test position

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Uri Blass
Posts: 10420
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: test position

Post by Uri Blass »

Dann Corbit wrote: Mon May 13, 2024 7:09 am Jeremiah Pennery wrote draw detection code using pawn walls for Crafty.
Mike Byrnes used draw detection code using pawn walls for Black-Diamond-XI-NN.
Those engines can detect these kinds of draws instantly.
Unfortunately, the net effect is negative for overall Elo, so I do not think any current engine is using it.
Seems simple enough in concept...
The wall detection part is simple. You need at least 7 pawns (or bishops playing a pawn in the wall) and they need to guard each other and be rammed against opponent pawns (five or 6 self-defends). The two friendly pawn sentries at the edge of the hole in the column must be kept safe if not guarded by friendly pawns and all the squares behind the hole on the open file must be guardable. Three shifts+ANDs to figure out the pawn wall, and the hole guarding can probably be figured out by search.

Try Black-Diamond-XI-NN on this position. You will see that it solves it instantly.
In this case there is a stalemate idea so you cannot detect a draw only based on the pawn structure.
Dann Corbit
Posts: 12606
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: test position

Post by Dann Corbit »

re: "In this case there is a stalemate idea so you cannot detect a draw only based on the pawn structure."
Odd, that is what I do when I look at the position.
An impermeable wall with one open slot.
Can I defend the slot?
Yes, easily.

When you look at the position, how do you decide?
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.