Calculation tree

A calculation tree represents possibilities as an inverted tree. The current position is the root. Every move considered creates a branch to a new position, and every opponent reply divides the analysis again. One concrete sequence from the root to a later point is a line.

Parts of the model

• Root: the position where the decision begins.
• Node: any resulting position within the analysis.
• Branch: an alternative that separates from another sequence.
• Leaf: a position where calculation stops and an evaluation, a judgment of which side stands better, is made.

The full tree of a position can be enormous because each side often has many legal moves. A person does not try to explore all of it. They select a small group of plausible moves, prioritize checks, captures, threats, and strong replies, and abandon clearly inferior paths. The candidate-move method formalizes that initial choice of branches.

The model helps prevent positions from being mixed. A reply that belongs to one branch cannot automatically be transferred to another where pieces occupy different squares. In training or post-game analysis, drawing a small tree exposes omitted replies, final positions judged too early, and alternatives compared after calculating one much farther than another.