Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: CCT-3 - An operator report - Der Bringer

Author: James T. Walker

Date: 12:23:29 05/28/01

Go up one level in this thread


On May 28, 2001 at 14:28:14, Peter Berger wrote:

>After reading all the interesting reports by various programmers I thought it
>might be
>interesting for some people to read a Joe User report :)
>
>If memory serves me well in late April Gerrit Reubold , the author of the
>program
>"Der Bringer" ( free download at http://www.reubold.onlinehome.de)  and friend
>of mine
>contacted me and asked if I wanted to operate Bringer in CCT-3 .
>
>CCT-3 was the second "official" tournament for Bringer - it had already
>participated in CCT-1
>some time ago .
>
>CCT-1 preparation had been kind of a chaos - I borrowed a fast computer
>( PIII500 then ) and had everything set up about 5 minutes before the start of
>the first
>game . Opening preparation - none until the very last games . So Bringer happily
>agreed to play a Benoni with black against archrival PostModernist and got
>spanked . Another
>unfortunate game was a draw out of a good position against Grok .
>In the end everything went a little smoother until in the last round Bringer had
>to play
>LambChop - it won a pawn and got a good position - but instead of winning it
>drew :( .
>It finished with 4/8 - good for a new starter , but not too outstanding either .
>
>For CCT-3 everything was starting much more fortunate . Bringer 1.8 had just
>been released - so a well-tested
>and stable version was availlable - and : Bringer has improved quite a lot since
>CCT-1 times .
>
>I planned to buy a new computer anyway so very fast hardware was availlable .
>
>Planned changes for Bringer 1.9 are book tuning and book learning . Book
>learning is not yet availlable ( no problem
>for a tournament ) but book tuning has been made way easier by Gerrit .
>Everything could be done in a comfortable way in
>the Beta's GUI .
>
>So the engine that participated was in fact Bringer 1.8 - the engine is
>identical to the release version so far .
>
>This time there was plenty of time to set everything up , tune the book a little
>and test it . So I was confident that
>this was the meanest Bringer beast the world has seen so far - ready to take any
>challenge :)
>
>The aim was : to get one big scalp this time - a win against one of the big ones
>- (5,5/8 was declared the dream result).
>
>( I personally hoped to get a chance to see a game of Bringer against Ferret,
>too ) .
>
>
>1st Round : Bringer-Chezzz 1-0
>
>A book victory in fact as Chezzz has no opening book yet and Bringer was set up
>for the Queen's Gambit - Chezzz happily
>took the pawn on c4 and tried to keep it ... . Although Chezzz showed some tough
>resistance it was futile .
>
>A good start !
>
>Chezzz' programmer expressed his fears to see this game repeated a few time
>against others but also took it as encouragement
>to work on his book code .
>
>
>2nd Round : Hiarcs8x - Bringer 0-1
>
>Second round and major tournament goal already reached :)). At this time it was
>unknown how strong the new Hiarcs is but as it
>lost only this game in the tournament and finished 3rd it can't be that weak .
>
>Bringer was set up to answer 1.c4 with 1. ..e6 to head for
>Nimzo-Indian/Queens-Indian position types instead of the tough
>English ones and this tactic worked well - Bringer was equal after the opening
>was finished .
>
>After the opening the operator of Hiarcs informed me that Hiarcs was set up for
>all or nothing .
>Dunno to what he referred , but if it was the contempt factor it was definitely
>not the right position to play with IMHO.
>
>[Event "ICS Match (Partie in 45 Minuten + 10 Sekunde(n) pro Zug)"]
>[Site "unbekannt"]
>[Date "2001.05.26"]
>[Round "1"]
>[White "Hiarcs8x(C)"]
>[Black "Bringer1.9ß02"]
>[Result "0-1"]
>[WhiteElo "2200"]
>[BlackElo "2481"]
>
>1. c4  e6  2. Nf3  Nf6  3. d4  b6  4. Nc3  Bb7  5. e3  d5  6. cxd5
>exd5  7. Bb5+  c6  8. Bd3  Be7  9. O-O  O-O  10. b3  Nbd7  11. Bb2
>Bd6  12. Ne2  Re8  13. Ng3  g6  14. Ng5  Qe7  15. Rc1  Nh5  16. Qg4
>Nxg3  17. hxg3  a5  18. Qh4  f6  19. Nf3  a4  20. Rc3  axb3  21. axb3
>Ra2  22. Rc2  Rea8  23. Rfc1  Ba6  24. Bxa6  R8xa6  25. Rb1  c5
>26. Rd2  Qe4  27. Qxe4  dxe4  28. Ne1  f5  29. dxc5  bxc5  30. Rbd1
>Bf8  31. Bc3  Rxd2  32. Rxd2  Nf6  33. Bb2  Kf7  34. Nc2  Rb6
>35. Na1  h5  36. Kf1  Ra6  37. Nc2  Ke6  38. Na1  Ra2  39. Bc3
>Ra8  40. Nc2  Nd5  41. Bb2  Rb8  42. Na1  Nb4  43. Bc3  Ra8  44. Rd1
>Nd5  45. Bb2  Ra2  46. Bh8  Be7  47. Rc1  Nb4  48. Re1  Nd3  49. Re2
>Ra8  50. Bg7  Kf7  51. Bc3  Ra3  { 2.29/17 0:00:47.0 } 52. Kg1
>Bf6  { 2.86/17 0:00:34.1 } 53. Bxf6  Kxf6  { 2.87/18 0:00:34.1 (PB: 0:00:26.7 )
>}
>54. Nc2  Rxb3  { 3.07/17 0:01:45.6 (PB: 0:01:45.5 ) } 55. f3
>Rb1+  { 3.14/16 0:01:17.9 } 56. Kh2  Ne5  { 3.18/17 0:01:04.4 (PB: 0:00:01.5 ) }
>57. Na3  exf3  { 3.19/17 0:00:49.9 (PB: 0:00:01.5 ) } 58. gxf3
>Rb3  { 3.31/16 0:00:36.1 (PB: 0:00:35.9 ) } 59. Nc2  Nxf3+  { 3.33/16 0:00:31.5
>(PB: 0:00:31.4 ) }
>60. Kg2  Ng5  { 3.48/16 0:01:36.1 (PB: 0:00:24.0 ) } 61. Na1
>Rb4  { 3.58/16 0:00:26.9 (PB: 0:00:01.6 ) } 62. Nc2  Rg4  { 3.88/16 0:01:04.7 }
>63. Kh2  h4  { 3.90/16 0:00:58.8 (PB: 0:00:01.5 ) } 64. Rg2  h3  { 3.95/16
>0:00:44.6 (PB: 0:00:01.5 ) }
>65. Rf2  Ra4  { 4.04/15 0:00:27.9 (PB: 0:00:08.1 ) } 66. Ne1
>Ra3  { 4.12/15 0:00:32.0 (PB: 0:00:31.7 ) } 67. Nf3  Ne4  { 3.98/14 0:00:28.9
>(PB: 0:00:28.8 ) }
>
>{Hiarcs8x resigns}0-1
>
>Well done, Gerrit !!
>
>
>3rd Round : Bringer - Deep Fritz 1/2
>
>An easy draw in fact ; game was about equal all the time .
>
>
>4th Round : Deep Shredder - Bringer 1/2
>
>After an equal opening Shredder slowly seemed to make major progress but
>couldn't get enough advantage to win .
>
>So after 4 rounds at the end of the first day Bringer had 3.0/4 after playing
>Hiarcs, Fritz and Shredder - quite a day :) .
>
>A funny experience was how stressing these games were for me :) ; I had to do
>_nothing_ but watch and chat but it felt like a
>day of hard work - in fact it was very similar to having to play for myself .. -
>but it definitely was exciting .
>
>
>5th Round : Bringer - Diep 1/2
>
>This probably was the game to decide whether there will be a good or a
>sensational tournament performance for Bringer.
>
>Both engines commited several serious blunders . Bringer got the better out of
>the opening . In fact Diep was out of book
>at about move 6 . Although Vincent informed me that he used a big book and all
>games were checked by search and at first
>he was quite optimistic as he said Diep had to fill its hashtables and that
>needed time.
>
>To me the time management of Diep looked strange as it used about half of its
>time for the first dozen moves or so .
>
>Bringer went for the dubious plan to play 13. Nb5 and 14.c5 - sacrifying the
>bishop pair for a passer on c6 .
>
>Black probably was already better here , but Bringer had several chances to
>reach a clear draw which it refused as it enjoyed
>its position because of the passer .
>
>Diep slowly advanced until it blundered with the miserable 35. ... Rh1 so it was
>even again .
>
>Then Diep even sacked a pawn because now _it_ was dreaming of a strong passer -
>I think at this stage Bringer had
>some serious winning chances .
>
>But it was Bringer's time to blunder again - it went for bad pawngrabbing and
>came into zugzwang .
>
>Diep went on to win : Bringer was at -2.80 at move 61 and even displayed some
>lines where it went to -5 :( .
>
>But much to my surprise Diep played 72. ...Kh8 ? allowing an easy perpetual .
>
>In the end I was quite happy with the result . Diep is strong - but this game
>didn't show good play by both programs .
>
>It was much fun to talk and analyze with Diep's programmer . Vincent is a very
>strong chessplayer and saw several things
>I didn't notice - and he is very emotional watching the games :)) .
>
>
>6th Round : Crafty - Bringer 1 - 0
>
>Only game with a serious hardware disadvantage against the Quad but the game
>seemed to start well as it went straight
>into an expected line in the Sicilian which is very drawish I think - and
>another draw would have been a nice result ;)
>
>Suddenly Crafty sacked a pawn for great compensation which Bringer grabbed . I
>was very happy to see Bringer
>playing 15. ...Nd5 ! returning the material but after a few pieces were
>exchanged and the dust settled it was Crafty with
>the bishop pair in an open position where it slowly advanced .
>
>I had some hopes after Crafty refused to play f6 and allowed Black to play 22.
>...f6 but after all it was one of those
>positions Crafty simply plays well and my impression was : it just played
>stronger than Bringer here .
>
>With hindsight a sharper opening line might have been a better idea .
>
>
>7th Round : Bringer-LambChop 1-0
>
>So it was Chop again ;) ! I didn't have great expectations for this game after
>the loss against Crafty and knowing how
>strong LambChop plays - also I thought of CCT-1 ..
>
>But much to my surprise LambChop's opening book blundered with 11. ...Nd7 ?? and
>the quite obvious 12. Nb5 ! saw
>Bringer with +1.5 up .
>
>This game was big fun to watch because PeterMcKenzie and Gerrit Reubold had
>interesting discussions about chess
>programming and it was fun to sit there and listen .
>
>On the chessboard it looked like an easy win for a long time .
>
>Although I didn't like 17. dxc5 Bringer seemed to improve its position nicely -
>but then it played 35. b6 ?! closing the
>queen side .
>
>Now it was clear that at some point it had to advance its kingsside pawns to
>make progress . IM Schroer commenting praised
>White and predicted an easy win , the chessprogrammers were more sceptical
>though .
>
>Problem was that Bringer thought it was +2.0 up positionally so why weaken this
>perfect position by "weakening" the kingside
>with pawn moves ?
>
>Fortunately Gerrit told us he had learned his lesson from the Grok game in CCT-1
>and Bringer now recognizes the 50 moves
>rule in search .
>
>It got quite exciting now ; now and then Bringer made a pawn move or what looked
>like a promising King move but the game
>went on and on without any noticeable progress . I don't know if Bringer would
>have won this ( no deeper analysis done so
>far ) but by this time LambChop had started to make some tiny weakening pawn
>moves without being forced to . Those didn't
>seem to hurt too much - until : 122. ...h4 ?? throwing away the game in one
>move.
>
>This game definitely looked like a computer's game .
>
>
>Final Round : SOS - Bringer 1-0
>
>As the LambChop game lasted that long there wasn't any time left to think about
>what to play against SOS so Bringer
>had to deal with the lines chosen against Crafty again . I don't know
>what to think about this game . Opening position was nearly equal, slightly
>worse for Bringer and then it went downhill slowly .
>I haven't analyzed this game so far - it felt like being smashed .
>
>
>So with the much better field of competitors in CCT-3 compaired to CCT-1 Bringer
>finally scored 4,5/8 .
>As it played nearly all the big ones it got a huge SOP/Buchholz and would have
>finished 9th .
>Ahead of Crafty and Shredder ;-) .
>Slightly ahead of GnuChess :)) .
>A shared 9th is still a good result .
>
>
>Bringer beat Hiarcs in a nice game .
>
>My private hope of getting a shot against Ferret wasn't fullfilled .
>
>I think this tournament was a nice success for Gerrit Reubold and showed Bringer
>isn't _that_ far away anymore .
>
>Maybe next time even better :) ??
>
>pete

Thanks for this interesting post Peter,
I have recently become a fan of Der Bringer mostly because of the info given by
the GUI.  I really like to see what the engine is doing during search.  The
engine is getting stronger as the result suggest.  I am interested in getting a
better book for it.  I'm using the medium book now.
Regards,
Jim



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.