Structural Analysis of Manzai Scripts

Extracting Patterns Associated with Higher Ratings in the M-1 Grand Prix

Abstract

This study analyzed 50 manzai routines from the 2015–2019 M-1 Grand Prix to identify structural patterns associated with higher audience evaluations. Transcripts were segmented into functional units, and sequential patterns were compared using length-normalized N-gram frequencies. High-rated routines more often resolved a boke (a deliberate deviation) quickly with a subsequent tsukkomi (an explicit corrective call-out) whereas low-rated routines exhibited prolonged set-up and additive or repetitive boke. A sensitivity analysis that collapsed consecutive identical labels (run-length encoding) yielded the same qualitative result: the advantage of short cycles that insert tsukkomi promptly persisted, while differences attributable to long identical runs diminished. These findings indicate that maintaining a short lag from boke to tsukkomi, and avoiding extended stretches of set-up or repeated boke, characterizes routines that receive higher evaluations.