The ability of consumers to anticipate the value they will draw from new product generations that expand the capabilities of incumbent goods is explored. Drawing on previous research in affective forecasting, the work explores a hypothesis that consumers will frequently overestimate the benefits they envision drawing from new added product features and underestimate the learning costs required to realize those benefits. This hypothesis is tested using a computer simulation in which subjects are trained to play a Pacman-like arcade game where icons are moved over a screen by different forms of tactile controls. Respondents are then given the option to play a series of games for money with an incumbent game platform or pay to play with an alternative version that offer either expanded (Experiments 1 and 2) or simplified (Experiment 3) sets of controls. As hypothesized, subjects displayed an upwardly-biased valuation of the new sets of controls as measured by actual versus forecasted usage rates and performance gains. Yet, when given the opportunity to be paid to trade down to a more efficient device in exchange, few accepted. We thus observe a paradox where the presence of forecasting mistakes in product adoptions does little to induce regret in ownership.