Reviewing Expert Chess Performance

Explaining expert chess players' dramatically superior skill represents an outstanding unsolved theoretical problem for cognitive psychology. This review extends and re-evaluates the current state of theories on chess skill, highlighting the strengths and weakness of various theories ranging from general abilities to chunking theory to more modern variants. After describing limitations with earlier approaches, a new theory is described based on Productions RElating STored Organizations (PRESTO) of chess pieces in memory. I discuss how this theoretical framework, extending and elaborating Ericsson and Kintsch's (1995) proposal for long-term working memory in chess, addresses established empirical findings, makes testable new predictions, and how it is related in several respects to the earlier theories. These predictions are examined in three studies, illustrating how the model can explain characteristics of chess problem solving and planning behavior as well as a historic increase at the highest level of skill.