Capture Go: The One-Rule Door into the Deepest Game
Go has a reputation problem. Everyone has heard it's the deepest board game in the world, with roots in China well over two thousand years ago and a professional scene that treats openings the way conservatories treat music. Almost nobody starts, because the full game seems to demand a mountain of study before the first satisfying moment. Capture Go is the door the teaching world cut into that mountain, and it needs exactly one sentence: surround a stone and you win. Our Capture Go plays it on a nine by nine board against the computer, first capture takes the game, and everything you learn transfers whole to the real thing.
Liberties are the entire rulebook
Stones sit on the intersections of the lines, not the squares, and every stone lives or dies by its liberties: the empty points directly beside it. A lone stone in the open has four. Stones connected along the lines share their liberties as a single group, and that pooling is the game's first deep idea, because a group is exactly as alive as the empty points around it. Fill the last liberty of any stone or group and it comes off the board. In Capture Go, that removal is the finish line: the first capture wins on the spot.
Two consequences follow immediately, and they are the same consequences that govern professional Go. First, the edge of the board is nobody's friend. A stone on the side starts with three liberties, in the corner just two, so fights near the edge tilt toward the attacker and beginners who hug the rim are volunteering for trouble. Second, you may not place a stone onto a point where it would have no liberties at all, no suicide, which in our version the board simply refuses. Those two facts, plus the capture rule, are the complete law of the land.
From them grow the two capturing shapes every Go player learns first. The ladder: a group reduced to one escape route runs diagonally across the board while the attacker checks it at every step, and whether it lives depends entirely on what's waiting in the far corner. The net: instead of chasing, the attacker places one calm stone that closes every exit at once, and the prey is dead without ever being touched. Chase a group in our version and you'll meet both shapes inside ten minutes; recognizing them is most of what beginner Go instruction is for.
One word you need: atari
Go has a small vocabulary and one word matters on day one. A stone or group with exactly one liberty left is in atari, one move from capture, and saying the word to yourself changes how you see the board: is that group of mine in atari, is theirs, and who profits if I answer versus ignore it. Most Capture Go games are decided by a player who missed an atari on their own stones while hunting the opponent, so make the check mechanical: after every enemy move, count your loosest group before planning anything. The habit costs two seconds and wins more games than any clever attack.
Board size is part of the teaching design too. Full Go is played nine teen lines by nineteen; our version uses the nine by nine board that Go schools start students on, big enough for real ladders and nets, small enough that every stone matters from move one and games finish in minutes. The scaling is honest: nothing you learn on the small board is unlearned on the big one, the big board just adds more room for the same truths to operate in.
The three instincts it installs
Capture Go is used by teachers because winning at it quietly installs the reflexes the full game runs on.
Count before you place. The question is never where looks good, it's how many liberties each group has after the move. Strong play is a running audit: my group three, yours two, so I can afford one more push before I must connect. The arithmetic habit is identical to counting the race in Quoridor, and once you have it, boards stop looking like patterns and start looking like numbers.
Know when to abandon ship. The beginner's ruin is adding a third stone to a doomed pair, then a fourth to the doomed three. Every stone added to a dying group raises the prize for capturing it. The moment a group's escape arithmetic fails, the strong move is to play somewhere else entirely, and learning to feel that moment without grief is, no exaggeration, the central emotional skill of Go.
Make walls face somewhere useful. Even in a race to one capture, stones do double duty: the group that chases also builds a wall, and a wall pointed at open board keeps paying after the chase ends. Aimless contact fighting builds walls that face nothing.
When the first capture stops feeling like the point, you've outgrown the training wheels, and the full game with its territory, its life-and-death puzzles and its two-eye rule is waiting; the same board, the same liberties, everything you've built still true. The neighboring aisles reward the same instincts: Gomoku for open-board threats, Bagh-Chal for asymmetric surround-and-trap, Reversi for the strange arithmetic of flanking. All free, all in the browser, all in the board room. The mountain has a door, and it's one rule tall.