With the increasing complexity of socioeconomic environments, multicriteria group decision-making (MCGDM) has attracted increasing attention from researchers. Experts’ weight and reliability are crucial to MCGDM and have an important influence on decision-making accuracy. In reality, an expert’s weight and reliability might vary with the influence of factors such as changes in expert psychology and the collection of additional information. Thus, this study proposes a dynamic MCGDM method with automatic reliability and weight calculation. First, we introduce a generalized combination rule into MCGDM and propose methods for automatically determining experts’ weight and reliability by mining evidence. Here, experts’ weight can be calculated according to the entropy of evidence, while experts’ reliability can be calculated according to evidence distance from the perspectives of horizontal comparison and longitudinal comparison. Then, the consensus-reaching process is taken into account in MCGDM; experts are allowed to modify and change their judgments, and experts’ weight and reliability can be automatically updated in each round of interaction. Finally, we provide an illustrative example and make some comparisons to demonstrate the applicability and advantages of the proposed method.