[an error occurred while processing this directive]

Computer Octi Tournaments

Three Computer Octi tournaments have been held annually, all at Yale University in New Haven, CT.

Fourth Tournament: 2003

The fourth Computer Octi tournament is scheduled for December 2003.

Third Tournament: 2002

The third Computer Octi tournament had three sections: Octi-9x9 for three bases, Octi-9x9 for one base, and Octi-6x7. This was the first year for Octi-9x9 one base, and the second for Octi-6x7. This was also the first year that the games were simulcasted on the Octi Online server, which archived them automatically.

The Octi-6x7 tournament was held on November 9, 2002, although two of the games were played later. Four people entered programs: Aaron Armading (Aaronbot), Charles Sutton (Casbah), Jeff Bacher (testme), and Martin Hsu (martinbot). The results of the tournament were:

Aaronbot Casbah martinbot testme TOTAL
Aaronbot 2-0 0-2 2-2
Casbah 2-0 0-2 2-2
martinbot 0-2 0-2 0-2 0-6
testme 2-0 2-0 2-0 6-0

The Octi 9x9 one-base tournament had two entrants, Casbah (Charles Sutton) and testme (Jeff Bacher). Both games were won by testme. Don Green described the games thus:

The first game had the Charlesbot playing first. Both sides emphasized stacking early, with Charlesbot putting all but one reserve down at the outset. Charlebot by turn 22 became enmeshed in an increasingly nasty set of traps, forcing it to "retreat" into the center to avoid compounding losses. By turn 32, big exchanges were threatened by both sides, with the Charlesbot hunkering down as the Jeffbot lashed out with attacks. With 3 captured pods and no reservese, Charlesbot nevertheless gained tempo on turn 38 and looked to be making a comeback. By turn 42, I felt that the Charlebot position had drawn even, given the lack of structure to the Jeffbot's position. (Do you agree?) But Charlesbot dithered to turn 52, allowing the Jeffbot to build a beachhead and play for exchanges. Even as late as turn 56, I felt that Charlesbot still had some life in it, but Jeffbot gathered a base-striking force in the Charlesbot's back lines and moved in for the kill. A very interesting see-saw battle.

Game 2 was rather different. Again, Charlesbot fell behind early, but gained tempo and an overwhelming advantage in the center by turn 38. But rather attacking by moving 46-45, it continued to dither and let the Jeffbot regroup. Never let an offensive opportunity slip away if you are behind in Octi, because your opponent will convert your wasted tempos into prongs and pods! Charlesbot's move into the center on turn 42 was poorly coordinated with the pods in its formation (the sequence of moves at 42 and 44 are in some sense backward, perhaps due to limited search horizon), and the Charlesbot quickly ran into snags (all the while, with an enemy pod in its back lines). Five turns of costly exchanges left the Charlesbot down 4-2 in pods and out of reserves.

The three-base tournament also had two entrants: Casbah and testme. This time, testme's deeper search led to an easy crush of Casbah. So testme swept all three sections of the tournament.

For reference, here are the game records for all the sections:


Octi-9x9 one-base

Octi-9x9 three-base

Second Tournament: 2001

The second Computer Octi Tournament was held on August 25, 2001. For the first time, there were two divisions: one for Octi for Kids programs, and one for programs that play the adults' game. The Octi division was a rematch of the previous year, Sutton vs. Chiou. This time Chiou won. In the kids division there were four entrants--Aaron Armading, Charles (Hal), Thomas Hallock, and Charles Sutton. Armading was the winner.

I have written a tournament report, which was published in the ICGA Journal.

Each of the 2001 entrants answered some questions about their programs.

First Tournament: 2000

The first annual Computer Octi Tournament was held in New Haven on October 21, 2000. There were two entrants, Charles Sutton and Charles Chiou, with Sutton the winner.

By Charles Sutton, at casutton@cs.umass.edu. Last modified 20 Sept 2003.

[an error occurred while processing this directive]