Horizon Effect in Artificial Intelligence

Horizon Effect problem involves an extremely long sequence of moves that lead to a strong advantage for one player, but where the sequence of moves, although potentially obvious to a human player, takes more moves than is allowed by the bounded search. This bounded Minimax Search problem is known as Horizon Effect, coined by Berliner in 1977.

Principle of Horizon Effect:

The basic principle behind Horizon Effect is that the search tree is only examined to a particular depth. All nodes at this depth are considered to be leaf nodes and are evaluated using a static evaluation function.

Horizon Effect Algorithm:

function bounded_minimax(current_node, max_depth)
   if is_leaf(current node)
     return static-evaluation(current node);
   if (depth-of(current node)==max-depth))
     return static-evaluation(current node);
   if (is-min-node(current node))
     return min(minimax(children-of(current node)));