Re: It is Strange

On May 2, 9:28=A0am, Sanny <softta...@[EMAIL PROTECTED]
> wrote:
> I remember once GetClub had 2 bishop and it was unable to give the
> Mate and Help Bot had to resign. Then Help Bot Complained this game is
> draw and I had to Resign to finish the game.

Help Bot is correct. If the program is so clueless that it cannot
execute the mate for KBBk then it is a draw under the 50 move rule.
>
> So now when Opponent has 2 Bishops Left I declared it a draw.
>
> Now, you are saying It is a Win?

YES! It is a fairly simple win for anyone with a modicum of chess
intelligence.
You basically squeeze the enemy king right into a corner and then
arrange a checkmate. KBNk is slightly harder (especially so in under
50 moves against an engine using tablebases).
>
> If I give it a Win in your case then why not give it a win for
> GetClubs case.

Because GetClub doesn't know how to win the position.
>
> Same situation both player has to be given same set of rules.
>
> If having 2 Bishops is draw when GetClub have them, then same is true
> for the Opponent.

Not at all. If the opponent has more than two brain cells to rub
together he will win. One such winning position with the unlucky loser
trapped in a corner looks like this.
_____________
| k |     |     |    |
|    | B |     |     |
|    | K |     | B |

It should be obvious what the last move for each side was.
>
> Moreove when the Game is Drawn your Rating is not affected. So in such
> case atleast you do not loose any point.

Yes. But I would feel cheated if having got to a KBBk, KBNk position
it was declared a draw by a program too stupid to win from that
position. Just because your program cannot win from these positions
(and so must accept a draw) does not mean that a human cannot win.

Your choices are to add a few choice heuristics to the endgame
evaluation function when the number of pieces is 4 or less and/or to
implement basic tablebases for the classic 3 & 4 piece endgames.

These would be roughly (in both KBBk and KBNk) to:

Make the winning sides king bolder and take the opposition with the
enemy king.
Restrict the mobility of the enemy king to 2 or 3 moves where
possible.
Restrict the region of the board available to the enemy king.
Winning side scores for forcing enemy king to edge of board, bonus for
corners.
Avoid stalemates (stating the obvious)

Losing side should aim to stay as near the centre of the board as
possible and maximise his mobility.

I suspect though I have not tried it that these heuristics in eval
would be enough to get a 10 ply search in range of seeing the mates
with minor pieces fairly reliably.

Regards,
Martin Brown