Will They or Won't They Couples Data

Many television shows follow the “will they or won’t they” trope, where the dynamic between a pair of main characters constantly shifts between friendship and something more throughout the run of the series. This trope has persisted throughout the decades, and examples include Sam and Diane from the 1980s show Cheers and Jess and Nick from the 2010s show New Girl. In some cases, the audience may wait multiple seasons before a couple like this gets together, and some suspect that producers delay the moment to create suspense and keep viewers engaged. Events marking major romantic milestones, such as the pair’s first kiss, often change the trajectory of the plot, influence the number of viewers tuning into the show, and drive up episode ratings. In this project, we scrape viewer ratings from the Internet Movie Database (IMDb) for 150 popular couples from 125 television series and then model the plot shifts following episodes with romantic milestones using causal inference methods. Specifically, we construct an interrupted time series model, where the interruption is the episode in which each couple has their first kiss. From this model, we assess whether these interruptions are associated with changes in viewer ratings on average.

诸多电视剧都采用了“爱情走向成谜”的叙事桥段：一对主角之间的关系张力会在整部剧集的播出周期内，于友情与暧昧情愫之间反复摇摆。该叙事桥段数十年来长盛不衰，典型案例包括1980年代剧集《干杯酒吧》（Cheers）中的山姆与戴安，以及2010年代剧集《杰茜驾到》（New Girl）中的杰茜与尼克。部分观众往往需要追完多季才能等到这类情侣最终确定关系，有观点认为制片方刻意延后这一关键节点，以制造悬念、维持观众的观剧粘性。这类标志性浪漫节点（如情侣的初吻）往往会改变剧集的剧情走向，影响剧集的收看人数，并推高单集收视率。本研究从互联网电影数据库（Internet Movie Database, IMDb）中爬取了125部剧集里150对热门情侣的观众评分数据，随后采用因果推断方法，对包含浪漫节点的剧集播出后的剧情变化进行建模。具体而言，我们构建了断点时间序列模型，以每对情侣初吻所在的剧集作为断点，基于该模型评估这类断点是否会整体上带来观众评分的变化。