Expert Chess Memory: Revisiting the Chunking Hypothesis

In the paper

“In their study, the skill of subjects (from 1000 to 2200 ELO) did not correlate with the recall of random positions 3 shown for a few seconds, but effectiveness of the search for the best move in these positions did correlate with skill.”

What the sources say

• Supported

  The Procedure section says players viewed the random positions for 8 seconds for recall, then studied the corrected positions while choosing moves. The Specific Recall Analyses section reports no significant correlation between recall and USCF rating (r = .10), but a significant correlation between good-move scores and rating (r = .75, p < .001). The Subjects section describes six rating classes, with mean ratings from 1,065 to 2,129. (LLM, conf=0.93)

  cites Holding and Reynolds (1982) source

In the paper

“Chunk presentation produced better recall than column presentation; and, in fact, better recall than presentation of the entire position for the same length of time (12 s).”

What the sources say

• Supported

  The Experiment II Results report higher mean recall with meaningful chunks (13.72 pieces) than with columns (10.69) or the entire-board presentation (12.77). The Discussion describes all three modes as six-slide displays and says the final slide lasted 2 seconds, consistent with the stated 12-second presentation. (LLM, conf=0.92)

  cites Frey and Adesman (1976) source

In the paper

“However, studies using interfering material in intervening tasks (Charness, 1974;Cooke, Atlas, Lane, & Berger, 1993;Frey & Adesman, 1976, Gobet & Simon, in press-a) have shown that this material does not interfere much with chess memory, thus implying that, as the interfering tasks reduce retention in STM, some information has to be transferred rapidly to LTM.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Charness, 1974
• ?Not checked — no full text of the sourcecites Cooke, Atlas, Lane, & Berger, 1993 source
• Supported

  The Experiment III Discussion reports that interpolated processing had little detrimental effect on chess-position recall. The General Discussion interprets this finding, together with Charness’s results, as suggesting that information is processed quickly and passed directly to long-term storage. (LLM, conf=0.93)

  cites Frey & Adesman, 1976 source

In the paper

“Wixted and Rohrer (1994) report inconclusive results on the latencies within clusters: in some studies, latencies increased with position, but did not in other studies.”

What the sources say

• Supported

  The CLUSTERING section reports that Graesser and Mandler found within-cluster interresponse times increased with output position, a finding replicated by Rubin and Olsen. It also reports that Gruenewald and Lockhead did not find that increase, matching the claim of mixed results. (LLM, conf=0.94)

  cites Wixted and Rohrer (1994) source

In the paper

“The large size of chunks we have found adds support to the template theory, inasmuch as the difference in the number of chunks by skill was small and as this number was within the supposed capacity of visual STM (Zhang & Simon, 1985).”

What the sources say

• Supported

  The Results section says visual or semantic short-term memory can retain two or three chunks, supporting the capacity estimate for which Zhang and Simon are cited. The large chunks, small skill-level difference, and implication for template theory are the citing authors’ own findings and interpretation, not findings attributed to this source. (LLM, conf=0.90)

  cites Zhang & Simon, 1985 source

In the paper

“BGP latencies are longer than those found by Chase and Simon (means around 3 s), which may reflect differences in strategies used by our subjects.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“About 79.5% of the corrected WGP latencies are now below 2 s, with a median of 1.37 s and a mode of 1.13 s, in reasonable agreement with the Chase and Simon's data.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“For our subjects, the median is 2.63 s and the mode is about 2.37 s.”

In the paper

“Although no subject actually refused to copy (or to recall) 4 a random position, this task elicited deep negative feelings.”

In the paper

“Their subjects studied the stimulus position for a short time (a few seconds), then replaced a few pieces on the copy board, repeating this cycle until all pieces had been replaced.Our subjects (especially the Masters) studied the stimulus position for some dozens of seconds before placing the first piece; later, they rarely revisited the stimulus.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“The times, consistent for our three skill levels, are close to Chase and Simon's, although a little slower even after the correction for the mouse.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“The corrected distribution of WGP latencies has now a median of 1.49 and a mode of 1.25.”

In the paper

“Whatever strategy was used, subjects had to correct numerous color errors.”

In the paper

“Stronger players tended initially to try to replace pieces by semantic groups; then, encountering difficulty, switched to reconstructing the position in a more systematic way, roughly line by line or column by column.”

In the paper

“In the present experiment, 79.5% of the WGP latencies (against only 1.11% of the BGP latencies) are less than 2 s, and 89.3% (against 4.4%) are below 2.5 s.”

In the paper

“The between-glance latency distribution has small peaks at 3.75 and 5.75 s and a median at 7.3 s.”

In the paper

“In the random positions, Masters are faster than the others (97.0, 98.8, 128.4 s), but slower than would be predicted from their times in game positions.”

In the paper

“We will show that our data generally agree well with the data from the earlier experiments of Chase and Simon (1973a and b), with some differences in sizes and numbers of chunks that are more compatible with the revised template theory than with the chunking theory in its original form.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a and b)

In the paper

“With random positions, Masters tend to study the stimulus in longer glances than those of weaker players, but to return less often to look at the stimulus position.”

In the paper

“Note also that the mean of the median largest chunks for random positions is constant across skill levels (6.3 pieces [sd = 2.1] in the copy task and 4.6 pieces [sd = 1.7] in the recall task, for all subjects pooled) and is well over what would be predicted by a theory postulating the visual encoding of individual pieces in STM.”

In the paper

“For game positions, the size of chunks varies with skill levels [F(2,23) = 11.81,p<.001], but size does not differ significantly between the copy and recall tasks [F(1,23) = 1.56, ns.]The interaction term is not statistically significant.”

In the paper

“The main effects of skill [F(2,23) = 4.84, MS e = 1.50, p <.05] and type of presentation [F(1,23) = 240.1,MS e = 1.68, p <.001] are significant, while the interaction is only marginally significant [F(2,23) = 3.23, MS e = 1.68, p <.06].”

In the paper

“For the Masters, the mean of the median largest chunks is 16.8 for the recall task, and 14 for the copying task.”

In the paper

“The percentage of placements classified consistently by the two methods is 72% for the task of copying game positions, 64% for copying random positions, 74% for recalling game positions, and 70% for recalling random positions.All four tables have chi-squares with probabilities below .0001.”

In the paper

“Below 1.8 s, placements having 3 or 4 relations dominate over placements with 0 or 1 relation.Above 2.2 s, the pattern is shifted.In the interval 1.8-2.2s, which includes the value of 2 s we have selected as a cut-off, the numbers with few and many relations are almost equal.”

In the paper

“Table 1, columns 2-3 and 5-6 show, averaged over all subjects (there was little difference in latency statistics between skill levels), the median latencies between the placement of two successive pieces for each combination of relations, for the Copy task in Random and Game positions, within and between glances.”

In the paper

“The within-chunk latencies do show a modest increase of about 50% over 30 pieces, but the between-chunk latencies increase by a factor of 2 over the same interval.”

In the paper

“Age also did not correlate significantly with recall of game positions (r = .06) or of random positions (r = -.36 ), though the observed correlation was stronger for the latter.”

In the paper

“Figure 3, which plots latency as a function of the number of relations in the recall task, for both short and long latencies combined (skill levels and types of positions are pooled), shows a clear negative correlation between number of relations and latencies, giving results similar to Figure 5 of Chase and Simon (1973a).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“From the Figure , we see that, for within-chunk conditions, the placements having few relations are below chance, while the placements having several relations are well above chance.Between-chunk placements, with a flatter trend, are overall much closer to chance.”

In the paper

“The strict systematic strategy was used by only one Master, on his second random board; but Experts and Class A players used this strategy in copying 38% of the random boards.”

In the paper

“The strategy was used by 22% of Experts and 33% of Class A players on the first board, 56% of Experts and 42% of Class A players on the second.”

In the paper

“The correlations with number of relations for the within-chunk conditions are: copy game, withinglance: .81;copy random, within-glance: .68;recall game, short latencies: .86;recall random, short latencies: .79.”

In the paper

“Leaving aside the data for recall of random positions, which were not computed by Chase and Simon, and the a priori probabilities, the correlation between the remaining items in our Table 2 and the corresponding items in Chase and Simon's (1973a) Table 7 is .78,accounting for 61% of the variance in the correlation coefficients.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon's (1973a)

In the paper

“For example, in 27 cases in game positions, two opposing pieces of the same kind attacked each other (and had no other relation), giving a probability of .001for the AS relation; and in 8,978 cases a pair of pieces had none of the five chess relations, giving a probability of .335for the null relation.”

In the paper

“In agreement with the theory, the within-chunk deviations from a priori probabilities are highly correlated with the number of relations, while this correlation is weaker for the between-chunk deviations.”

In the paper

“The correlations with the between-chunk conditions are: copy game, between-glance: .61;copy random, betweenglance: .56;recall game, long latencies: .69;recall random, long latencies: .31.”

In the paper

“We find the same pattern in our data: all four correlations of latencies with numbers of relations are negative, although some are not statistically significant.”

In the paper

“Finally, all but one of the latencies of the within-glance condition of the copying task are below 2 s, the exception being the case where there is only a relation of attack with game positions (this case occurs only in 0.5 % of the observations).”

In the paper

“We have computed the comparable data for our experiments, but as the two sets of data are very similar, we will summarize the comparison with Chase and Simon instead of reproducing our table in full.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“We obtained the following correlations with age: (a) for game positions: interpiece latency (.22), time to move the mouse (.10);(b) for random positions: interpiece latency (.26); time to move the mouse (.01).”

In the paper

“Subjects produce fewer chunks in the recall task than in the copy task, which simply reflects the fact that multiple chunks do not have to be held in STM in the copy task.”

In the paper

“For the withinglance copy placements in game positions, Spearman's rho correlation is -.77, and in the random positions the correlation is -.84.”

In the paper

“The shortest times are obtained with the PCS and DPCS relations, which mainly appear with pawn formations.”

In the paper

“Although we did find, as Chase and Simon did, some differences in the number of chunks between skill levels, the ordering of numbers by skill was wholly different in the two studies.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“The number of chunks replaced is inversely related to skill level for the copy task, while it shows an inverted U-curve for the recall task, with Class A players recalling the least chunks.Note that the difference between Masters and Class A players is small in the recall task: less than one chunk.”

In the paper

“In general, there was little difference among skill levels in the nature of the chunks: all show the same pattern of relations, which led us to pool the skill levels in the presentation of results.”

In the paper

“In summary, the relations of Same Type, Color and Proximity play a major role in predicting the latency when successive placements belong to the same chunk, but not when they belong to two different chunks.”

In the paper

“Besides, Masters show an interesting pattern: in contrast with the other players, their BGPs are much slower with random positions than with game positions.This difference accounts for almost the whole of the (marginal) interaction effect of Type of position x Skill : Experts and Class A players keep almost the same rhythm for the BGPs in both game and random positions.”

In the paper

“There was some indication that they were slightly superior also with random positions, a result that did not appear in Chase and Simon's data.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“The importance of relations for sequence of placements is closely related to the frequency of their occurrence, although proximity has a larger role than its frequency would predict.”

In the paper

“Second, numbers of chess relations in within-glance placements in the copy task and in placements within 2 s in the recall task were strongly correlated, as were numbers of relations in between-glance placements in the copy task and placements over 2 s in the recall task.”

In the paper

“important difference between WGP and BGP: WGP latencies are much shorter than BGP.ANOVA indicates this main effect of Mode of placement [F(1, 22) = 90.74,MS e = 10.3, p<.001].”

In the paper

“In particular, Masters recalled nearly three times as many pieces as Class players in game positions, but only 1 1/2 times as many in random positions.”

In the paper

“Although the 2-s boundary is only an approximation, it seems to be a reasonable approximation, for in our replication of the copy task, 79.5 % of the latencies for placements within a glance (as against only 1.11% of the latencies for placements between glances) were less than 2 s.”

In the paper

“Third, the size of chunks increased with skill, accounting for skilled players' superiority in the recall task.”

In the paper

“The mean number of references to the stimulus decreases with chess skill, and game positions require fewer references (2.5, 4.9, 7.2) than random positions (6.8, 10.8, 12.0).”

In the paper

“This does not imply that there will be any systematic increase over the period of recall in within-chunk latencies for successive chunks, nor do our data show substantial differences of this kind.”

In the paper

“Third, strong players tend make faster between-glance placements in game positions than amateurs, but not within-glance placements.”

In the paper

“Our data showed that, when we plot the latencies of pieces that have 2 or more chess relations with their predecessors, we find a small trend in size over time; when we plot the latencies of pieces that have 0-1 chess relations with their predecessors, we find a strong positive trend, with the latter slope more than twice the former.”

In the paper

“With game positions, the percentages of pieces correctly recalled are 92.0,57.1 and 32.2 for Masters, Experts and Class players, respectively.The corresponding percentages for random positions are 19.0,13.8 and 12.4.”

In the paper

“No main effect of Skill [F(2, 22) = 0.32, MS e = 13.5] or of Type of position [F(1, 22) = 1.63,MS e = 2.15] are found.A marginal interaction is signaled for Skill x Type of position [F(2, 22) =3.17, MS e = 2.15, p=.062].There are no other two-way or three-way interactions”

In the paper

“Masters created larger clusters at all levels of partitioning than did class B players.In addition, the chunks they detected at the basic level corresponded to the chunks identified by Chase and Simon (1973a).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Freyhoff, Gruber and Ziegler (1992)
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“although even small chunks can suggest good moves in tactical situations, and chunks or small constellations of them allow recognition of positions of particular types.”

In the paper

“Chase and Simon carried out little more than an exploratory experiment.They studied only a single Master, a single Expert and a single Class A player.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“However, the number of units decreased and their size increased with expertise, stronger players using overlapping sorting criteria that grouped chunks together.”

In the paper

“Pattern recognition then facilitates the generation of moves and plans during search and allows a rapid and precise evaluation of positions met during search.”

In the paper

“A computer program, MAPP (Simon and Gilmartin, 1973), simulated several experimental findings, including the percentage of pieces recalled by a class A player, 1 the types of pieces replaced and the chess relations between successive pieces in the reconstruction.”

What the sources say

• ?Not checked — no full text of the sourcecites Simon and Gilmartin, 1973 source

In the paper

“Pawn structures provide information about the squares on which pieces should be placed (e.g. a Knight in front of an isolated pawn) and also about typical pawn moves for given structures.”

In the paper

“In addition, as the subjects used actual chess boards and pieces, the maximum number of pieces they could grasp in one hand could have limited apparent chunk sizes.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“Charness found better recall in the chunking condition than in the column condition, and poorest recall in the random condition.The same results were found when pieces were presented visually, one at a time (Charness, 1974).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Charness, 1974

In the paper

“Gruber (1991) also found that, when allowed to ask questions about an as yet unseen position, chess experts asked about the past and future path of the game, about plans and evaluations, while novices asked about the locations of single pieces”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Gruber (1991)

In the paper

“Holding (1985,1992) has challenged this assumption on the grounds that (a) most chess patterns consist of pawns, 2 and pawn structures do not generate many moves; (b) that most chess patterns found by Chase and Simon (1973a,b) are too small to provide useful information; and (c) that pattern recognition is not sufficient to explain chess skill, because it applies only to the initial moves from the”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Holding (1985,1992)
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a,b)

In the paper

“In particular, 74.6% of the pieces within partitions shared three or more relations, as compared with 67.6% in Chase and Simon's data for the recall task.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Freyhoff, Gruber and Ziegler (1992)
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“First, their size was, on average, 3.6 pieces for masters, and 2.7 pieces for class B players-reasonably close, given differences in the types of positions used, to Chase and Simon's 2.5 pieces for the Master and 2.1 pieces for the Class A player.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Freyhoff, Gruber and Ziegler (1992)
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“Although Chase and Simon only mention chunks as eliciting (initial or subsequent) moves, chunks, especially the large chunks we call templates, can also provide information about the class to which the position belongs, about heuristics, plans, partial evaluation of the position and so on.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“Chase and Simon (1973b) proposed that Masters access information in long-term memory (LTM) rapidly by recognizing familiar constellations of pieces on the board, the patterns acting as cues that trigger access to the chunks.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973b)

In the paper

“Because storing one chunk in STM gives access to a number of pieces, masters perform remarkably well in recall tasks.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973b)

In the paper

“On the contrary, Chase and Simon (1973b, pp. 268-272) stated explicitly that recognition of patterns is used not only to generate base moves but also subsequent moves triggered by patterns in the”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973b)

In the paper

“One central thesis of the chunking model is that chunks act as cues that, when recognized, evoke access to heuristic suggestions for good moves.”

In the paper

“Simon and Gilmartin estimated that expertise in chess would require between 10,000 and 100,000 chunks in memory (in the literature, this range is often reported simply as 50,000 chunks).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Simon and Gilmartin

In the paper

“the finding of Chase and Simon (1973a) that the differences in ability of chess players at different skill levels to copy and to recall positions are attributable to the experts' storage of thousands of chunks (patterned clusters of pieces) in long-term memory.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“Moreover, the Master was rather inactive in chess at the time of the experiments and performed substantially less well than other Masters who have been tested in the same or similar tasks.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“According to the theory, pairs of pieces that have numerous relations are more likely to be noticed together, hence chunked.”

In the paper

“Finally, Chase and Simon's theory of memory implies that chunks, upon recognition, would suggest good moves to the masters.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“Because these chunks are associated with possible moves, chess masters can generally choose good moves with only moderate look-ahead search.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973b)

In the paper

“Chase and Simon (1973b) proposed that, during the brief presentation of a chess position, players recognize already familiar chunks on the board and place pointers to these chunks in a short-term memory of limited size.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973b)

In the paper

“Gobet and Simon (in press-a) propose a modified model that is in accord with other recent models of expert memory (e.g., Richman, Staszewski & Simon, 1995), and accounts for the rapid encoding shown by chess masters.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Gobet and Simon
• ?Not checked — no full text of the sourcecites Richman, Staszewski & Simon, 1995 source

In the paper

“Replications are rare in chess psychology research (Charness, 1988;Gobet, 1993), and the data supporting the 2-s boundary for delimiting chunks have never been replicated experimentally.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Charness, 1988
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Gobet, 1993

In the paper

“Second, subjects in the original study replaced pieces by picking up several of them simultaneously.Hand capacity will limit the number of pieces that can be grasped, hence the estimated size of chunks.”

In the paper

“Note also that, because of the design of the program, it was difficult to access the stimulus position and come back to the reconstruction board in less than one second, which may have provided one motive for fewer and longer references to the stimulus.”

In the paper

“Finally, templates contain pointers to symbols representing plans, moves, strategical and tactical concepts, as well as other templates.These pointers are also acquired at normal learning rates (i.e., 5 to 10 seconds per chunk).”

In the paper

“The template hypothesis and the evidence supporting it predict that strong players should replace positions in chunks (templates) larger than the ones identified by Chase and Simon (1973a).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“The modified theory asserts that-as in the chunking theory-chunks are accessed through a discrimination net.”

In the paper

“Several authors (Freyhoff, Gruber & Ziegler, 1992;Gold & Opwis, 1992;Holding, 1985;Reitman, 1976) have seen difficulties in Chase and Simon's method for defining chunks, among the most important of them: (a) difficulty in identifying chunks by reaction times, (b) impossibility of capturing overlapping or nested chunks, (c) difficulty in assigning pieces erroneously replaced and (d) the assumption”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Freyhoff, Gruber & Ziegler, 1992
• ?Not checked — no full text of the sourcecites Gold & Opwis, 1992
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Holding, 1985

In the paper

“This could account for the observed larger average size of the early than of the late chunks.”

In the paper

“This difference in behavior may be related to the differences in the ways in which stimuli were presented and responses made in the two sets of experiments.”

In the paper

“Still, there is warrant for testing further the validity of Chase and Simon's method for identifying chunks: infelicities in the original study; criticisms of the technique used; evidence that Masters perceive a position at a higher level than 4-5 piece chunks; a different prediction of the template theory about the size of chunks; the small number of subjects.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“In addition, if the close relation between the number of relations joining a pair of pieces and the likelihood of the pair being perceived in rapid succession were confirmed, then numbers of relations, on the one hand, and latencies, on the other, provide converging evidence about the numbers, sizes and character of the chunks that experts perceive.”

In the paper

“The two positions (the model and the position being reconstructed) were slightly shifted and of a different size, in order to avoid subjects' using iconic memory to superimpose one on the other.”

In the paper

“Chase and Simon found that pairs of pieces within chunks identified by the copy and recall methods showed the same pattern of relations, but a different pattern from that shown by pairs of pieces belonging to different chunks.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“To remove learning effects, the first position is omitted from the following analysis.”

In the paper

“These objections raise serious difficulties if the goal is to cut a chess position into precise chunks but are not fundamental for analyses that relate chunks to the distributions of relations between pieces, as is the case for Chase and Simon's (1973a) study and the present one.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“Notice that, although slots in templates can be filled rapidly, hence serve essentially as augments to STM in the domain of expertise, the templates themselves are built up slowly, at normal LTM learning rates.”

In the paper

“This is of course at variance with their findings (which, we have noted, were based on the performance of only a single Master).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“These results undermine the hypothesis that a pattern of pieces can be recognized independently of its position on the chessboard and add support to the estimate of 50,000 chunks.”

In the paper

“Moreover, as we have seen above, various alternative techniques (partitioning, sorting) provide converging evidence that supports the original results of Chase and Simon.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“We decided to pool the results from the two experiments, as there was no difference between the two experimental groups nor any interaction of experimental group with the variables discussed below.”

In the paper

“Our new experimental procedure eliminates these two potential problems.”

In the paper

“As in the case of the copy task, we judged the confounding due to the non-random order of presentation to be acceptable, because we were not primarily interested in comparing the levels of reconstruction of the game and random conditions.”

In the paper

“According to the chunking model, pieces are encoded in a STM of limited size during the recall task and no new information is then added to LTM.”

In the paper

“In their data, 80% of the WGP latencies were less than 2 s, with a median around 1 s and a mode around 0.5 s (estimated from their graph).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“Templates, besides containing information about a pattern of pieces, as chunks do, possess slots (variables that can be instantiated) in which some new information can be stored in a matter of seconds.”

In the paper

“Most of the criticisms we have reviewed were either due to misinterpretation of the chunking theory or pointed toward the necessity for postulating some kind of rapid encoding into LTM, a requirement that is now met in the template theory (Gobet & Simon, in press-a).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Gobet & Simon

In the paper

“Holding (1985Holding ( , 1992) ) argued that this number is much too large, and that as few as 2,500 chunks may account for the results obtained in recall experiments, by assuming that the chunks encode relations between pieces but not the location of these pieces.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Holding (1985, 1992)

In the paper

“Table 6 of Chase and Simon (1973a) gives the probabilities of the presence of different combinations of relations in the various experimental conditions.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“In spite of the difference in apparatus, the statistics derived from the current experiments agree closely with those reported by Chase and Simon (1973a).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“Chase and Simon had found that small average latencies in the copy task correlate with a large number of relations between successive pieces, while large average latencies correlate with few relations.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“In the recall task, 5 Chase and Simon (1973a) did find a difference in chunk size between their subjects, their Master obtaining, for the first chunk replaced with middle game positions, a mean of 3.8 pieces, while the Class A player had a mean of 2.6 pieces and the beginner a mean of 2.4 pieces.The median largest chunk per position was 5 pieces for the Master with game positions.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“The chunking hypothesis predicts that there would be many more relations of attack, defense, proximity, shared color and shared type of piece between successive pieces within the same chunk than between successive pieces on opposite sides of a chunk boundary.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“In addition, if the same processes determine the latencies in both the copy and recall tasks, then there should be a high correlation between the relations in the two tasks.”

In the paper

“The chunking theory predicts that the inter-piece latencies will stay more or less constant when pairs of pieces belong to a chunk, but that the inter-piece latencies between chunks may increase as a function of the number of pieces previously placed.”

In the paper

“If chunk boundaries are indicated by latencies > 2 s, then the relations between successive pieces should be different with short than with long latencies between them.”

In the paper

“The template theory predicts that experts develop larger chunks with practice than are predicted by the original chunking theory.”

In the paper

“Thus, the two methods of defining chunks produce quite similar segmentations of the output, and the findings reported in Chase and Simon's paper and in this paper would hold about equally well if we defined a chunk as a set of consecutively placed pieces each of which has 2 or more chess relations with the piece previously placed.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“The slowdown between chunks would follow if the players first replace salient chunks, then have to search a little longer to find the less salient, and therefore less easily recognized chunks.”

In the paper

“This was one of the most troublesome of their findings, for the original model postulated that the difference in recall between players of different skills was to be explained by chunk size differences not by differences in the chunk capacity of short-term memory.”

In the paper

“Finally, in two analyses extending Chase and Simon's, we showed that there was a considerable agreement in predicting whether a piece belongs to a chunk using either the number of relations or the latencies, and that the between-chunk latencies lengthen significantly over time, but the withinchunk latencies only slightly.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“All of these correlations show that within-chunk patterns of relations, whether in the copy or recall task, are quite different from between-chunk patterns, the latter resembling more closely the relations between pairs of pieces selected randomly.”

In the paper

“We do not show the statistics for the recall task, as the separation of within-chunk from between-chunk placements in that task on the basis of latency would confound independent with dependent variables.”

In the paper

“Specifically, the relations for within-glance placements in the copying experiments should correlate with those for rapid placements (≤ 2 s) in the recall experiments and the relations for between-glance placements in the former should correlate with those for slow placements (> 2 s), in the latter.”

In the paper

“As none of the correlations were significant at p = .10,we may conclude that age did not much affect time to move the mouse or compromise our estimate of chunk cut-offs.”

In the paper

“The former follows from the fact that chunks are postulated to be stored in LTM, and speed of recovery of their successive elements should be independent of the time when they are copied or replaced.”

In the paper

“As the number of chunks in the recall task differ little between the strongest skill level and the weakest, this supports the hypothesis that strong players use templates, supplemented by smaller chunks, to encode information rapidly.”

In the paper

“While this is somewhat possible for the positions of the copy task, the positions we have used in the recall task do not, as far we can judge, belong to typical openings situations (see Appendix A), and it is unlikely that the typicality of positions has inflated our estimate of the size of chunks.”

In the paper

“Such a time-sharing strategy should however level the inter-piece latencies and destroy chunks.”

In the paper

“Another possible factor in the differences in chunk size at Master level between the Chase and Simon experiments and ours is that their Master was somewhat out of practice (a score of only 71.5% for recall of middle game positions, as compared with 92% for our Masters), and, being in his mid-forties, may have been slower in replacing pieces, causing some chunks to be divided in scoring them.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“In both cases, the player has spent enough time studying this position or type of position to have acquired large, well differentiated templates for it.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase & Simon, 1973a

In the paper

“The correct claim, and the one actually made by Chase and Simon and in this paper, is that skill in playing chess depends both on (a) recognizing familiar chunks in chess positions while playing games, and (b) exploring possible moves and evaluating their consequences.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“(It cannot be assumed that Masters have superior motor skills in placing pieces, for experience with the use of computer and mouse was evenly distributed among the subjects in our experiments-skill level accounts for less than 0.3 % of the variance when used to predict the time to move the mouse.)”

In the paper

“Summarized in one sentence, the message of this paper is that chunks are larger than estimated by Chase and Simon, but that, as they showed, the pattern of relations between two pieces placed successively are radically different when the pieces do or do not belong to the same chunk.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon

In the paper

“Our explanation is that the large chunks are built around templates that encode information, acquired by strong players over years of practice, about typical and familiar positions, and that provide rapidly fillable slots for additional chunks of information about the current position.”

In the paper

“When a physical chessboard and pieces are used, the maximum chunk size is limited by the number of pieces the hand can grasp; with our present apparatus, it is not limited in this way.”

In the paper

“The superiority of strong players for game positions is then due not to the number of chunks (Masters replaced only one chunk more than Class A players, in the recall task), but to the presence of a few large chunks.”

In the paper

“These results, based on a larger sample (26 subjects) than that of the earlier experiment, support the major findings of Chase and Simon (1973a) and corroborate their hypothesis that the same information processing mechanisms, operating on chunks stored in long-term memory, determine the time intervals in both the copy and recall tasks.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Chase and Simon (1973a)

In the paper

“It is unclear why this should be the case more with Masters than with Class A players.The latters' larger chunks (at most 6-7 pieces) are not unusually large and can easily be explained, for example, by a few common patterns, like castling positions.”

In the paper

“However, as the largest chunks of highly skilled players are much more complex than those of weaker players, important qualitative differences may be present in these large templates.”