WBEC Ridderkerk new results.

[Leo Dijksman]

Geschrieben von:/Posted by: Leo Dijksman at 13 July 2004 18:02:46:

1st Division:
Round 5:
Movei 0.08.247 0111 3.0/4
Dragon 4.5CF 1000 1.0/4
---------------------------------
WildCat 4.0 110½ 2.5/4
Abrok 5.0 001½ 1.5/4
---------------------------------
Zarkov 4.67 1111 4.0/4 !!
Pepito 1.59 0000 0.0/4
---------------------------------
Comet B.68 10 1.0/2 = 2 games to go!
Gothmog 1.0b7 01 1.0/2
---------------------------------

3rd Division:
Round 5:
Scidlet 3.6 0110 2.0/4
Knightx 1.83 1001 2.0/4
---------------------------------
Esc 1.16 0110 2.0/4
Averno 0.70a 1001 2.0/4
---------------------------------
Round 6:
TheCrazyBishop 0052 10½1 2.5/4
Chispa 4.055 01½0 1.5/4
---------------------------------
Cerebro 1.26g 1½ 1.5/2 = 2 games to go!
Spike 0.5 0½ 0.5/2
---------------------------------

5th Division:
Round 9:
Eeyore 1.39 - Tinker 4.55 1-0 73
Plywood 1.73 - Gully 2.15 pl10 1-0 54
Gully 2.15 pl10 - Plywood 1.73 0-1 55
Matheus 2.3 - SseChess 2.04.5 1-0 140
SseChess 2.04.5 - Matheus 2.3 0-1 72
MrChess 2.1 - Deuterium 04.03.06.68 1-0 32
Deuterium 04.03.06.68 - MrChess 2.1 0-1 87
Armageddon 2.251 - Asterisk 0.4b =-= 180
Asterisk 0.4b - Armageddon 2.251 0-1 39
DrunkenMaster 1.1 - Nesik 0.6.4 0-1 29
Nesik 0.6.4 - DrunkenMaster 1.1 0-1 43
--------------------------------------------------------------

6th Division Group B:
Round 1:
Alex 1.60.05a - Marvin 1.1.0 1-0 53
Marvin 1.1.0 - Alex 1.60.05a 1-0 61
Round 2:
Nero 5.3 - Faile 1.4 0-1 52
Faile 1.4 - Nero 5.3 1-0 51
Bace 0.46 - Alex 1.60.05a 0-1 46
Alex 1.60.05a - Bace 0.46 1-0 33
Fencer 2.0 - Mustang 2.67 =-= 32
Mustang 2.67 - Fencer 2.0 1-0 15
WESP 0.8 - Kanguruh 1.87 =-= 111
Kanguruh 1.87 - WESP 0.8 1-0 56
LarsenVB 0.05.01 - ChessRikus 1.4.63 0-1 36
ChessRikus 1.4.63 - LarsenVB 0.05.01 =-= 65
MiniChessAI 1.18 - Cefap 0.7.2 0-1 28
Cefap 0.7.2 - MiniChessAI 1.18 1-0 79
Round 3:
Adam 1.6 - Bace 0.46 1-0 29
Bace 0.46 - Adam 1.6 0-1 33
Faile 1.4 - Marvin 1.1.0 0-1 70
--------------------------------------------------------------

Crosstables and pgn on the WBEC homepage.
Enginelist, latest updates/new engines:
13/07/2004:
Updated: Jan Willem v1.06b
Updated: Alex v1.60.08e
Updated: DanChess v1.06
Updated: Protej v0.4.2
Best wishes,
Leo.



WBEC Ridderkerk homepage.

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 13 July 2004 19:45:42:
Als Antwort auf:/In reply to: WBEC Ridderkerk new results. geschrieben von:/posted by: Leo Dijksman at 13 July 2004 18:02:46:

1st Division:
Round 5:
Movei 0.08.247 0111 3.0/4
Dragon 4.5CF 1000 1.0/4
---------------------------------

Something seems to be too good with the results of Movei
Movei is one of the weakest programs in the world as was proved in WCCC when a very similiar version to Movei0.08.247 was only place 11-12 out of 14
and movei even could not beat an old program of 1997(woodpusher).
It knows almost nothing and it's time management is broken as I learned in WCCC
because I usually do not test with ponder on except checking that the thing does not crash and WCCC was my first experience as operator and I saw it more than once using more than half hour for a single move(equivalent to more than 10 minutes for a single move in WBEC).
How can it be second place in this strong division?
I guess that more games will correct the error but still it seems too good to be correct.
Uri

[Dann Corbit]

Geschrieben von:/Posted by: Dann Corbit at 13 July 2004 20:47:46:
Als Antwort auf:/In reply to: Re: WBEC Ridderkerk new results. geschrieben von:/posted by: Uri Blass at 13 July 2004 19:45:42:

1st Division:
Round 5:
Movei 0.08.247 0111 3.0/4
Dragon 4.5CF 1000 1.0/4
---------------------------------

Something seems to be too good with the results of Movei
Movei is one of the weakest programs in the world as was proved in WCCC when a very similiar version to Movei0.08.247 was only place 11-12 out of 14
and movei even could not beat an old program of 1997(woodpusher).
It knows almost nothing and it's time management is broken as I learned in WCCC
because I usually do not test with ponder on except checking that the thing does not crash and WCCC was my first experience as operator and I saw it more than once using more than half hour for a single move(equivalent to more than 10 minutes for a single move in WBEC).
How can it be second place in this strong division?
I guess that more games will correct the error but still it seems too good to be correct.

Movei is stronger than you think and it belongs in that division.
With a 4x hardware advantage, the already superior programs of the WCCC (by perhaps 150 Elo or more) got a free 120 Elo boost, at least. Shredder is far and away superior to any free program that I have tested, including Ruffian.
Now, if you face someone 270 Elo better than you are, the win expectancy is 17% and that is less than 1/5 of the points.
On even hardware, I would expect Movei to get 1/4 of the points or more even against top programs. And so you could very well see some surprising upsets.
Starting with the model of TSCP [a great thing to start with to write a simple engine but not my choice of model for a world-beater], in a very short time, you have created a program close to Rebel and Aristarch in ability. That is nothing short of amazing. With your latest version, let me remind you again of this result:

Program Elo + - Games Score Av.Op. Draws
1 Ruffian_210 : 2709 112 164 24 79.2 % 2477 25.0 %
2 Crafty-1914 : 2626 122 159 25 68.0 % 2495 16.0 %
3 DeepSjeng : 2614 124 163 25 66.0 % 2499 12.0 %
4 List512 : 2614 124 163 25 66.0 % 2499 12.0 %
5 Delfi : 2582 130 124 25 62.0 % 2497 28.0 %
6 Aristarch 4.50 : 2575 130 124 25 62.0 % 2490 28.0 %
7 Smarthink-017a : 2570 136 122 24 60.4 % 2497 29.2 %
8 Thinker : 2556 139 113 24 58.3 % 2497 33.3 %
9 Moveinow : 2522 143 126 25 54.0 % 2494 20.0 %
10 Rebel12_CB : 2481 109 146 24 45.8 % 2510 33.3 %
11 Yace : 2475 109 146 24 45.8 % 2504 33.3 %
12 Ufim500 : 2439 113 139 24 41.7 % 2497 33.3 %
13 Movei 8.174 : 2437 115 133 25 40.0 % 2508 32.0 %
14 Gothmog_10-b7 : 2426 102 133 25 40.0 % 2497 40.0 %
15 ELChinito 3.25 : 2425 124 130 25 38.0 % 2510 28.0 %
16 Jonny 2.62 : 2403 150 127 25 36.0 % 2503 16.0 %
17 Fruit-1_5-intel: 2374 178 122 25 32.0 % 2505 8.0 %
18 Tcb0050-win32 : 2349 147 121 24 29.2 % 2504 25.0 %
19 Patzer_361 : 2324 143 114 25 26.0 % 2506 28.0 %

Other results I have seen in contest by others are similar (e.g. Leo's games in reconfiguration and the current tournament).
You are much too hard on yourself. Time control is a complex issue. I suggest that you take a look at the code of Pepito for time control. It is simple and yet flexible. It can be improved greatly, but it is a good start.
If you had better time control and no hardware failures, movei would probably have done better.


my ftp site {remove http:// unless you like error messages}

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 13 July 2004 21:52:09:
Als Antwort auf:/In reply to: Re: WBEC Ridderkerk new results. geschrieben von:/posted by: Dann Corbit at 13 July 2004 20:47:46:

1st Division:
Round 5:
Movei 0.08.247 0111 3.0/4
Dragon 4.5CF 1000 1.0/4
---------------------------------

Something seems to be too good with the results of Movei
Movei is one of the weakest programs in the world as was proved in WCCC when a very similiar version to Movei0.08.247 was only place 11-12 out of 14
and movei even could not beat an old program of 1997(woodpusher).
It knows almost nothing and it's time management is broken as I learned in WCCC
because I usually do not test with ponder on except checking that the thing does not crash and WCCC was my first experience as operator and I saw it more than once using more than half hour for a single move(equivalent to more than 10 minutes for a single move in WBEC).
How can it be second place in this strong division?
I guess that more games will correct the error but still it seems too good to be correct.

Movei is stronger than you think and it belongs in that division.
With a 4x hardware advantage, the already superior programs of the WCCC (by perhaps 150 Elo or more) got a free 120 Elo boost, at least. Shredder is far and away superior to any free program that I have tested, including Ruffian.
Now, if you face someone 270 Elo better than you are, the win expectancy is 17% and that is less than 1/5 of the points.

Yes but I expected at least 4 out of 11.
2.5/3 were achieved against weaker programs.
fibchess is an easy win for every player
1.5/2 against Crazybishop and woodpusher is normal(they used the tournament hardware)
I got only 0.5 out of 8 in the remaining games and that is a clear disappointment.
17%*8=1.36 and considering the fact that 4 of the opponents( Sjeng,Jonny,Sos Isichess) used a single processor that is only near 1.5 times faster than the tournament hardware the result is a clear disappointment.
Uri

[Dann Corbit]

Geschrieben von:/Posted by: Dann Corbit at 13 July 2004 22:40:50:
Als Antwort auf:/In reply to: Re: WBEC Ridderkerk new results. geschrieben von:/posted by: Uri Blass at 13 July 2004 21:52:09:

[snip

Yes but I expected at least 4 out of 11.
2.5/3 were achieved against weaker programs.
fibchess is an easy win for every player
1.5/2 against Crazybishop and woodpusher is normal(they used the tournament hardware)
I got only 0.5 out of 8 in the remaining games and that is a clear disappointment.
17%*8=1.36 and considering the fact that 4 of the opponents( Sjeng,Jonny,Sos Isichess) used a single processor that is only near 1.5 times faster than the tournament hardware the result is a clear disappointment.
Uri

Not surprizing with the hardware disadvantage, and the small number of games, together with the hardware problems and the time control bug.
I am super-competitive (more than is good for me) and so I feel your pain. I used to be a pretty good runner long ago (when in the USAF, I had the fastest time in the 1.5 mile yearly fitness race at every base I went to). After I left the USAF, I used to run road races for a while (little local city races, not things like the Boston marathon). I did not usually win them, but I would place in the top 5 most of the time. I always felt like I had a good chance to win. Once I got to the point where I thought I had no chance to win, I no longer had any motivation to run in them.
I can't run a race to complete it. I have to feel that I might win somehow (other than something absurd like all the others spraining their ankles).
You have lots of good programming and chess ideas. If you devote enough energy to it, eventually you can have a real contender for the world championship.
I am guessing that nobody every won the Computer Chess world championship on their first try at chess programming. Probably, they "went back to the drawing board" a few times before ultimate success.
Probably, on your first try at correspondence chess, you did not win a championship either. But eventually, you became a national champion. So you have the determination and concentration to do something great if you put your mind to it.



my ftp site {remove http:// unless you like error messages}

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 13 July 2004 23:26:55:
Als Antwort auf:/In reply to: Re: WBEC Ridderkerk new results. geschrieben von:/posted by: Dann Corbit at 13 July 2004 22:40:50:

[snip

Yes but I expected at least 4 out of 11.
2.5/3 were achieved against weaker programs.
fibchess is an easy win for every player
1.5/2 against Crazybishop and woodpusher is normal(they used the tournament hardware)
I got only 0.5 out of 8 in the remaining games and that is a clear disappointment.
17%*8=1.36 and considering the fact that 4 of the opponents( Sjeng,Jonny,Sos Isichess) used a single processor that is only near 1.5 times faster than the tournament hardware the result is a clear disappointment.
Uri

Not surprizing with the hardware disadvantage, and the small number of games, together with the hardware problems and the time control bug.
I am super-competitive (more than is good for me) and so I feel your pain. I used to be a pretty good runner long ago (when in the USAF, I had the fastest time in the 1.5 mile yearly fitness race at every base I went to). After I left the USAF, I used to run road races for a while (little local city races, not things like the Boston marathon). I did not usually win them, but I would place in the top 5 most of the time. I always felt like I had a good chance to win. Once I got to the point where I thought I had no chance to win, I no longer had any motivation to run in them.
I can't run a race to complete it. I have to feel that I might win somehow (other than something absurd like all the others spraining their ankles).
You have lots of good programming and chess ideas. If you devote enough energy to it, eventually you can have a real contender for the world championship.
I am guessing that nobody every won the Computer Chess world championship on their first try at chess programming. Probably, they "went back to the drawing board" a few times before ultimate success.
Probably, on your first try at correspondence chess, you did not win a championship either. But eventually, you became a national champion. So you have the determination and concentration to do something great if you put your mind to it.

Note only that the time control bug is only with pondering and is relevant also for Leo's tournament(maybe it is less important there because it is 40/40 time control).
I think that I fixed the time control bug(at least most of it) during the tournament before the game against Jonny
There was still one problem against isichess when movei used a long time for move 61(1095.12 seconds) and I guess that it assumed that the remaining time is for a single move and not for all the game at move 61 but it is probably a bug in the manual code and not in the program in general and when I release a new version I will define MANUAL to be 0.

I did 2 changes from the last code that I sent to you before the game against Crafty
1)multiplying the maximal time per move by 3/2 that seems better for pondering
(the original code was used against Crafty and I saw that movei did not used enough time and after fixing the bug time management was ok against Jonny).
2)Correcting a bug that can cause movei to evaluate wrong positions with less than 5 pieces after a null move.
I noticed the bug after movei failed high against Jonny for no reason from a draw score in position when it expected draw by KNN vs K.
I did not work much about the code during the tournament and I also did not work about the code today because I mislocated my leg and prefered not to travel and the computer that I use now has not C installed on it.
Uri