In two-player sequential games, a ply refers to one turn taken by one of the players. The word is used to clarify what is meant when one might otherwise say "turn".
The word "turn" can be a problem since it means different things in different traditions. For example, in standard chess terminology, one move consists of a turn by each player; therefore a ply in chess is a half-move. Thus, after 20 moves in a chess game, 40 plies have been completed—20 by white and 20 by black. In the game of Go, by contrast, a ply is the normal unit of counting moves; so for example to say that a game is 250 moves long is to imply 250 plies.
The word "ply" used as a synonym for "layer" goes back to the 15th century. Arthur Samuel first used the term in its game-theoretic sense in his seminal paper on machine learning in checkers in 1959, but with a slighty different meaning: the "ply", in Samuel's terminology, is actually the depth of analysis ("Certain expressions were introduced which we will find useful. These are: Ply, defined as the number of moves ahead, where a ply of two consists of one proposed move by the machine and one anticipated reply by the opponent").
In computing, the concept of ply is important because one ply corresponds to one level of the game tree. The Deep Blue chess computer which defeated Kasparov in 1997 would typically search to a depth of between six and sixteen plies to a maximum of forty plies in some situations.