EndgameEnthusiast2357 wrote:
What I don't understand is why this site wouldn't decide on the most fundamental rule for this when programming it, yet they put in stats pages for 3+ chess and king of the hill variants? I also didn't know USCF considers 2 knights insufficient mating material. Because it is easier to blunder into that KNKB for example.
They did decided on the rules to use, long before there were stats pages for those features. I can't ever say they won't revisit it, but it seems very unlikely.
Do you think of they saw some of my examples in this thread they would reconsider it?
Jun 17, 2023
0
#39
EndgameEnthusiast2357 wrote
Do you think of they saw some of my examples in this thread they would reconsider it?
They are completely aware there are some edge cases where a mate is forced but the algorithm gives a draw in timeout conditions.
Here is a famous endgame study that shows the problem. Immediately after bg8, the site will declare the game a draw even if neither side lost on time. Here's another cool one with the same ending position:
In both positions inherent moves lead to various forced wins, yet in some cases the losing side could actually deliberately let his time run out to get a draw as a result of this site's implementation!