It terminates the game more quickly.

It resets all previously calculated alpha and beta values.

It automatically assigns the worst possible outcome to the minimiser.

It reduces the search space by not exploring certain branches.

It extends the search to shallower levels in the tree.

Only leaf nodes.

Both maximiser and minimiser nodes.

Only minimiser nodes.

Any node with multiple children.

Only maximiser nodes.

It automatically assigns the best possible outcome to the maximiser.

It resets all previously calculated alpha and beta values.

It terminates the game immediately.

It extends the search to deeper levels in the tree.

It reduces the search space by not exploring certain branches.

When the game reaches a stalemate condition.

When both players agree on the optimal strategy.

When the maximiser's best option (alpha) is greater than or equal to the minimiser's best guarantee (beta).

When all children of a node have been evaluated.

When the minimiser's best guarantee (beta) is less than or equal to the maximiser's best option (alpha).

Only leaf nodes.

Both maximiser and minimiser nodes.

Only minimiser nodes.

Any node with a single child.

Only maximiser nodes.

When the maximiser's best guarantee (alpha) is greater than the minimiser's worst option (beta).

When both players agree on a suboptimal strategy.

When all children of a node have been evaluated.

When the minimiser's worst option (beta) is less than or equal to the maximiser's best guarantee (alpha).

When the game reaches a draw condition.

At any middle level in the tree.

At the deepest level of the tree where leaf nodes reside.

At the topmost level of the tree near the root.

Alpha cutoffs can occur at any level except at the root.

Only at nodes that are directly above leaf nodes.

A reduction in the beta value at a maximiser node.

A situation where the alpha value is reset to negative infinity.

A method to increase the search depth dynamically.

A scenario where a minimiser node stops evaluating further because a better value is already assured elsewhere.

A scenario where a maximiser node stops evaluating further because a better value is already assured elsewhere.