Unofficial World Championship Of Computer Programs In International Draughts 2022

[clp]

Hi Krzysztof,
Can you tell how "No O.(W)" is calculated, I mean the formula?

What the formula means when not all the results are known is as yet shrouded in mystery.

[Joost Buijs]

Hi Krzysztof,
Can you tell how "No O.(W)" is calculated, I mean the formula?

What the formula means when not all the results are known is as yet shrouded in mystery.

Indeed, up until now it looks a bit fuzzy to say the least.

[Krzysztof Grzelak]

Tournament table after the first round of the tournament " The Unofficial World Championship Of Computer Programs In International Draughts 2022".

Krzysztof

[clp]

As we collect more data, the column O.(W) can be better understood. The number in front of the dot indicates the position in the ranking, such that equal players are given the same rank. For a tiebreaker, this number is redundant.
The logic is shown in the table below.

+------------------------+---------------------------+------------------------+
|    12e round           | 18e round                 | 20e round              |
| Pl To No O.(W) .(W)    | Pl   To No O.(W) .(W)     |Pl  To No O.(W) .(W)    |
|  1 18  --       19     | 1    27           19      |1   31           19     |
|  2 17  18.11 =  18 <-- | 2    26           18      |2   28  18.13 =  18 <-- |
|  3 17  18.10    17     | 3    25           17      |3   28  18.09    17     |
|  4 17  18.08    16     | 4    24  16.08 =  16 <--  |4   27  16.11 =  16 <-- |
|  5 17  18.08    15     | 4    24  16.08    15      |5   27  16.09    15     |
|  6 16           14     | 6    24  16.07    14      |6   27  16.07    14     |
|  7 15  13.08 =  13 <-- | 6    24  16.07    13      |7   26  13.18 =  13 <-- |
|  8 15  13.04    12     | 8    23           12      |8   26           12     |
|  9 14           11     | 9    21  11.18 =  11 <--  |9   24           11     |
| 10 13  10.18 =  10 <-- | 10   21  11.17    10      |10  23           10     |
| 11 13  10.18     9     | 11   19           9       |11  21            9     |
| 12 11   8.18 =   8     | 12   18           8       |12  20            8     |
| 13 11   8.18     7     | 13   15            7      |13  16            7     |
| 14 8   -         6     | 14   11            6      |14  12            6     |
| 15 6   -         5     | 15   9             5      |15  9   5.11  =   5 <-- |
| 16 4   -         4     | 16   8             4      |16  9   5.08      4     |
| 17 3   -         3     | 17   4             3      |17  5             3     |
| 18 1             2     | 18   1             2      |18  1             2     |
+------------------------+---------------------------+------------------------+     
The number after the dot follows the same logic, by player per row in the "Cross table according to places" table.

Unofficial World Championship Of Computer Programs In International Draughts 2022

[clp]

After the 20th round, Ares is clearly in the lead.

Note that Dragon and Odeon share 7th and 8th place with 26 points. Odeon has no value in the No O.(W) column. This does not seem right to me.

Code: Select all

Kórnik - Poland 2022-09-01/2022-09-18
Rapid, 100, Round robin, 10 minutes + 10 seconds on 85 moves
Arbiter: Krzysztof Grzelak

Cross table - round 20
Place   No. 
        Program                         1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 N   T  No O.(W) Wins
1   1   NED NED Ares 1.53b NNUE           1 1 1 1 1 1 1 2 3 1 2 2 4 2 2 4 2 20  31           11
2   2   USA USA Kingsrow 1.63           1   1 1 2 1 1 2 1 3 1 1 1 2 2 2 2 4 20  28 18.13      8
3   3   NED NED Damage 16.1 NNUE        1 1   1 1 2 1 1 1 1 2 3 3 2 2 2 2 2 20  28 18.09      8
4   4   FRA FRA Scan 3.1                1 1 1   1 1 2 2 2 1 1 2 1 2 3 2 2 2 20  27 16.11      7
5   5   NED NED Maximus 2.02            1 2 1 1   1 1 1 1 1 3 1 2 1 2 2 2 4 20  27 16.09      7
6   6   NED NED Sjende Blyn             1 1 2 1 1   2 1 1 1 1 1 2 2 4 2 2 2 20  27 16.07      7
7   7   NED NED Dragon Pro 4.6.2        1 1 1 2 1 2   1 1 1 1 2 2 2 2 2 2 2 20  26 13.18      6
8   8   BLR BLR Edeon Sport 1.1.333     1 0 1 2 1 1 1   2 1 1 1 2 2 2 4 2 2 20  26 ??.??      7
9   9   NED NED Moby Dam (May 17 2019)  0 1 1 0 1 1 1 2   1 2 1 2 2 1 2 2 4 20  24            6
10  10  NED NED Truus voor Windows 1.0  1 1 1 1 1 1 1 1 1   1 2 1 2 2 2 2 2 20  23            5
11  11  NED NED Flits 3.02              1 1 0 1 1 1 1 1 2 1   0 1 1 2 3 2 2 20  21            4
12  12  SUI SUI TD King 2020            0 1 1 0 1 1 2 1 1 0 2   1 1 2 1 3 2 20  20            4
13  13  NED NED Tornado 7.11            0 1 1 1 0 0 0 0 0 1 1 1   2 1 2 3 2 20  16            3
14  14  RUS RUS Aurora Borealis         0 0 0 0 1 0 0 0 0 2 1 1 2   0 1 2 2 20  12            2
15  15  FRA FRA Dream 3.X.3             0 0 0 1 0 0 0 0 1 0 0 0 1 2   1 1 2 20  9  5.11       2
16  16  RUS RUS Plus 500 6.54.j.5       0 0 0 0 0 0 0 0 0 0 1 1 0 1 3   1 2 20  9  5.08       2
17  17  NED NED Cerberus 2.3            0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1   1 20  5             0
18  18  NED NED GWD 4.01                0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1   20  1             0
DraughtsArbiter Pro (v.5.44b1) (C) A.Curyło/FMJD  https://fmjd.org/

[BertTuyt]

Ares, is doing quite well this year.

Basically the pattern is the same as last year, there are some very good programs ( 7 - 8, Ares, KIngsrow, Damage, Scan, Maximus, Sjende Blyn, Dragon and Edeon).
These programs mostly draw against each other, and in general win against the somewhat weaker programs.
So in the end the tournament is decided by the man in the middle, Moby Dam, Truus, Flits, TD King and Tornado.

These 5 programs are also quite strong, but on a random basis they tend to make some mistakes, which determines game outcome.
I'm sure that these 5 all could become much better, as the base is already there, if one takes the time to analyses these wrong moves in more detail.

Unfortunately some programmers stopped with their hobby, so Truus and Flits did not change over the last 10-20 years, although they have much potential.

Anyway compliments and credits towards Krzysztof for organizing and taking so much time for this tournament.

Bert

[BertTuyt]

Here an example where i believe a draw was possible.

It is a complex (at least for me) late middle game position from the game Damage - TD King, round 19.
In the below diagram TD-King (black) played 42. 22-27?, whereas 38-43 seems to be a draw.

After 22-27 the engine thinks it has a huge advantage.

Maybe other programs could confirm this?



Bert

[Joost Buijs]

Here an example where i believe a draw was possible.

It is a complex (at least for me) late middle game position from the game Damage - TD King, round 19.
In the below diagram TD-King (black) played 42. 22-27?, whereas 38-43 seems to be a draw.

After 22-27 the engine thinks it has a huge advantage.

Maybe other programs could confirm this?



Bert

Although white has a clear positional advantage Ares with 6P EGDB sees after 13 seconds at depth 28/44 that the position after 38-43 is a draw. My guess is that TD-King didn't look deep enough to see that 38-43 draws the game.

The fun part is that without EGDB Ares sees the draw after 38-43 in 7 seconds at depth 27/45, so I wonder how much of an improvement the EGDB really gives.

[edit]
Sorry, I have to correct the above statement, it was a second run with EGDB enabled, my mistake, it doesn't see the draw without EGDB, but it plays the right move of course.

After 22-27 with 6P EGDB:

Code: Select all

depth 20/33  score   136  time   0.26  nodes   36058659  Mn/s 140.7 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 21/36  score   140  time   0.35  nodes   49820890  Mn/s 142.8 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 22/36  score   139  time   0.52  nodes   74255000  Mn/s 143.5 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 23/40  score   161  time   1.62  nodes  237214280  Mn/s 146.3 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 24/46  score   161  time   2.21  nodes  320929579  Mn/s 145.4 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 25/44  score   210  time   4.56  nodes  666984657  Mn/s 146.1 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 26/45  score   210  time   5.99  nodes  867736657  Mn/s 144.7 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 27/47  score   211  time   8.47  nodes 1222815312  Mn/s 144.3 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 28/47  score   210  time  13.16  nodes 1901193477  Mn/s 144.4 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 29/50  score   212  time  20.53  nodes 2963563115  Mn/s 144.3 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36
depth 30/50  score   221  time  32.64  nodes 4733320352  Mn/s 144.9 PV: 11-06 27-32 06-01 32x41 01-23 41-47 23x49 47-36

After 22-27 without EGDB:

Code: Select all

depth 20/35  score   123  time   0.27  nodes   38767103  Mn/s 146.1 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 21/35  score   123  time   0.35  nodes   50817465  Mn/s 147.0 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 22/36  score   125  time   0.47  nodes   68829002  Mn/s 147.8 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 23/39  score   123  time   0.69  nodes  102595891  Mn/s 149.7 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 24/39  score   125  time   1.00  nodes  149893592  Mn/s 149.5 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 25/41  score   124  time   1.99  nodes  299296831  Mn/s 150.7 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 26/44  score   145  time   4.59  nodes  695799012  Mn/s 151.5 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 27/45  score   143  time   6.24  nodes  941597781  Mn/s 150.9 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 28/45  score   150  time   9.21  nodes 1380761198  Mn/s 149.8 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 29/50  score   149  time  13.97  nodes 2084852056  Mn/s 149.2 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44
depth 30/47  score   149  time  25.13  nodes 3717589809  Mn/s 147.9 PV: 11-06 27-32 06-01 32x41 01-23 28-33 23x46 33x44

With EGDB it sees a somewhat higher score, but nothing conclusive.

[Krzysztof Grzelak]

I have to apologize to everyone watching the tournament. Unfortunately something happened in the 22nd round - and the games and the results of the games are gone. He tried to restore all the games and see what the cause of this fault was.

Krzysztof

[Krzysztof Grzelak]

Everything is back to normal. Tomorrow is another day of the tournament.

[Joost Buijs]

I have to apologize to everyone watching the tournament. Unfortunately something happened in the 22nd round - and the games and the results of the games are gone. He tried to restore all the games and see what the cause of this fault was.

Krzysztof

In a previous round (I don't remember which one) I saw something strange happening too, there was a game going on, and the next game (that still had to be played) showed the same moves. I don't know what kind of software is used to broadcast these games, but there seem to be some irregularities in it.

The clocks of the applet show funny behavior too, I understand that after each move 10 seconds operator time is added, in between moves the clocks often jump up and down, sometimes by more than 20 seconds, this is not easy to explain. I assume a time stamp is send with each update of the applet, of course there will be transmission delays, but delays of tens of seconds seem a bit awkward.

[Krzysztof Grzelak]

In a previous round (I don't remember which one) I saw something strange happening too, there was a game going on, and the next game (that still had to be played) showed the same moves. I don't know what kind of software is used to broadcast these games, but there seem to be some irregularities in it.

The clocks of the applet show funny behavior too, I understand that after each move 10 seconds operator time is added, in between moves the clocks often jump up and down, sometimes by more than 20 seconds, this is not easy to explain. I assume a time stamp is send with each update of the applet, of course there will be transmission delays, but delays of tens of seconds seem a bit awkward.

It is difficult for me to relate to this. As for the clock, it is set equally to all programs playing in the tournament.

[Joost Buijs]

In a previous round (I don't remember which one) I saw something strange happening too, there was a game going on, and the next game (that still had to be played) showed the same moves. I don't know what kind of software is used to broadcast these games, but there seem to be some irregularities in it.

The clocks of the applet show funny behavior too, I understand that after each move 10 seconds operator time is added, in between moves the clocks often jump up and down, sometimes by more than 20 seconds, this is not easy to explain. I assume a time stamp is send with each update of the applet, of course there will be transmission delays, but delays of tens of seconds seem a bit awkward.

It is difficult for me to relate to this. As for the clock, it is set equally to all programs playing in the tournament.

It's not important, I understand that it is difficult to grasp what it all does. I assume you use a DGT timer that is transmitting the used time or the remaining time of the players to the website that runs the tournament software.

Anyway thanks for organizing this tournament, it is a wonderful yob you are doing, kudos!

[Krzysztof Grzelak]

It's not important, I understand that it is difficult to grasp what it all does. I assume you use a DGT timer that is transmitting the used time or the remaining time of the players to the website that runs the tournament software.

Anyway thanks for organizing this tournament, it is a wonderful yob you are doing, kudos!

I use these applications as in this movie https://www.youtube.com/watch?v=s-Vx-uFNddk

[Joost Buijs]

It's not important, I understand that it is difficult to grasp what it all does. I assume you use a DGT timer that is transmitting the used time or the remaining time of the players to the website that runs the tournament software.

Anyway thanks for organizing this tournament, it is a wonderful yob you are doing, kudos!

I use these applications as in this movie https://www.youtube.com/watch?v=s-Vx-uFNddk

So you are using a DGT electronic draughts board, I didn't know that something like this exists. In the past I had a DGT electronic chess board, I assume the capabilities of the draughts board are about the same.