<p>A Pythagorean fuzzy set is not only an extension of intuitionistic fuzzy set, but also an effective tool for more comprehensively reflecting fuzzy information of things, especially when dealing with multi-attribute decision-making problems over wider area, it plays a huge role. It is elaborated that there are some flaws in the existing score functions of Pythagorean fuzzy numbers, and a novel geometric score function and its ranking criterion are proposed by the curved trapezoidal area corresponding to each Pythagorean fuzzy number, and then study the monotonicity of the score function. Secondly, the two discrete Choquet integral ordered weighted average operators are put forward in the the Pythagorean environment, and discussed the basic properties of this integral operator. Finally, a novel evaluation approach is given according to the suggested geometric score function, ranking criterion and discrete Choquet integral ordered weighted average operator, and the availability and advantages of the method are verified in the example of automobile performance evaluation. Two main contributions: one is to establish a novel score function and its sorting rule based on curved trapezoidal area geometric method, which surmounts some disadvantages of existing ranking approachs in a Pythagorean environment, and solves the chaos induced by the popular intuitionistic fuzzy numbers and Pythagorean fuzzy numbers mixed ranking for many years; the other is to establish the discrete Choquet integral ordered weighted average operator by introducing Choquet integral to handle the integration problem of multiple information, and applied them to practical issues in automotive performance evaluation.</p>