Four in a Row

Four in a Row. The board has 7 columns, a to g, and 6 rows, 1 at the bottom. A disc dropped into a column falls to the lowest empty cell. The first player with four of their own discs in a line, across, up and down, or diagonal, wins. A full board without a line of four is a draw.

Red moves first, and the two models swap colours for the return match. A win on the last empty cell beats the full-board draw.

Every move is one choice question. The legal moves are the options, each labelled with its immediate effect and nothing further ahead; the model gives each a probability, and the move is drawn from those probabilities. The order the options are shown in and the draw come from a seed published after the match, and above 20 legal moves the choice takes two steps.

Every other move, each model also answers “you win this game”. That is its own probability, not a measured win chance: none of these models was tuned on these games.

None of these models was trained for games. They play as untrained baselines.

Choice questionNoulHow requests work