GS-4 tournament scoring problems

[Gabor Szots]

Geschrieben von: / Posted by: Gabor Szots at 29 March 2000 11:04:07:
I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?
4. Crux exited unexpectedly 3 times. That means, three of its opponents played only 6 games altogether, therefore being handicapped compared to the others. Should they have been awarded a win?
Can someone show me a light?
Gabor

[Colin Frayn]

Geschrieben von: / Posted by: Colin Frayn at 29 March 2000 16:20:39:
Als Antwort auf: / As an answer to: GS-4 tournament scoring problems geschrieben von: / posted by: Gabor Szots at 29 March 2000 11:04:07:

I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?

I don't know how your system is set up, but on mine Xboard (or Winboard) will automatically stop the game when there is insufficient material left (or any other kind of draw). ColChess would have scored its game against Averno as a draw (and might have even played towards this ending, knowing it to be drawn), but would not claim a draw as I'm used to Xboard doing that for me. It would probably have said that it was a draw if you were playing in text mode, but it wouldn't have done so in Winboard mode.
I think they should all be scored as draws personally, but for the official word from the FIDE rules.....
"9.6. The game is drawn when a position is reached from which a checkmate cannot occur by any possible series of legal moves, even with the most unskilled play. This immediately ends the game."
Strictly speaking, the other two positions (by repetition) have to be claimed by either player, so you might not want to count those as draws, but the ColChess vs. Averno game was a strict draw.

Not that I'm begging for 1/2 a point or anything....
Cheers,
Colin

ColChess Homepage

[José Carlos]

Geschrieben von: / Posted by: José Carlos at 29 March 2000 17:41:51:
Als Antwort auf: / As an answer to: Re: GS-4 tournament scoring problems geschrieben von: / posted by: Colin Frayn at 29 March 2000 16:20:39:

I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?

I don't know how your system is set up, but on mine Xboard (or Winboard) will automatically stop the game when there is insufficient material left (or any other kind of draw).
ColChess would have scored its game against Averno as a draw (and might have even played towards this ending, knowing it to be drawn), but would not claim a draw as I'm used to Xboard doing that for me. It would probably have said that it was a draw if you were playing in text mode, but it wouldn't have done so in Winboard mode.
I think they should all be scored as draws personally, but for the official word from the FIDE rules.....
"9.6. The game is drawn when a position is reached from which a checkmate cannot occur by any possible series of legal moves, even with the most unskilled play. This immediately ends the game."
Strictly speaking, the other two positions (by repetition) have to be claimed by either player, so you might not want to count those as draws, but the ColChess vs. Averno game was a strict draw.

Not that I'm begging for 1/2 a point or anything....
Cheers,
Colin

I might be wrong, but I don't think Winboard has that "knowledge". Instead, the program has to claim the draw by sending to winboard "1/2-1/2" and the reason why it is a draw (repetition or whatever). In that case, Winboard accepts the result even if it's wrong.
Try implementing what I said above. It is very easy if you already know that the position is a draw. My program doesn't know it, and so it goes happily to that ending, thinking "hey, I'm winning! I'm a knight up!
Anyway, I've implemented claiming draw in case of repetition or 50 moves rule.
Ok, ok... You can have your 1/2 point, Colin But I'll start right now "to teach" my program that such position is a draw... (stupid Averno...)
José C.

[Colin Frayn]

Geschrieben von: / Posted by: Colin Frayn at 29 March 2000 18:29:36:
Als Antwort auf: / As an answer to: Re: GS-4 tournament scoring problems geschrieben von: / posted by: José Carlos at 29 March 2000 17:41:51:

I might be wrong, but I don't think Winboard has that "knowledge". Instead, the program has to claim the draw by sending to winboard "1/2-1/2" and the reason why it is a draw (repetition or whatever). In that case, Winboard accepts the result even if it's wrong.
Ok, ok... You can have your 1/2 point, Colin But I'll start right now "to > teach" my program that such position is a draw... (stupid Averno...)

You may well be right - I've just seen it claim automatic draws in a lot of games between ColChess and other programs, but mostly it's been Crafty and similar, so you'd expect them to claim pretty much everything correctly.
Hehe thanks
Cheers,
Colin

ColChess Homepage

[WYx]

Geschrieben von: / Posted by: WYx at 29 March 2000 20:05:05:
Als Antwort auf: / As an answer to: GS-4 tournament scoring problems geschrieben von: / posted by: Gabor Szots at 29 March 2000 11:04:07:

I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?
4. Crux exited unexpectedly 3 times. That means, three of its opponents played only 6 games altogether, therefore being handicapped compared to the others. Should they have been awarded a win?
Can someone show me a light?
Gabor

Hi
Gabor.
Warum wickelst du so deine Tour ab, dass eines Programm mit Zeit verlieren kann?
Bei mir es ist unbekannt!!
Ich probiere immer gerechte Tour machen!!
Probiere auch, bitte.
Gruss
WYx

[Gabor Szots]

Geschrieben von: / Posted by: Gabor Szots at 30 March 2000 08:19:46:
Als Antwort auf: / As an answer to: Re: GS-4 tournament scoring problems geschrieben von: / posted by: Colin Frayn at 29 March 2000 16:20:39:
(snip)

I don't know how your system is set up, but on mine Xboard (or Winboard) will automatically stop the game when there is insufficient material left (or any other kind of draw). ColChess would have scored its game against Averno as a draw (and might have even played towards this ending, knowing it to be drawn), but would not claim a draw as I'm used to Xboard doing that for me. It would probably have said that it was a draw if you were playing in text mode, but it wouldn't have done so in Winboard mode.
I think they should all be scored as draws personally, but for the official word from the FIDE rules.....
"9.6. The game is drawn when a position is reached from which a checkmate cannot occur by any possible series of legal moves, even with the most unskilled play. This immediately ends the game."
Strictly speaking, the other two positions (by repetition) have to be claimed by either player, so you might not want to count those as draws, but the ColChess vs. Averno game was a strict draw.

Not that I'm begging for 1/2 a point or anything....
Cheers,
Colin

Hi Colin,
In fact, I did award a draw. Only I forgot to change the result tag in the PGN file sent to Frank, which he used for his calculations of final standings and cross table.
Besides, I don't know how to set Winboard to stop the game automatically when there is insufficient material.
Cheers,
Gabor

[Gabor Szots]

Geschrieben von: / Posted by: Gabor Szots at 30 March 2000 10:46:37:
Als Antwort auf: / As an answer to: GS-4 tournament scoring problems geschrieben von: / posted by: Gabor Szots at 29 March 2000 11:04:07:

I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?
4. Crux exited unexpectedly 3 times. That means, three of its opponents played only 6 games altogether, therefore being handicapped compared to the others. Should they have been awarded a win?
Can someone show me a light?
Gabor

Well, I decided to award a draw in insufficient material positions (whether claimed or not), but wait for a claim in threefold repetition and 50-move rule situations (in accordance with FIDE rules). To avoid losses on time in such situations, I think it is better not to play game in x minutes games, but use time increment or conventional clock instead.
I still don't know what happens when one program claims a draw based upon one of the above conditions, but its opponent doesn't understand this claim. Will Winboard stop the clocks? Why? WB is not capable of checking the validity of a claim. There may be a false draw detection. And in that case I probably won't be able to make the engines continue playing (if I am capable at all to check for the validity of the claim) since the same erroneous detection will occur. Am I right?

As for item No4, I think such games should be replayed unless in a a clearly won position for the innocent side. If 'exitus' happens say three times in a row, the opponent is awarded the full point. And if a program makes such exits regularly, it should be disqualified, regrettably.

[José Carlos]

Geschrieben von: / Posted by: José Carlos at 30 March 2000 22:15:44:
Als Antwort auf: / As an answer to: GS-4 tournament scoring problems geschrieben von: / posted by: Gabor Szots at 30 March 2000 10:46:37:

I faced the following problems during my GS-4 tournament:
1. ColChess-Averno was won by black on time, although black had nothing left but king and knight, while white had only the lone king. Should I have awarded a draw?
2. Faile-Sjeng was won by white on time after they had played almost 150 (!) moves without capturing or moving a pawn. Should I have awarded a draw?
3. Once again, Crux-Snail was won by black on time, after they had repeated moves 4 times. Should I have awarded a draw?
4. Crux exited unexpectedly 3 times. That means, three of its opponents played only 6 games altogether, therefore being handicapped compared to the others. Should they have been awarded a win?
Can someone show me a light?
Gabor

Well, I decided to award a draw in insufficient material positions (whether claimed or not), but wait for a claim in threefold repetition and 50-move rule situations (in accordance with FIDE rules). To avoid losses on time in such situations, I think it is better not to play game in x minutes games, but use time increment or conventional clock instead.
I still don't know what happens when one program claims a draw based upon one of the above conditions, but its opponent doesn't understand this claim. Will Winboard stop the clocks? Why? WB is not capable of checking the validity of a claim. There may be a false draw detection. And in that case I probably won't be able to make the engines continue playing (if I am capable at all to check for the validity of the claim) since the same erroneous detection will occur. Am I right?

As for item No4, I think such games should be replayed unless in a a clearly won position for the innocent side. If 'exitus' happens say three times in a row, the opponent is awarded the full point. And if a program makes such exits regularly, it should be disqualified, regrettably.

If a program claims a result, winboard automatically accepts the result as correct, even if it is not correct. I think you can't do anything about that, except deciding the result yourself...
When two programs repeat positions for many moves, and both of them don't know it is a draw, I usually asing a draw result. If you set a time control with increment, you will probably wait forever.
José C.

[Gabor Szots]

Geschrieben von: / Posted by: Gabor Szots at 31 March 2000 07:46:10:
Als Antwort auf: / As an answer to: Re: GS-4 tournament scoring problems geschrieben von: / posted by: José Carlos at 30 March 2000 22:15:44:

If a program claims a result, winboard automatically accepts the result as correct, even if it is not correct. I think you can't do anything about that, except deciding the result yourself...
When two programs repeat positions for many moves, and both of them don't know it is a draw, I usually asing a draw result. If you set a time control with increment, you will probably wait forever.
José C.

You are right. That means, there must be human intervention until the programs will be 'perfect'.
Best regards,
Gabor