Misconceptions in physics are persistent and resistant to conventional instruction. This study investigated whether a gamified digital environment facilitates conceptual change compared to a structurally equivalent non-gamified version. Ninety-four secondary school students (Mage = 13.5) completed pre- and post-tests while engaging with eight levels, each targeting a common physics misconception. No overall learning gains were observed. Item-level analyses, however, revealed heterogeneous effects: some misconceptions appeared more responsive to gamified features, particularly when perceptually grounded, whereas others benefited more from the non-gamified format. Correlational trends suggested potential interactions between task complexity and instructional design, though estimates were imprecise due to the small item pool. The findings indicate that gamification is not inherently beneficial, but contingent on the alignment of game mechanics with the epistemic structure of specific misconceptions.