The paper argues for an "adequacy-for-purpose" view of model evaluation, where a model's quality is assessed based on its suitability for specific purposes rather than solely on its representational accuracy. It introduces two notions of adequacy (success in a particular use and reliability in a type of use) and highlights how user, methodology, and circumstances influence a model's adequacy.

The author acknowledges funding from the European Research Council, but no specific conflicts of interest related to the research topic are apparent.

This paper presents a valuable perspective on model evaluation, emphasizing the importance of considering a model's fitness for its intended purpose rather than solely focusing on representational accuracy. It offers a well-reasoned argument and introduces helpful conceptual distinctions, such as between different types of adequacy. While it lacks practical implementation details and a mid-level theory, its conceptual contributions warrant a strong rating. No evidence of anti-cheating measures was found.