Tic-Tac-Toe
One move wins it. Finding that move is the puzzle.
A new puzzle every day. Each board arrives with pieces already in play, and somewhere on it is a single move that forces a win. Find it, finish in as few moves as you can, and build your streak.
How to play
- You play as X. The AI plays as O.
- Get 4 in a row to win, horizontally, vertically or diagonally.
- Each puzzle starts with pieces already placed. One move creates a forced win.
- The AI plays perfectly after your move, so find the right line before it counters.
- Fewer moves means a better record for that puzzle.
Your stats
Puzzle records — fewest moves to win
No records yet. Solve today's puzzle to set one.
What is this game?
A tactical variant of Tic-Tac-Toe on a 6×6 grid, where four in a row wins. The classic 3×3 version is solved and everybody knows it; a bigger board brings back real strategy, because threats can arrive from several directions at once and the AI will take any gap you leave it. Each puzzle loads a pre-arranged board with one specific job attached to it: find the single move that creates a fork nothing can stop.
A brief history
Tic-Tac-Toe in its 3×3 form is one of the oldest strategy games we know of, with roots in ancient Egypt. The name comes from a nineteenth-century British pencil-and-slate game played blind. The 4-in-a-row variant on larger grids is a cousin of the Japanese game Gomoku, five in a row on a 15×15 board, and of Connect Four. Both of those have been solved mathematically and shown to be first-player wins under optimal play.
A 6×6 board with a win length of four sits between those extremes. Small enough to hold in your head, large enough to have real tactics in it.
Core strategy
Everything in 4-in-a-row comes back to the fork: a position where you make two threats at once and only one of them can be blocked. One threat is easy to counter. Two aren't. Strong play usually means building a diagonal scaffold that can branch into either a horizontal or a vertical completion, so your opponent has to pick which threat to stop while the other one lands.
When you're looking at the pre-placed pieces, trace every partial line of three that belongs to X and ask one question of each: which extension creates a second line at the same time?
The AI defends and attacks well. It blocks any winning threat it sees, and it builds its own if you leave it room, so extending the obvious line won't get you there. You need a move that wins in spite of the AI's best reply. In puzzle mode that move is guaranteed to exist: every board has exactly one first move that forces the result no matter what the AI does.
One puzzle a day
One puzzle a day, drawn from the full set, and it's the same board for everyone who visits that day. Your first game against it is the one that counts: a win updates your streak and records the fewest moves you needed. After that the play button keeps dealing boards from the same set, as many practice games as you want, none of them touching your streaks or records. Good for studying a position, or just for a rematch with nothing at stake.
Tips for beginners
Start by scanning for any line of three that already exists, yours and the AI's both. If the AI has three in a row and can complete it, you have to block, unless your own winning move is available on this exact turn.
Then look at the diagonals. The pre-placed X pieces are always arranged diagonally, so the winning move is nearly always the one that extends that diagonal while threatening a cross direction at the same time. And when you genuinely can't see it, try whatever move sits closest to the centre of your existing cluster. Centre mass gives you the most branches.