Previous studies suggested that spoilers might increase the enjoyment of novels. However, the problem of spoilers has not been sufficiently clarified. The objective of our work is to clarify the effect of comic spoilers and to apply clarified knowledge for applications. In our past work, we constructed a spoiler dataset and investigated the spoilers’ effect by changing the spoiler timing for readers. However, in the past work, we could not clarify the characteristics of spoilers. In this work, we clarified that spoilers reduced the interest in continuing reading the comics and analyzed the characteristics of spoilers by using the dataset. Then, we considered how to construct a comic spoiler dataset and investigated how to determine the spoiling pages automatically using image processing, and character detection and so on.