Traveling there is out of the question for me. I'm also quite certain
I would not place well. I can contribute an executable if someone
wants to run my bot, they could keep any prize money.
Sent from my iPhone
On Jun 9, 2008, at 1:25 PM, Peter Drake <[EMAIL PROTECTED]> wrote:
We'd like to get an estimate of numbers. Who's planning to enter the
US Go Congress Computer Go Tournament?
Here is the latest version of the info on the tournament:
Computer Go Tournament
Tournament Director
Peter Drake (AND CO-DIRECTOR?)
Description
This 19x19 tournament is for computer programs only. While there
have been notable breakthroughs in recent years, computer Go remains
an open problem in artificial intelligence research.
Location
NEED LOCATION (at US Go Congress in Portland, Oregon)
Schedule
Rounds will be played as games are completed, beginning at 1:00 PM,
Sunday, August 3. Time permitting, we hope to run a double round-
robin tournament.
Time Limits
60 minutes per player, no byo-yomi. If there are many entrants, we
may reduce this to 45 minutes per player.
Rules
To give programmers as much time as possible to work out networking
bugs, we will use the same rules as the monthly KGS computer go
tournaments. These rules can be found at http://www.weddslist.com/
kgs/. Of particular note are that programs must implement GTP and
that Chinese (area) scoring is used.
All hardware must be physically present in the competition room.
Programs may not rely on any remote resources. Programmers must be
able to roughly explain and demonstrate the program’s “thought
process”, e.g., by showing logging output as the program responds to
an arbitrary position. Programmers may bring their own hardware; we
will also provide some machines (NEED DESCRIPTION).
Programs must be entered by the authors, although authors who cannot
attend the Congress may sent alternate operators. If a program is
based on another program (e.g., GNU Go or MoGo), it must contain a
significant amount of new code. The tournament director has final
say over what constitutes a "significant amount"; as a guideline, a
radically different search algorithm or life-and-death evaluator
would be significant, but tweaking some parameters or adding new
patterns to a database would not. Such derivative entries must
include written permission from the authors of the original program.
If you have any doubts about the eligibility of your program,
contact the TD before buying an airline ticket!
Prizes
Prize money has been donated by Hierarchical Systems Research
Foundation and other anonymous donors. At a minimum, the following
prizes will be awarded:
1st place $400
2nd place $200
3rd place $100
4th place $50
Peter Drake
http://www.lclark.edu/~drake/
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
