[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [computer-go] Acornsoft 1984
I see that you are missing a lot of information on many
tournaments. Attached are some of the tournament
reports from my archives. This should fill in most of your blanks. In
1990, Many Faces of Go was 4th, but I have
no other information.
I have sgf files for many of the ING competitions if there are any you are
missing.
I also have a collection of old programs that could be made available with
permission of their authors.
David Fotland
At 11:43 AM 2/25/2004 +0000, you wrote:
I have published at
http://www.intelligentgo.org/en/computer-go/events/acornsoft/index.html
an account of the first-ever computer Go Tournament, held in London in
1984. This includes all the game records; previously, I think only the
game between the finalists had been published.
I have transcribed all the games from the original kifus. This was harder
than might be expected, for 13x13 games; one of the games had over 200 moves.
Nick
--
Nick Wedd nick@xxxxxxxxxxxxxxxxx
_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://computer-go.org/mailman/listinfo/computer-go
The 1992 US Computer Go Championship - by David Fotland
This year the US Computer Go Championship was held August 2, 3, and 4
at Williamette University in Salem Oregon. This was the 5th Computer Go
Championship held in the USA at the Go congress. Previous winners were:
1988 - Cosmos (Many Faces of Go)
1989 - Go Intellect
1990 - Go Intellect (actually won by Goliath, a program from Europe)
1991 - Many Faces of Go
This year's results were:
1) Many Faces of Go, by David Fotland
2) Go Intellect, by Ken Chen
3) Nemesis, by Bruce Wilcox,
4) Contender, By Lynn Beus
5) Poka, by Howard Landman
6) Prototype, by Arthur McGrath
7) Fumiko, by Peter Dudey
With 7 programs entered there was time to play a full round robin with one
game between each pair of programs. Time limits were 1 hour to play 125
moves for each program, with the game adjudicated after move 250 if the
programs were too slow or the games were too long. Ing's SST Laws of Wei Chi
were used since these are the rules used by the World Computer Go Championship.
None of the games had to be adjudicated.
Five programs have participated in this tournament
several times before: Many Faces of Go, Go Intellect, Nemesis,
Contender, and Poka. Go Intellect has several second place finishes
in world competition and was the favorite to win this year. Many Faces of
Go was based on the new second release that had been completed the week before,
and is about 3 stones stronger than last year's Many Faces of Go. Nemesis
was part way through a major overhaul to allow him to add full board
lookahead, so Bruce was not expecting to do very well.
Poka was pretty much the same program as last year.
There were two new programs this year, Fumiko and Prototype. Although
neither of them beat any of the established programs, both of their
authors were inspired by the competition, and plan to do much better
next year.
Ken chose white for Go Intellect whenever he could because the tournament was
played by the Ing SST Laws of Wei Chi with an 8 point komi. His
opponents all chose black when they had the choice of colors, so Go
Intellect played white in every round.
I knew that Go intellect would be my toughest competition this year. Many
Faces beat Go Intellect last year only because Go Intellect got into time
trouble and stopped reading and dropped two big groups in the endgame.
Before that it was 100 points ahead. Last year's Many Faces was
definately weaker than Go Intellect. I didn't expect to have any problem
beating Nemesis, Contender, Poka, or the new programs.
I had hoped to have some time to tune Many Faces against Go Intellect
before the competition, but the second release for Ishi Press was a little
late. I delivered the final Many Faces to Ishi on July 27, after 3 months
of intense bug fixing, and spent the rest of the week relaxing in front of
the TV watching the Olympics. On Friday I played a test game between
Many Faces and last year's Go Intellect, and Many Faces lost by more than 30
points. I fixed problems in this game from Friday night right up until
the game against Go Intellect on Monday afternoon.
Unfortunately, one of the changes I made before the round against Contender
was incorrect and caused the program to crash, so Many Faces lost to
Contender. Programs can be restarted after a crash, with the clock
still running, but cannot be modified during a round. Many Faces survived
two restarts, but the third time it crashed for good. I fixed the bug after
the round and played Go Intellect the next round.
Go Intellect and Many Faces were the only two programs that do full board or
life and death reading, and they crushed all other opponents. Many Faces
beat Nemesis by over 140 points. The game between Go Intellect and
Many Faces was very close. Go Intellect was ahead by about 3 points in the
late endgame when it tried to live inside a large secure
territory. It gave up a lot of points and lost by 11. Go Intellect and
Many Faces ended up tied with 5 wins each, and tied on sum of opponent's
scores. Since Go intellect lost to Many Faces and Many Faces lost to
Contender, Many Faces won on sum of defeated opponent scores.
Many Faces and Nemesis were the only programs to implement the Standard
Go Modem protocol, so only one game was played over a cable. All other
games required the authors to act as operators. (A copy of the Standard
Go Protocol specification, and sample code from Many Faces of Go and Nemesis,
is available from either Bruce Wilcox or myself).
Howard was also tuning up Poka during the tournament between rounds.
In the first round Fumiko had an incorrect position on the board and
couldn't correct it, and so forfeited. They continued playing, and Fumiko
suicided. Poka wouldn't accept it, but had already won. Howard fixed
Poka to accept suicide, but unfortunately did it wrong, and later lost to
Nemesis when it took both the capturing and captured stones off the
board in a ko fight.
Contender looked quite a bit better this year. The game between Contender
and Many Faces looked close before Many Faces crashed. After I fixed the
bug we completed the game and Many Faces killed a big group and won. Contender
got way ahead of Poka, but gave back a big dead group in the endgame and
lost. Poka would have finished above Contender if Many Faces hadn't
crashed.
Prototype plays a large moyo game, and looked very good against Many Faces
in the opening, but let most of its stones get captured in the middle game.
It also made a huge moyo against Nemesis, but defended the boundary incorrectly
and let Nemesis break through the wall. The game between Poka and Prototype
was interesting since both programs make big moyos and both are weak at
tactics and life and death. It came down to a big group of Poka's which had
a single 3 space eye. Prototype didn't kill it, and Poka didn't make it
live since Poka thinks groups only need one eye to live. Finally after all
dame were filled, Poka played to make two eyes as a safety play (since in
Chinese rules it doesn't cost to play inside your own territory) and won.
Fumiko was written during the summer and was a very young program. It crashed
or ran out of time in its first four games, and forfeited the last two
without playing. It was running on my portable 386 SX, which was not
fast enough to allow it to finish its games.
I borrowed an HP9000/720 "Snake" workstation to run Many Faces, but ended
up not using it. I couldn't duplicate the programming environment from my
PC onto it, and since I was making changes up to the last minute I traded
machines with Howard. He used my Snake, and I used his 40 Mhz 486. I played
all tournament games at level 15. At one point Howard played two games
simultaneously on the Snake.
1 2 3 4 5 6 7 Wins
1 Go Intellect X w-loss w-win w-win w-win w-win w-win 5
2 Many Faces b-win X b-win w-crash b-win w-win win 5
3 Nemesis b-loss w-loss X w-win b-win w-win w-win 4
4 Contender b-loss b-win b-loss X w-loss w-win win 3
5 Poka b-loss w-loss w-crash b-win X w-win b-win 3
6 Prototype b-loss b-loss b-loss b-loss b-loss X b-win 1
7 Fumiko b-loss forfeit b-loss forfeit w-loss w-loss X 0
Game between go Intellect and Many Faces of Go (Round 3):
Black 17: This is one of the last minute changes I made. The prior
version prefers to extend toward the edge rather than the center. Go
Intellect frequently plays this diagonal attachment.
Black 29: There was a bug in the pattern database. MFGO did not expect
28 in answer to 27, and does not know how to respond.
Black 33: Since the black stone at 7 is not captured, MFGO thinks the
white stones are weak and tries to surround them.
Black 69: The urgent extension to prevent a hane at the head of two
stones is missing from the pattern database.
Black 71: Making sure of two eyes, and 75, making sure of the
connection are both too conservative. Black should have made some
territory in the lower center.
Black 97,99: Finally sealing off the top. If white had jumped in first
it would have had the lead. Black is quite a bit ahead at this point.
Black 111: MFGO picks this way to live since it is sente.
Black 125: To prevent the connection, but not a very good way to do it.
Black 133: Since the white group is undercut at the top and black can
push in from the center, MFGO thinks it can kill it with this atari by
reducing it to one eye. Wrong of course. After white 136 it realizes
that white is alive and plays elsewhere.
Black 149: A major pattern bug. This urgent shape pattern neglects to
check if the black group below can be captured. Black was ahead, but
now the game is close when white captures 3 stones.
White 220: At this point White is a few points ahead, but launches an
invasion of the big territory at the top. Black picks up over 5 points
during this invasion.
Black 233: Hits the key point to prevent white from living.
Black 259: The program picks a strange way to connect. This gives
white a chance to make a big ko. White plays to set up the ko, then
at 264 fills the ko rather than starting it. After black defends at 265
the game is over.
Although the game was close almost all the way, both programs missed
big chances and both made many slack moves. The state of the art in
American computer go demonstrated here still has lots of room for
improvement.
EVENT 1992 US GO Congress Computer Go Championship, Round 3.
User
MFGO level 15
Enduser
Boardsize 19
Handicap 0
Rules Chinese
Black Many Faces of Go, Release 2, Rev 8.0, level 15
White Go Intellect
COM
Game played: 8/3/1992 at 16:11
ENDCOM
B 1 D17
W 2 R16
B 3 C4
W 4 Q3
B 5 P17
W 6 C15
B 7 C13
W 8 E3
B 9 C16
W 10 R5
B 11 D6
W 12 D15
B 13 F16
W 14 C2
B 15 E14
W 16 Q17
B 17 P16
W 18 Q18
B 19 L16
W 20 C9
B 21 K3
W 22 H3
B 23 N3
W 24 C11
B 25 C7
W 26 D13
B 27 B2
W 28 B3
B 29 P18
W 30 R10
B 31 F7
W 32 C12
B 33 E11
W 34 B13
B 35 B15
W 36 E13
B 37 F13
W 38 B14
B 39 J2
W 40 F12
B 41 G13
W 42 E12
B 43 C3
W 44 A2
B 45 O5
W 46 D2
B 47 R8
W 48 R13
B 49 P8
W 50 B4
B 51 C5
W 52 B16
B 53 B17
W 54 A15
PRISONER B15
B 55 E9
W 56 D8
B 57 G9
W 58 P10
B 59 P14
W 60 P13
B 61 O14
W 62 O13
B 63 N8
W 64 H2
B 65 F5
W 66 A17
B 67 B18
W 68 G12
B 69 B8
W 70 H5
B 71 B6
W 72 B9
B 73 S9
W 74 S10
B 75 M5
W 76 J6
B 77 P2
W 78 Q2
B 79 P3
W 80 Q4
B 81 A9
W 82 J12
B 83 H12
W 84 H11
B 85 H13
W 86 G11
B 87 Q1
W 88 R1
B 89 P1
W 90 R2
B 91 Q19
W 92 K12
B 93 M13
W 94 M12
B 95 N14
W 96 J9
B 97 L13
W 98 L12
B 99 K14
W 100 A10
B 101 S19
W 102 E10
B 103 F10
W 104 D10
B 105 Q6
W 106 F11
B 107 R6
W 108 A8
PRISONER A9
B 109 A7
W 110 A9
B 111 E4
W 112 F3
B 113 E15
W 114 M9
B 115 M8
W 116 D16
B 117 O10
W 118 O11
B 119 N10
W 120 C17
PRISONER C16
B 121 D18
W 122 S6
B 123 N12
W 124 S7
B 125 S8
W 126 N13
B 127 N11
W 128 Q14
B 129 T10
W 130 T11
B 131 O12
W 132 P11
B 133 T12
W 134 S11
B 135 Q15
W 136 R15
B 137 L9
W 138 L8
B 139 M10
W 140 P12
B 141 K8
W 142 J8
B 143 L7
PRISONER L8
W 144 T9
PRISONER T10
B 145 H1
W 146 G1
B 147 J1
W 148 J3
B 149 K4
W 150 K2
B 151 L2
W 152 K1
PRISONER H1
PRISONER J1
PRISONER J2
B 153 L1
W 154 T8
B 155 K13
W 156 R7
B 157 Q7
W 158 R9
B 159 Q8
W 160 J13
B 161 J14
W 162 P4
B 163 O4
W 164 E8
B 165 F8
W 166 C8
B 167 B7
W 168 E7
B 169 E6
W 170 G6
B 171 F6
W 172 G5
B 173 R18
W 174 S17
B 175 J4
W 176 C18
B 177 C19
W 178 C16
B 179 Q5
W 180 R4
B 181 E16
W 182 A18
B 183 B19
W 184 P5
B 185 O6
W 186 Q16
B 187 H7
W 188 G7
B 189 J7
W 190 P15
PRISONER Q15
B 191 O15
W 192 K6
B 193 H8
W 194 K9
B 195 K10
W 196 J10
B 197 K7
W 198 L6
B 199 M6
W 200 K11
B 201 L10
W 202 G10
B 203 K5
W 204 L5
B 205 M4
W 206 F9
PRISONER F10
B 207 J1
W 208 J2
B 209 F10
PRISONER F9
W 210 S18
B 211 F9
W 212 H4
B 213 H9
W 214 D14
B 215 P6
W 216 R17
B 217 R19
W 218 B5
B 219 A5
W 220 E19
B 221 E18
W 222 G18
B 223 K18
W 224 F18
B 225 F4
W 226 H17
B 227 G4
W 228 F17
B 229 E17
W 230 G16
B 231 H15
W 232 K17
B 233 J18
W 234 J17
B 235 L17
W 236 G15
B 237 F14
W 238 G14
B 239 H14
W 240 H18
B 241 H16
W 242 J19
B 243 L19
W 244 F19
B 245 D19
W 246 M18
B 247 L18
W 248 T19
B 249 T18
PRISONER T19
W 250 T17
B 251 F15
W 252 T19
PRISONER T18
B 253 P19
W 254 P9
B 255 G3
W 256 G2
B 257 O9
W 258 N17
B 259 N18
W 260 O18
B 261 N19
W 262 O16
B 263 M19
W 264 Q15
B 265 N16
W 266 M14
B 267 O17
PRISONER O16
W 268 D3
B 269 A4
W 270 B1
PRISONER B2
B 271 Q9
W 272 A3
B 273 A6
W 274 D7
B 275 C6
W 276 L11
B 277 Q10
W 278 L4
B 279 L3
W 280 D4
B 281 Q11
W 282 R12
B 283 Q12
W 284 Q13
B 285 M11
W 286 J11
B 287 R11
W 288 S12
B 289 T18
PRISONER T19
W 290 H6
B 291 D5
W 292 T19
PRISONER T18
B 293 A19
W 294 K16
B 295 A16
PRISONER A17
PRISONER A18
W 296 A17
PRISONER A16
B 297 T18
PRISONER T19
W 298 L15
B 299 K15
W 300 T19
PRISONER T18
B 301 T7
W 302 T6
PRISONER T7
B 303 T18
PRISONER T19
W 304 D9
B 305 T19
W 306 G8
B 307 H10
W 308 J5
B 309 A18
W 310 H19
B 311 A16
PRISONER A17
W 312 S2
B 313 A17
W 314 J16
B 315 M15
W 316 G17
B 317 K19
W 318 M16
B 319 L14
PRISONER L15
PRISONER M14
W 320 H1
PRISONER J1
B 321 M17
PRISONER M18
PRISONER M16
PRISONER N17
W 322 C14
PRISONER C13
B 323 J15
W 324 T13
PRISONER T12
B 325 G19
PRISONER J19
PRISONER H19
PRISONER E19
PRISONER F19
PRISONER G18
PRISONER F18
PRISONER F17
PRISONER H18
PRISONER H17
PRISONER J17
PRISONER K17
PRISONER K16
PRISONER J16
PRISONER G17
PRISONER G16
PRISONER G15
PRISONER G14
W 326 D11
PRISONER E11
B 327 O19
PRISONER O18
W 328 pass
B 329 N9
PRISONER M9
W 330 pass
B 331 pass
USER
MFGO PATTERNS 2860
ENDUSER
Summarized from July 1984 AGA Go Journal report by Bruce Wilcox
June 16, 1984 in Salt Lake City, at the Usenix conference.
All programs round robin, eliminate one program, round
robin again with colors reversed. 1 hour CPU time
each, then 10 second per move. Games were often over
400 moves.
Jim was far slower than others
ogo mostly played mirror go
Nemesis crashed once, otherwise won big
4-1 Nemesis Bruce Wilcox
3-2 Goanna Bruce Ellis
2-3 Ogo Peter Langston
0-3 Jim Hank Dietz
1985 ICGC, only 9x9 tournament:
Programmer Program Name State or region
First place: Won Roshee Friday? Taiwan lose one game on time, won rest
(Wang Loh-Tsi)
Second place: Tsao Kwo-Ming ? Taiwan
3rd place: Allen Scarff Microgo UK
8 total participants. Nemesis by Bruce Wilcox participated, lost only to Friday, and
a taiwanese program that suicided. 6 programs from Taiwan.
$2500 first prize,
$1500 second
$500 third
Return-Path: <hpda!ucbvax!bellcore!psl>
Received: from hpda.UUCP by hpihoa.UUCP (4.30/3.4)
id AA04156; Mon, 15 Jul 85 22:28:46 pdt
Received: by hpda.UUCP (4.30/3.4)
id AA12844; Mon, 15 Jul 85 22:26:35 pdt
From: hpda!ucbvax!bellcore!psl
Received: by UCB-VAX.ARPA (4.24/5.2)
id AA00398; Mon, 15 Jul 85 22:18:44 pdt
Date: Mon, 15 Jul 85 22:18:44 pdt
Message-Id: <8507160518.AA00398@xxxxxxxxxxxxxxxxx>
To: ucbvax:hpda!hpihoa!fotland
Subject: Re: Usenix go tournament
Nope, nemesis didn't win; oog did.
Here's a copy of the article that will appear in login.
.po 1.2i
.sz +2
.sz +4
.b
.ce
Second Annual Usenix Go Tournament
.sz -4
.r
.pp
The second Usenix Computer Go Tournament was held June 12 at the Summer 1985
Usenix meeting in Portland, Oregon.
This is the second full-board computer go competition ever to be held
(the first being last year's tournament).
There were five programs entered:
\fIgoanna\fR, \fIgorilla\fR, \fInemesis\fR, \fIogo\fR, and \fIoog\fR.
Each entrant played every other entrant twice,
once with white and once with black.
.sp 0.1i
.pp
\fIGoanna\fR and \fIgorilla\fR were written by Bruce Ellis.
They employed a pattern-matching strategy that required an astonishingly
small amount of cpu time;
the game between these two programs was true speed go.
\fIGorilla\fR was designed to play a more cutthroat game in which
killing opposition stones was a paramount goal.
.pp
\fINemesis\fR was written by Bruce Wilcox and is available commercially
for a variety of microcomputers.
It entered the tournament as the reigning champion and clear favorite,
having demolished the opposition last year and having played well
in a recent tournament with humans.
Although judged to be the strongest entrant, it seemed to
have trouble with the tense tournament atmosphere and
crashed frequently (in 6 out of 8 games), thereby threatening its
domination of the standings.
.pp
\fIOgo\fR was written by Peter Langston.
It plays a peculiarly thoughtless strategy in which it
mirrors its opponent's moves whenever possible.
This is not unheard of, even in professional play.
Unfortunately, some last minute "enhancements" caused \fIogo\fR to
commit suicide with rhythmic regularity.
\fIOog\fR was also written by Mr. Langston (who happened to be the
tournament organizer); it actually tried to play go (a novel approach
for that author).
.sp 0.1i
.pp
It was far more difficult to judge the results in this tournament
than in an ordinary tournament.
Should a loss due to a blatantly illegal move be comparable to a loss from
falling into a complete catatonic stupor and forgetting to move at all?
How should either compare to uttering "out of buffers" before discretely
resigning?
The judges finally decided to pay most attention to genuine go competence.
The results were then:
.TS
center;
l l c c c.
Finish Entrant Won Lost Other
1 \fIoog\fR 5 2 1
2 \fInemesis\fR 2 0 6
3 \fIgoanna\fR 0 2 6
4 \fIgorilla\fR 0 2 6
5 \fIogo\fR 0 1 7
.TE
Here ``Other'' refers to a variety of unfortunate outcomes
which were scored by a zany scheme concocted by the judges to reflect
their notions of how each outcome reflected go skill or
lack thereof.
.pp
Thus \fIoog\fR is the champion, coming from complete obscurity to snatch
the title from the favorite.
Both of \fIoog\fR's losses were to \fInemesis\fR; one of these games actually
had a strong resemblance to a go game.
The other genuinely interesting game was a practice game between \fInemesis\fR
and itself;
it was a close contest but finally black gave up the ghost and fell into
a silent torpor.
.sp 0.1i
.pp
The game of go is difficult to play well.
A sense of broad strategical issues is very important; apparently none of
the programs here used the familiar tree search techniques so popular
for handling tactical situations.
It seems that \fINemesis\fR plays at about the level of 20 kyu judging both
from its performance here and its performance in an earlier tournament
in which its opponents were human.
This is substantially above the level of a beginning player but still very,
very far away from being able to beat players drastically weaker than the
program's author.
Perhaps next year's tournament will bring forth new and stronger programs
to astound the Usenix go community,
but the champ doesn't seem worried, "Hah! Oog'll cream those wimps!"
.sz -2
Return-Path: ucbvax!research!rob
Received: from hpda.HP.COM by hpihoc.HP.COM; Wed, 29 Jun 88 12:42:38 pdt
Received: by hpda.HP.COM; Wed, 29 Jun 88 12:41:34 pdt
From: ucbvax!research!rob
Received: from ucbvax with uucp; Wed, 29 Jun 88 09:26:44
Received: by ucbvax.Berkeley.EDU (5.59/1.28)
id AA15681; Wed, 29 Jun 88 07:51:38 PDT
Date: Wed, 29 Jun 88 07:51:38 PDT
Message-Id: <8806291451.AA15681@xxxxxxxxxxxxxxxxx>
Apparently-To: hpda!fotland
the games were perfectly regular:
g2.25 won all its games
g2.5 was defeated only by g2.25, and so on for
goo
goo with symmetry (goog)
goanna
i managed to forget that you'd given me the tape for posterity
and used it to bring the results back. so your program
is safely unavailable. (p.s. could you send me a tape of
that version please? sorry to be so forgetful.)
i destroyed the pc floppy.
there was a japanese amateur shodan at the tournament who played
on the pc. he said your program was between 9 and 12 kyu.
he played both with the parameter set to 5 and to 50, and said
its level of play didn't change much. interpret that as you will.
i've asked usenix to dig deep into its pocket and buy a real trophy.
i figure the first time the best program wins deserves special
commemoration.
-rob
1986 ICGC, 9x9 tournament:
1st place: Du Kwei-Chung ? Taiwan
2nd place: Liu Dong-Yue Dragon Taiwan
3rd place: Tsao Kwo-Ming ? Taiwan
1986 ICGC, 19x19 tournament:
1st place: Du Kwei-Chung ? Taiwan
2nd place: Liu Dong-Yue Dragon Taiwan
3rd place: Bruce Wilcox Nemesis USA
Summary of US Computer Go Championships:
1988 - Berkeley CA
1989 - Rutgters, NJ
1990 - Denver CO
1991 - Rochester, NY
1992 - Willamette, OR
1993 - Holyoak, MA
1994 - Washington DC
1995 - Seattle WA
1996 - Cleveland, OH
1997 - Lancaster, PA
88 89 90 91 92 93 94 95 96 97
David Fotland Many Faces of Go 1 2 3 1 1 4 2 1 1 1
Ken Chen Go Intellect 1 1 2 2 2 1
Martin Mueller Explorer 2 2
Kao Stone 3 1
Bruce Wilcox Ricigo/Ego 3-4
Bruce Wilcox Nemesis 2 3 2 5 3 6
Lynn Beus Contender 6 4 4 3-4
Art McGrath Prototype 6 3
Howard Landman Poka 4 5 5 3 3
Jim Gunderson Petronius 7 4
Peter Dudey Fumiko 7
Herb Enderton Golem 5
Ken Schatten Go Guru 7
B. Ladendorf Tsunami 8
Gary Boos Teamgo 5 4 2
10 programs, 9 round round robin, full report soon
Final results:
1) Handtalk 8-1
2) Go4++ 8-1
3) Go Intellect 7-2
4) Silver Igo 7-2
5) Many Faces 5-4
6) Modgo 4-5
7) Fungo 3-6
8) Star of Poland 2-7
9) Explorer 1-8
10) Super Ego 0-9
Super Ego is Bruce Wilcox' new program.
Silver Igo is Japanese
Fungo is Korean
RESULTS OF THE 1996 ING WORLD COMPUTER GO CHAMPIONSHIP.
The 1996 World Computer Go Congress was held on Nov 16 and 17 at Zhongshan
University, in Guangzhou, China. The Ing foundation provided prize funds and
money to run the competition, and Chen Zhi Xing, author of handtalk, organized
it. We had a 6 round swiss tournament, with the top 4 programs from last
year seeded in a 12 program field. The tournament was well organized,
with 133 Mhz Pentium computers provided. There were no computer problems,
and the rounds started on time. Many programs implemented the computer
go modem protocol, which helped make the rounds go faster.
The final result was:
1st, 6 wins, Handtalk, by Chen Zhi Xing
2nd, 5 wins, Go Intellect, by Ken Chen
3rd, 4 wins, Stone, by Kao Kuo Yuan
4th, 3 wins, Go4++, by Michael Reiss
5th, 3 wins, Many Faces of Go, by David Fotland
6th, 3 wins, Go Master, by Jee Wonho
7th, 3 wins, Jimmy, by Yan Shi-jim
8th, 3 wins, Wulu, by Chen Guo Bao, Lei Xue Ru, and Fan Jin Chao
9th, 2 wins, Sason, by Ha Youngkyun
10th, 2 wins, DFGO, by Chen Tie Jun
11th, 1 win, WQ, by Wang Sheng Jun
12th, 1 win, Wu, by Wu Tong Kuan
This year Handtalk was clearly superior, easily winning all its games. The
closest was a 21 point win over Go4++. Kao has completetly rewritten Stone,
working 10 hours a day since June. This is his first finish in the top
3 places. His new program focuses on fighting well, and reading life and
death accurately. WuLu is another good new program. It had bad luck in losing
one game on time, but had a very close game against go4++, so it is stronger
than it appears from this result. It is written by Prof Chen's assistants
and is only 2 years old.
Go Intellect has been ported to windows-95 on a PC, and is stronger than
last year.
The closest game was between Many Faces of Go and Go4++, which Many Faces
won by only 1 point, playing black.
Go4++ has a territory oriented style, while the other top programs all depend
on fighting strength. Last year Go4++ did very well, but the other programs
have had a year to get used to playing against this style, so it was not as
effective this year. Go4++ has a stylized opening, always taking 2 adjacent
3-3 points and splitting the other side on its first 3 moves. Handtalk
prepared a counter fuseki that prevented Go4++ from starting a moyo anywhere.
This year the human players were quite strong, and won all 3 games against
Handtalk with an 11 play handicap.
The full tournament grid (not in order of final result):
Wins SDS SOS-SDS Place
1) Handtalk +2 +4 +7 +10 +11 +6 6 1
2) Sason -1 +3 -4 -6 -7 +5 2 3 9
3) DFGO -4 -2 +5 +12 -6 -9 2 2 10
4) Many Faces +3 -1 +2 +7 -10 -11 3 7 5
5) Wu -6 -8 -3 -9 +12 -2 1 1 12
6) Jimmy +5 -7 -8 +2 +3 -1 3 5 12 7
7) Go4++ +8 +6 -1 -4 +2 -10 3 8 4
8) WuLu -7 +5 +6 -11 -9 +12 3 5 10 8
9) Go Master -10 -12 -11 +5 +8 +3 3 6 6
10) Go Intellect +9 +11 +12 -1 +4 +7 5 2
11) Stone +12 -10 +9 +8 -1 +4 4 3
12) WQ -11 +9 -10 -3 -5 -8 1 3 11
-David Fotland
RESULTS OF THE 1995 ING WORLD COMPUTER GO CHAMPIONSHIP.
The 1995 World Computer Go Congress was held on November 4 and 5 in
Seoul, South Korea. This year Acer no longer co-sponsors the
competition. Instead, there are several new sponsors, with the main
one being Hangul and Computer Company (HNC), the largest Korean language
software company. In addition to the usual prozes provided by the
Ing Foundation, HNC paid 1/2 air fare to any program that won two
games.
This year's competition had a smaller than usual field of 10 programs.
Star of Poland missed its first congress in over 10 years, since
the author, Janusz Kraszek, was injured in an automobile accident.
Rumor has it that he might host next year's competition in Poland.
This year the tournament was organized a little later than usual, and
invitations to the qualifying tournament were not available until early
October, which may have prevented some programs from entering.
Since there were only 10 programs, and the results of the first 3 places
were clear after 5 rounds, the sixth round was not played.
The final result was:
1st, 5 wins, Handtalk, by Chen ZhiXing
2nd, 4 wins, Go4++, by Michael Reiss
3rd, 3 wins, Go Intellect, by Ken Chen
4th, 3 wins, Many Faces of Go, by David Fotland
5th, 3 wins, Stone, by Kao Kuo Yuan and Chih-wen Hsueh
6th, 2 wins, Explorer, by Martin Muller
7th, 2 wins, Sason, by Chang Ho Lee
8th, 2 wins, Big Stone, by Dong Chul Lee
9th, 1 win, Rex 95, by Han Jung Kim
10th, 0 wins, Right Thought, by Min Ho Kwak
The top programs are still fairly close. Handtalk won all its games,
but only beat Go Intellect and Go4++ by 9 points each. Go4++ beat
Go Intellect by 15 points, and in 3 additional friendship games after
the tournament, Go intellect won two (by 35 and 23 points), and Go4++
beat Go Intellect again by 60 points. Many Faces and Go4++ were not
paired in the tournament, but in 3 friendly gams afterwards, Go4++ won
two and Many Faces won one.
Handtalk won 200,000 NT dollars for first place, then went on to
challenge 3 human experts at 15 stone handicap. The 15 stone prize has
been attempted 4? times before (by Goliath, Handtalk, and Go Intellect)
without success. This year, the human experts were 3 youth champions,
9 or 10 years old. They were given a few minutes to practice, then
the games began. Handtalk looked like it was doing very well, then
around move 100, a reporter stepped on the main power cord and all the
computers shut down.
All 3 games were started from the beginning, and Handtalk went on to
win 2 of the three games, losing to the 9 year old, and winning
another 100,000 NT dollars. Handtalk immediately challenged at the
13 stone handicap and won 2 of three again, for another 133,333
NT dollars. Next year the Human challenge will be at 11 stones
handicap.
Full tournament grid (not in order of final result):
1 2 3 4 5 Wins SDS
1) Handtalk +4 +7 +3 +2 +5 5
2) Go Intellect +6 +3 +5 -1 -4 3 8
3) Many Faces +9 -2 -1 +6 +7 3 6
4) Go4++ -1 +10 +6 +7 +2 4
5) Stone +10 +8 -2 +9 -1 3 3
6) Explorer -2 +9 -4 -3 -8 2 2
7) Sason +8 -1 +10 -4 -3 2 1
8) Right Thought -3 -6 -9 -10 -6 0
9) Big Stone -3 -6 +8 -5 +10 2 1
10) Rex 95 -5 -4 -7 +8 -9 1
-David Fotland
_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://computer-go.org/mailman/listinfo/computer-go