The use of instructional videos is rampant in education; however, their interaction is limited by weak instructional design. Gagne has never insisted on using his renowned 9 Events of Instruction slavishly in situations as a viable paradigm for utilization in video design. Connecting grade level, video length, and interaction, this study seeks to determine the relevance of Gagne's prescribed 9 event sequence in instructional mathematics videos. We scrutinized 50 instructional mathematics videos on YouTube geared towards middle school pupils ranging between 5th and 8th grades. We used quantitative media content analysis for video analysis. In data analysis, partial least squares were used. Bayesian estimation was also resorted to for cross checking. The data revealed that one-third of Gagne's instructional design steps were not always present: "activating prior knowledge," "eliciting performance," and finally "providing feedback." A mediation analysis between grade level and video length revealed that 6 events fully mediated the association between the two. We also elicited the impact of these variables on affective and behavioral interactions in videos. This study assists in creating an idiosyncratic instructional design model, called "Birgili's 8 steps for instructional video design," and in infusing this with a melange of four theories. In contrast with the status quo attesting that the literature abounds with scholarly works touting "the shorter is the better" mantra, the results substantiated that longer may be better in leveraging video interactions provided that the length is judiciously used to conform to instructional design principles.