Academic research in games should be challenging our need for commonly accepted principles like Skinner boxes or certain narrative structures. Historically it tends to be focused on finding only certain types of quantifiable data, which can be distracting from the medium's overall progress.