Forecasting sales of new virtual goods with the Elo rating system

With the rapid growth of online games, firms increasingly sell virtual goods for use within their online game environments. Determining prices for such virtual goods is inherently challenging due to the absence of explicit supply curve as the marginal cost of producing additional virtual goods is negligible. Utilizing sales data, we study daily revenue of a firm operating a virtual world and selling cards. In particular, we analyze the impact of new product releases on revenue using ARIMA with intervention analysis. We show that during initial days after a new product release, the firm’s daily revenue significantly increases. Using a quality measure, based on the Elo rating method, we determine the relative good prices according to good usage data. Applying this method, we show that the rating of a product can be a good proxy for the number of units sold. Our quality‐based measure can be adopted for pricing other virtual goods.