Players in a game are "in equilibrium" if they are rational, and accurately predict other players' strategies. In many experiments, however, players are not in equilibrium. An alternative is "cognitive hierarchy" (CH) theory, where each player assumes that his strategy is the most sophisticated. The CH model has inductively defined strategic categories: step 0 players randomize; and step k thinkers best-respond, assuming that other players are distributed over step 0 through step k Ϫ 1. This model fits empirical data, and explains why equilibrium theory predicts behavior well in some games and poorly in others. An average of 1.5 steps fits data from many games.

No abstract

Self-tuning experience weighted attraction (EWA) is a one-parameter theory of learning in games. It addresses a criticism that an earlier model (EWA) has too many parameters, by fixing some parameters at plausible values and replacing others with functions of experience so that they no longer need to be estimated. Consequently, it is econometrically simpler than the popular weighted fictitious play and reinforcement learning models. The functions of experience which replace free parameters "self-tune" over time, adjusting in a way that selects a sensible learning rule to capture subjects' choice dynamics. For instance, the selftuning EWA model can turn from a weighted fictitious play into an averaging reinforcement learning as subjects equilibrate and learn to ignore inferior foregone payoffs. The theory was tested on seven different games, and compared to the earlier parametric EWA model and a one-parameter stochastic equilibrium theory (QRE). Self-tuning EWA does as well as EWA in predicting behavior in new games, even though it has fewer parameters, and fits reliably better than the QRE equilibrium benchmark.

No abstract

The complexity of managing a category assortment has grown tremendously in recent years due to the increased product turnover and proliferation rates in most categories. It is an increasingly difficult task for managers to find an effective assortment due to uncertain consumer preferences and the exponential number of possible assortments. This paper presents an empirically based modeling framework for managers to assess the revenue and lost sales implication of alternative category assortments. Coupled with a local improvement heuristic, the modeling framework generates an alternative category assortment with higher revenue. This framework, which consists of a category-purchase-incidence model and a brand-share model, is calibrated and validated using 60,000 shopping trips and purchase records. Specifically, the purchase-incidence model predicts the probability of an individual customer who purchases (and who does not purchase) from a given product category during a shopping trip. The no-purchase probability enables us to estimate lost sales due to assortment changes in the category. The brand-share model predicts which brand the customer chooses if a purchase incidence occurs in the category. Our brand-share model extends the classical Guadagni and Little model (1983) by utilizing three new brand-width measures that quantify the similarities among products of different brands within the same category. We illustrate how our modeling framework is used to reconfigure the category assortment in eight food categories for five stores in our data set. This reconfiguration exercise shows that a reconfigured category assortment can have a profit improvement of up to 25.1% with 32 products replaced. We also demonstrate how our modeling framework can be used to gauge lost sales due to assortment changes. We find the level of lost sales could range from 0.9% to 10.2% for a period of 26 weeks.Retailing, Product Assortment, Brand Reconfiguration, Purchase Incidence, Brand Share, Logit Model

St. Louis, and Wharton for their helpful suggestions. The authors are especially grateful to John Lynch for his help in developing a behavioral underpinning for their model. David Bell, Pete Fader, and Bruce Hardie generously provided the data. The authors also thank three anonymous for their many suggestions. Wagner Kamakura provided numerous detailed and helpful suggestions. A Parsimonious Model of Stock-Keeping Unit Choice Teck H. Ho and Juin-Kuan Chong Managerial SummaryThis article develops a model to describe and predict consumer stock-keeping unit (SKU) choice in frequently bought product categories. The model posits that a product category consists of several salient attributes and that each attribute has different levels and represents a SKU as an attribute-level combination. Our latent-class 2-segment model has a fixed number of 59 parameters for a category with 3 salient attributes, and in general has 11 + 12 (K+1) parameters for a K-attribute product category. We achieve this model parsimony by neither discarding data nor aggregating the level of analysis beyond the SKU level. Since the number of parameters does not depend on either the number of SKUs in the category or number of levels in each salient attribute, it is particularly useful for demand forecasting and inventory planning in large product categories that have hundreds of SKUs.Our model relies on three behavioral premises on how consumers choose products over time. First, the consumers accumulate not only a product-level experience but also attribute-level experiences. Second, these experiences have both consumption and shopping components. While consumption occurs only for the chosen attribute levels and product, shopping applies to all familiar and available attribute levels and products.Third, the consumption and shopping experiences increase with attribute-level and product familiarities.Using an extensive panel-level data set of sixteen categories involving more than one hundred and thirty thousands purchase records, we show that our model can describe and predict SKU choice well. In addition, we benchmark our model against the classical Guadagni and Little's model and its extension Fader and Hardie's model in a subset of seven small categories (where the number of the parameters for the latter models are less than 200). On average, our model fits 7% better in-sample and predicts 8% better out-ofsample in hit probability. In terms of adjusted pseudo R-square, the model is 8% and 11%higher in-sample and out-of-sample, respectively. This superior performance requires only one-half the number of parameters.Below are several ways how brand managers can use our model in practice:• Base volume forecasting: Our model can be used to forecast regular sales volume (i.e., base volume) of any SKU in a product category. Our model reveals the relative contribution of each attribute level to the base volume while controlling for the marketing mix effects.• Relative importance of each attribute: Using the model, one can easily analyze the relative im...

Noncooperative game theory combines strategic thinking, best-response, and mutual consistency of beliefs and choices (equilibrium). Hundreds of experiments show that in actual behavior these three forces are limited, even when subjects are highly motivated and analytically skilled (Camerer, 2003). The challenge is to create models that are as general, precise, and parsimonious as equilibrium, but which also use cognitive details to explain experimental evidence more accurately and to predict new regularities. This paper describes three exemplar models of behavior in one-shot games (thinking), learning over time, and how repeated "partner" matching affects behavior (teaching) (see Camerer et al., 2002b).

Although variation in finishing techniques has been shown to affect microleakage, little research has been published on the influence of finishing time on the marginal sealing ability of new generation composite bonding systems. The objective of the present study was to evaluate the influence of finishing time on the enamel and dentine marginal sealing ability of four new generation composite systems. Two class V preparations, which were solely in enamel or dentine, were made on the buccal surfaces of 96 freshly extracted molar teeth. The teeth were randomly divided into four groups of 24 and restored with composite resin (Silux Plus) utilizing the following bonding systems: Scotchbond Multi-purpose, Fuji Bond LC, Prime & Bond 2.0 and Bisco One-step. Half of the restorations in each group were finished immediately after light polymerization and stored for 1 week. For the remaining restorations, finishing was delayed for 1 week. The storage medium was isotonic saline at 37 degrees C throughout the experiment. All restorations were then thermocycled, subjected to dye penetration testing, sectioned and scored. The results suggest that the finishing of composite restorations, bonded with the bonding systems evaluated, should be carried out immediately after light polymerization. Delayed finishing does not improve but instead can be detrimental to the marginal seal of the restorations. The effects of delayed finishing are, however, both bonding system and tissue specific.