Chess Expertise

History: Building up the literature

How did chess expertise began as a form of study and on what path has the research taken? An even better question, where is the literature heading?

Getting Started: De Groot

Given these unique properties of chess, it becomes understandable why there is such a depth of literature on chess expertise. But the seminal work upon which most current articles are built start with the work of De Groot (1966). Looking at the qualitative description of processes chess players carry out when deciding on what move to make, he found no difference between players of various strengths in terms of search. He did, however, find that better skilled players could recall positions on a chess board better than novices.

A new approach: Simon and Chase
Graph of latency and chunk
Simon and Chase argued for "chunks" by saying that pieces placed on the chess board at roughly the same time (little latency) were highly associated (more relations) thus "chunked" together.
(Simon and Chase, 1973)

As discussed in the overview of the four main theories of expertise, De Groot’s work lead to Simon and Chase’s seminal 1973 articles which spelled out a recognition-based paradigm in conceptualizing chess expertise. These “chunks” (or units of related items) explained how experts can retain more than what was thought of as the limit at the time set by George Miller at 7 +/- 2. (Miller, 1956). Simon and Chase (1973) employed a “standard chess memory task” in which players simply recalled a game position presented for 5 seconds. Afterwards, they had to construct the game position on a board while their timing and eye movements were recorded. Simon and Chase argued that the grouping (or fast placement of certain pieces) on the board at or close to the same time, followed by a pause, indicated the existence of a “chunk.”

Role of Perception and Knowledge

But as stated earlier, chess is made up of numerous different strategies and plans which cannot rest purely on low-level perception. A rich knowledge-base that was well searched and evaluated was more akin to what expert processing was doing and explained why experts seem to possess a qualitatively different organization of knowledge when compared with novices (Chi et. al, 1981; Fryhoff et. al, 1992) This emphasis on the role of knowledge and search lead to Holding’s SEEK theory which due to lack of elaboration, still remains difficult to model.

Robust Memory

As interference data came to light (Charness, 1976) and the use of rapid presentations showed that quick LTM (long-term memory) encoding seemed to be important to expertise. Ericsson would shortly follow with his theory of long-term working memory which posited the idea of a “retrieval structure” that aided experts in memorization (Ericsson & Kintsch, 1995). Building on the ideas already found in skilled memory (Chase & Ericsson, 1981), LT-WM meant the inclusion of elaboration of retrieval cues and unlike “chunking” theory, accounted for the lack of interference with chess material in that information was rapidly encoded into long-term memory upon initial presentation.

Generalizing and Beyond

Regardless of which theory seems to hold for chess expertise, the domain of chess has helped to bring about questions that serve a usefulness in terms of asking the same questions across domains. Interference and eye movement data isn’t unique to chess, but was first asked in this domain which then naturally extends itself to asking in other domains. Because of the vast amount of research done in the area, chess remains one of the testing grounds of any new theory that has any hope of attaining validity.

Back to Chess Expertise...(Back)


Other Applications in Expertise

Valid XHTML 1.0! Valid CSS! Level Double-A conformance icon, 
          W3C-WAI Web Content Accessibility Guidelines 1.0 Bobby 508 Compliant! © Copyright Victor Long, Chandra Singh, and David Snitkof 2005
All Rights Reserved