Help from the 8 pieces endgame database ?

[Ed Gilbert]

In the game Gantvarg vs Chizhov, round 16, according to the WC 2011 web site Gantvarg at move 58 played 37-32 which is a white loss, and is also the last move of the game, but the players agreed to a draw. Is this actually what happened?

[Rein Halbersma]

In the game Gantvarg vs Chizhov, round 16, according to the WC 2011 web site Gantvarg at move 58 played 37-32 which is a white loss, and is also the last move of the game, but the players agreed to a draw. Is this actually what happened?

That would be really surprising, as the win is a "mate-in-one".

[BertTuyt]

Damage also has a mechanism to detect a theoretical Draw (don't know which word to use after so many remarks ).
However this was ( i guess) a somewhat crippled design, and I hardly used it.
Ed, if you browse my source code of the search routine you will still see the traces of this attempt.

Bert

[BertTuyt]

When you look at the Sjende Blyn tab the move apparently played was 37-28.

Bert

[Ed Gilbert]

That's a very recent change at the web site then, because IIRC what first alerted me to the strange move was a large spike at the end of the SB graph, and now that is gone.

-- Ed

[hendrikv]

I removed the spike from the database, because i really doubt the move is really played. The LDF file shows no time difference between the moves, so i suspect the clock was already stopped.

I dont know why the database shows 37-28. i didn't change that.

Hendrik

[Piet Bouma]

In the game Gantvarg vs Chizhov, round 16, according to the WC 2011 web site Gantvarg at move 58 played 37-32 which is a white loss, and is also the last move of the game, but the players agreed to a draw. Is this actually what happened?

That would be really surprising, as the win is a "mate-in-one".

I have just got an e-mail from the organisation that the last move was 58.37-28.

[Ed Gilbert]

I was surprised that Artykow resigned in this 6x5 position, where both sides will soon have 1 king. Kingsrow thinks it is a draw. What do the other programs think?


black to move.

Edit: Oh, never mind. I just saw that Artykow probably ran out of time.

[BertTuyt]

Ed, you are right he lost for time reasons.
Damage also thinks it is a draw (although I don't have the 8p DB ).
Most likely sequence 44-49 7-2 49x46 2x24

Is this in line with KingsRow...

Bert

[Ed Gilbert]

Bert, kingsrow wants to play 44-50 7-1 50-33 1-18 33-47 18x4 47x36. Your 44-49 leads to an instant 8-piece draw, where 44-50 delays it for some moves.

-- Ed

[BertTuyt]

Herewith a position from the most recent KingsRow - Damage match. White (Damage) to play
Hope that the 8p DB owners know the theoretical outcome.
Anyway black (KingsRow) won this game.



Bert

[Ed Gilbert]

W:W7,24,45,49:B3,9,14,15,K20

Black win.

[Ed Gilbert]

I recently received this email from a kingsrow user:

Two years after...your debate about Famous classical position , do- you still maintain the same idea that Wouldouby is draw and Fabre-De Haas also ?

The user is referring to this forum discussion: viewtopic.php?f=65&t=1915&p=56191&hilit ... aas#p56191. The position from the Woldouby game is already covered in the kingsrow help file, but I had not looked at the De Haas vs. Fabre game in a while. The position in question:


White to move.

I will paste below one of my emails to this person. The email explains the process that I used to conclude that the position is a draw.

I am on my lunch break at work, and I decided to look at the Fabre De Haas position that you asked about. It was easy to see that this position is a draw using kingsrow and about 10 minutes of time. I don't have the full 8/9pc databases here at work, and this computer only has 2GB of usable memory, so the analysis would have gone more quickly if I was at home, but still it was pretty easy. I will explain what I did so you can try the same.

Here is the position:

[FEN "W:W25,27,28,30,32,33,34,35,37,38:B13,14,15,16,18,19,21,23,24,26"]
You can copy/paste this position directly into Kingsrow using normal Windows copy/paste mouse or key commands.

Start by searching the position in Analysis Allscores mode. You can see that the only white move that has any possibility of winning is 34-29, as all other moves show either draw or worse for white.

Make the move 34-29 and the 2 captures 23x34 30x39. Now it's black's move, and it is clear from kingsrow's search scores that black's best defense is 18-23. So move 18-23.

Now it's white's move, and you can immediately see that white's only chance for a possible win is with 27-22, as all other moves show either draw or worse.

Make the move 27-22, and now it is black's move. Clearly from the kingsrow scores black's only reasonable move is 23-29.

Make the move 23-29. After a 1 minute search, we can see that white has 2 possibly good moves; 37-31, and 39-34. All other white moves show database draw or worse. So we will have to play into both these lines to see where they lead.

Make the move 37-31, and 26x36 32x41. It's now black's move and it is clear that his only good move is 13-18.

Make the moves 13-18 22x13 19x8.

It is white's move, and from the scores we can see that white's only hope for a win is with 28-22. Play 28-22, then change kingsrow's search mode to Analysis (not Allscores). The reason is that Allscores is a slower search, and we want to concentrate all of kingsrow's power on this one position and see if we can get a database result. It takes less than 1 minute to see a convincing database draw result. So now we have shown that one of the two earlier white attacks is a draw.

Use the Back command to move back to the game position that exists after the move 23-29. Now we will try the other white alternative. This time, instead of 37-31, play 39-34 29x40 35x44. Set the search mode back to Analysis Allscores so you can see all the scores.

It's clear that black's only good move is 13-18. Play 13-18 22x13 19x8.

It is immediately clear that white's only good move is 28-22, as all others are draw or worse. Play 28-22.

Black's best move is 8-12. Play 8-12

White's only good move is 32-28, as all others are draw or worse. Play 32-28.

At this point we want the most powerful search, so change the search mode to Analysis.

Kingsrow is showing 12-17 as the best move, so play 12-17 22x11 16x7.

Kingsrow now shows a conclusive database draw result. This concludes the analysis. We have looked at every white possible line of attack, and they have all shown to be draws.

-- Ed

I later tried the same analysis using only the 7pc database. I had to let a few of the searches run for a little longer period (2 - 3 minutes max), but I got the same definitive results.

-- Ed

[Rein Halbersma]

I recently received this email from a kingsrow user:

Two years after...your debate about Famous classical position , do- you still maintain the same idea that Wouldouby is draw and Fabre-De Haas also ?

The user is referring to this forum discussion: viewtopic.php?f=65&t=1915&p=56191&hilit ... aas#p56191. The position from the Woldouby game is already covered in the kingsrow help file, but I had not looked at the De Haas vs. Fabre game in a while. The position in question:


White to move.

-- Ed

Hi Ed,

Note that Wieger Wesselink also analyses this position at (with the same result of course):
http://10x10.dse.nl/analysis/2011/wkwom ... index.html

For a future Kingsrow release, it might be nice to add your analyses of the Woldouby, DeHaas-Fabre and Otgonyabar-Podolski classical games (and their mirror positions) to the opening book so that users can quickly browse these lines.

Rein

[Ed Gilbert]

Hi Rein,

It looks like the position I copied from the forum discussion is the colors-reversed mirror of the position that Wieger references as "de Haas - Fabre 0-2 (1921)". No matter, the analysis process and conclusion are the same, just the move numbers are different.

-- Ed