Introduction

To be honest, post-game interviews do not really matter, right? Coaches select a bunch of phrases from their repertoire, only deciding which to pick based on whether they got the win or not. Not too many insights to gather, and, who reads those statements by the coaches from their press conferences? Yet this short article is about post-game interviews by coaches in the Basketball Euroleague (Europe’s best cross-country competition). How comes so? I argue, that while the single interview (i.e. released as a press statement) you may likely not read, reading all interviews from the past ten seasons might be more interesting. Maybe there are larger-scale patterns? What are the most common words or phrases? How do coaches differ in their language? How do post-game interview language differ based on whether your team won or lost? To try answering all of these questions, I scraped data of post-game interviews from the Basketball Euroleague website. In total, 4641 post-game statements: 496781 words. While this would take days to be read, computational text analysis does it within seconds, remembers every word, and can provide me descriptive insights on the questions sketched above, and more. The dataset I created (that is available here) then contains variables on the coach, the team, whether the game was won or lost and by what margin, the season, the round (regular season gameday, playoffs, Final Four), and the interview statement itself.

Within the following sections, I will descriptively cover some of the aspects and questions I sketched out above in more of an exploratory data analysis. To do so, simple word-frequency counting, sentiment analysis as well as (tf-idf) are applied. To start things out, let’s count some words…

Common words