When video models lack shot-type instructions, they tend to default to the safe choice of a medium shot, resulting in a lack of visual rhythm variation in the footage.