Aesthetics in mate-in-3 combinations part II: Normality

Abstract

In part I (see Iqbal, 2010a), experimental results showed that the correlation strength of the scores generated by a computational aesthetics model (for mate-in-3 combinations in chess) with the mean human-player aesthetic ratings alone can be misleading. Moreover, it was shown that the use of weights or multipliers (even those provided by domain experts) to adapt aesthetic features is unreliable. In this article, the probability distribution of the human ratings is explored as a third criterion to substantiate the envisaged model's viability (i.e., after achieving of a minimum qualifying standard, and by having a reasonably good correlation with the human ratings). Only one approach from the thousands of alternatives tested was found that resembled the human ratings in this way. It combined a specific technique (viz. a 'random-alternating' technique using a specific probability-split) with selections of features that are both added and subtracted. The new and unexpectedly adequate stochastic approach contrasts with the author's deterministic existing model that generates only precise aesthetic scores. Given (a) the new model's closer resemblance to the human ratings, (b) its ability to 'change its mind' now slightly, and (c) the otherwise equivalent performance to the existing model, the new model was considered an overall improvement and a recommended modification. Additionally, this article highlights a curious 30-70 'strictness rule' which suggests that humans appreciate only the top 30% of aesthetic features associated with an object, and simultaneously penalize it for (up to) the remaining 70% that 'try' but fail to 'impress'.