The Unofficial World Championship Of Computer Programs 2024

Discussion about development of draughts in the time of computer and Internet.
Joost Buijs
Posts: 471
Joined: Wed May 04, 2016 11:45
Real name: Joost Buijs

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Joost Buijs » Thu Sep 19, 2024 10:10

Krzysztof Grzelak wrote:
Thu Sep 19, 2024 09:51
Joost Buijs wrote:
Thu Sep 19, 2024 09:45
Hi Krzysztof,

Do you have all the games from the applet in .txt format? if you do please mail them to me than I can convert all the games to PDN format.
It is too much work to check all the games in the PDN by comparing them manually with the moves shown in the applet.

Joost
I only have it in pdn format. You can copy everything to notepad and you will have a file .txt.
Then it will still have the errors in it. Now I understand that you typed in all moves by hand and that the moves shown in the applet are not saved to disk at all. Well, than it can only be solved by copying and pasting the moves from the applets to a .txt file, which is for 240 games a lot of work too.

Krzysztof Grzelak
Posts: 1368
Joined: Thu Jun 20, 2013 17:16
Real name: Krzysztof Grzelak

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Krzysztof Grzelak » Thu Sep 19, 2024 10:20

Joost Buijs wrote:
Thu Sep 19, 2024 10:10
Then it will still have the errors in it. Now I understand that you typed in all moves by hand and that the moves shown in the applet are not saved to disk at all. Well, than it can only be solved by copying and pasting the moves from the applets to a .txt file, which is for 240 games a lot of work too.
Look here and you will understand viewtopic.php?f=53&t=8679 .

Joost Buijs
Posts: 471
Joined: Wed May 04, 2016 11:45
Real name: Joost Buijs

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Joost Buijs » Thu Sep 19, 2024 10:30

Krzysztof Grzelak wrote:
Thu Sep 19, 2024 10:20
Joost Buijs wrote:
Thu Sep 19, 2024 10:10
Then it will still have the errors in it. Now I understand that you typed in all moves by hand and that the moves shown in the applet are not saved to disk at all. Well, than it can only be solved by copying and pasting the moves from the applets to a .txt file, which is for 240 games a lot of work too.
Look here and you will understand viewtopic.php?f=53&t=8679 .
I understand it perfectly, and was just looking for .txt files that contain the moves as they are shown in the applet, but as it seems there aren't any.

Krzysztof Grzelak
Posts: 1368
Joined: Thu Jun 20, 2013 17:16
Real name: Krzysztof Grzelak

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Krzysztof Grzelak » Thu Sep 19, 2024 10:32

The DGT board saves on the server and on the disk, but in this way.

[Event "The Unofficial World Championship Of Computer Programs In International Draughts 2024"]
[Site "Kornik - Poland"]
[Round "ronde20"]
[White "Ares 1.81 NNUE"]
[Black "Dream Gaia 1.3"]
[BoardId "1143"]
[GeneratedBy "EBS2.05"]
[Result "2-0"]
[FEN "W:W31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50:B1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20"] {[%clock W0:10:10 b0:10:10]}
1. 31-26 {[%clock w0:10:04 B0:10:10]}
19-23 {[%clock W0:10:12 b0:09:54]}
2. 34-30 {[%clock w0:10:05 B0:10:03]}
20-25 {[%clock W0:10:14 b0:09:51]}
3. 37-31 {[%clock w0:10:10 B0:10:01]}
25x34 {[%clock W0:10:19 b0:09:55]}
4. 39x30 {[%clock w0:10:14 B0:10:04]}
15-20 {[%clock W0:10:23 b0:09:47]}
5. 44-39 {[%clock w0:10:19 B0:09:56]}
10-15 {[%clock W0:10:28 b0:09:41]}
6. 41-37 {[%clock w0:10:14 B0:09:51]}
13-19 {[%clock W0:10:23 b0:09:35]}
7. 30-25 {[%clock w0:10:13 B0:09:44]}
08-13 {[%clock W0:10:22 b0:09:28]}
8. 40-34 {[%clock w0:10:12 B0:09:37]}
02-08 {[%clock W0:10:20 b0:09:25]}
9. 46-41 {[%clock w0:10:06 B0:09:34]}
20-24 {[%clock W0:10:15 b0:09:20]}
10. 34-29 {[%clock w0:10:04 B0:09:30]}
23x34 {[%clock W0:10:14 b0:09:25]}
11. 39x30 {[%clock w0:10:01 B0:09:34]}
05-10 {[%clock W0:10:11 b0:09:21]}
12. 45-40 {[%clock w0:09:59 B0:09:30]}
18-23 {[%clock W0:10:09 b0:09:19]}
13. 32-28 {[%clock w0:10:09 B0:09:23]}
23x32 {[%clock W0:10:13 b0:09:22]}
14. 37x28 {[%clock w0:10:04 B0:09:32]}
13-18 {[%clock W0:10:14 b0:09:13]}
15. 49-44 {[%clock w0:09:56 B0:09:22]}
18-23 {[%clock W0:10:06 b0:09:08]}
16. 44-39 {[%clock w0:09:52 B0:09:17]}
23x32 {[%clock W0:10:01 b0:09:11]}
17. 38x27 {[%clock w0:09:44 B0:09:21]}
17-21 {[%clock W0:09:53 b0:09:08]}
18. 26x17 {[%clock w0:09:49 B0:09:17]}
12x32 {[%clock W0:09:57 b0:09:06]}
19. 42-38 {[%clock w0:09:48 B0:09:16]}
11-17 {[%clock W0:09:57 b0:09:02]}
20. 38x27 {[%clock w0:09:57 B0:09:06]}
07-12 {[%clock W0:10:02 b0:08:59]}
21. 47-42 {[%clock w0:09:45 B0:09:08]}
17-21 {[%clock W0:09:54 b0:08:56]}
22. 42-38 {[%clock w0:09:42 B0:09:04]}
21x32 {[%clock W0:09:48 b0:09:04]}
23. 38x27 {[%clock w0:09:46 B0:09:08]}
01-07 {[%clock W0:09:55 b0:08:57]}
24. 41-37 {[%clock w0:09:46 B0:09:06]}
07-11 {[%clock W0:09:55 b0:08:51]}
25. 37-32 {[%clock w0:09:43 B0:09:00]}
12-17 {[%clock W0:09:53 b0:08:43]}
26. 43-38 {[%clock w0:09:41 B0:08:54]}
08-13 {[%clock W0:09:50 b0:08:36]}
27. 31-26 {[%clock w0:09:30 B0:08:45]}
17-21 {[%clock W0:09:39 b0:08:30]}
28. 26x17 {[%clock w0:09:34 B0:08:39]}
11x31 {[%clock W0:09:44 b0:08:32]}
29. 36x27 {[%clock w0:09:20 B0:08:40]}
13-18 {[%clock W0:09:30 b0:08:29]}
30. 33-28 {[%clock w0:09:22 B0:08:38]}
19-23 {[%clock W0:09:31 b0:08:27]}
31. 30x19 {[%clock w0:09:21 B0:08:35]}
18-22 {[%clock W0:09:30 b0:08:20]}
32. 27x29 {[%clock w0:09:23 B0:08:28]}
14x45 {[%clock W0:09:32 b0:08:08]}
33. 28-22 {[%clock w0:09:20 B0:08:17]}
06-11 {[%clock W0:09:30 b0:08:04]}
34. 32-28 {[%clock w0:09:20 B0:08:13]}
09-13 {[%clock W0:09:29 b0:08:00]}
35. 39-33 {[%clock w0:09:12 B0:08:09]}
11-17 {[%clock W0:09:21 b0:07:59]}
36. 22x11 {[%clock w0:09:05 B0:08:08]}
16x07 {[%clock W0:09:14 b0:08:01]}
37. 38-32 {[%clock w0:09:05 B0:08:11]}
07-11 {[%clock W0:09:14 b0:07:53]}
38. 32-27 {[%clock w0:08:55 B0:08:02]}
10-14 {[%clock W0:09:04 b0:07:45]}
39. 27-22 {[%clock w0:08:53 B0:07:54]}
11-16 {[%clock W0:09:02 b0:07:39]}
40. 22-18 {[%clock w0:08:28 B0:07:48]}
13x22 {[%clock W0:08:31 b0:07:48]}
41. 28x17 {[%clock w0:08:30 B0:07:52]}
03-08 {[%clock W0:08:39 b0:07:38]}
42. 35-30 {[%clock w0:08:24 B0:07:48]}
04-09 {[%clock W0:08:33 b0:07:35]}
43. 33-28 {[%clock w0:08:23 B0:07:45]}
14-19 {[%clock W0:08:33 b0:07:28]}
44. 30-24 {[%clock w0:08:24 B0:07:37]}
19x30 {[%clock W0:08:32 b0:07:31]}
45. 25x34 {[%clock w0:08:25 B0:07:40]}
15-20 {[%clock W0:08:35 b0:07:24]}
46. 28-22 {[%clock w0:08:29 B0:07:32]}
09-13 {[%clock W0:08:38 b0:07:16]}
47. 48-43 {[%clock w0:08:31 B0:07:25]}
13-19 {[%clock W0:08:41 b0:07:17]}
48. 22-18 {[%clock w0:08:37 B0:07:27]}
19-24 {[%clock W0:08:45 b0:07:21]}
49. 18-12 {[%clock w0:08:33 B0:07:30]}
08-13 {[%clock W0:08:43 b0:07:25]}
50. 12-07 {[%clock w0:08:39 B0:07:34]}
13-18 {[%clock W0:08:48 b0:07:20]}
51. 07-02 {[%clock w0:08:42 B0:07:29]}
18-22 {[%clock W0:08:51 b0:07:25]}
52. 02x30 {[%clock w0:08:43 B0:07:33]}
22x11 {[%clock W0:08:52 b0:07:24]}
53. 30-25 {[%clock w0:08:46 B0:07:33]}
20-24 {[%clock W0:08:55 b0:07:29]}
54. 25-03 {[%clock w0:08:48 B0:07:37]}
16-21 {[%clock W0:08:57 b0:07:27]}
55. 03x26 {[%clock w0:08:53 B0:07:36]}

Joost Buijs
Posts: 471
Joined: Wed May 04, 2016 11:45
Real name: Joost Buijs

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Joost Buijs » Thu Sep 19, 2024 10:37

This format will be fine, if you have these files please compress them with zip or rar and mail them to me.

TIA

Joost

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 12:03

As Joost pointed out, the differences in the top are small, and maybe you need some luck , and most likely when you play the tournament again the outcome could be different.

Below an example, where Damage was lucky.

Image

This is a position from the game Tornado - Damage, move 43 of the game.
Basically this position is a draw, but with the move 38-32 Tornado lost the game, whereas 43. 37-32 would be a draw.

I will post some more examples, hope that the files are correct which i used.

Bert

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 12:32

Even the winner Ares was not without fortune.
See the game against Dream.

Image

Dream (playing with white) had a draw position from around move 45.
The decisive failure was made at move 55 (see diagram).
Playing 32-27 would yield a draw, but with the move 33-28 Dream lost the game in the end.

Bert

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 13:01

A case of bad luck, was seen in the game Dragon - Edeon (as was also found by Ed).
Dragon was clearly winning, but in the diagram position a mistake was made at move 54.

Image

Dragon played 33-29? and the game ended in a draw, whereas 34-29 would be winning.

As Dragon has a 8 piece database and the search is multi-core, it is not impossible that this was a (hidden) bug in the program.

Bert

Krzysztof Grzelak
Posts: 1368
Joined: Thu Jun 20, 2013 17:16
Real name: Krzysztof Grzelak

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Krzysztof Grzelak » Thu Sep 19, 2024 13:08

Bert,
the program Dragon uses from not full 8 figure base of endings.

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 13:22

Krzysztof, thanks for your post, you are right.
But even with a 7 piece database one would expect that Dragon would be able to convert the position to a win.

Bert

Krzysztof Grzelak
Posts: 1368
Joined: Thu Jun 20, 2013 17:16
Real name: Krzysztof Grzelak

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Krzysztof Grzelak » Thu Sep 19, 2024 13:28

I think the only program that surprised me was the Tornado program in the tournament. But I won't write why. Of course I forgot to mention the very good game of the program Moby Dam.

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 13:59

Krzysztof, was your surprise regarding Tornado positive or negative?
I know that Tornado has much potential and the author Frank Mesander is hard working to improve the program further.
Unfortunately the hard work did not pay out yet this time
But I'm sure that Tornado will make huge steps in the near future.

And you are right MobyDam made a huge step (already last year), and it can now be considered a top-program.

Bert

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 15:35

The game MobyDam - Cerberus was more or less a roller-coaster of emotions.
From move 50 it was a draw, but then Cerberus made an mistake and it was a MobyDam win.
But then MobyDam made an error and the position was a draw again.
Finally Cerberus (with black) made the last error in the diagram position.

Image

65. ... 40-45 was played and hereafter MobyDam won, but with 40-44 the game outcome would be a draw, if no more errors would occur after this move. For which I'm not 100% certain.

Bert

BertTuyt
Posts: 1592
Joined: Wed Sep 01, 2004 19:42

Re: The Unofficial World Championship Of Computer Programs 2024

Post by BertTuyt » Thu Sep 19, 2024 15:44

Another interesting game was Maximus - Damage.
See the diagram (white to move, move 67) for a strange ending...

Image

Here Maximus played 1-12 and Damage replied with 30-39.
Maximus kept the king on the diagonal for good reasons and responded with 21-1, and Damage decided to do the same 39-30.
As this continued for several moves, the game ended in a draw.

Here Damage lost one point as the position is a win. But winning difficult endgames without repeating king-moves is also a quality, and here Damage needs to improve.

Bert

Krzysztof Grzelak
Posts: 1368
Joined: Thu Jun 20, 2013 17:16
Real name: Krzysztof Grzelak

Re: The Unofficial World Championship Of Computer Programs 2024

Post by Krzysztof Grzelak » Thu Sep 19, 2024 16:11

BertTuyt wrote:
Thu Sep 19, 2024 15:44
Another interesting game was Maximus - Damage.
See the diagram (white to move, move 67) for a strange ending...

Image

Here Maximus played 1-12 and Damage replied with 30-39.
Maximus kept the king on the diagonal for good reasons and responded with 21-1, and Damage decided to do the same 39-30.
As this continued for several moves, the game ended in a draw.

Here Damage lost one point as the position is a win. But winning difficult endgames without repeating king-moves is also a quality, and here Damage needs to improve.

Bert
I would like to comment on this game - but writing honestly what I think of it. Whether I can.

Post Reply