Catching Stray Balls: Football, fandom, and the impact on digital discourse

Mark J. Hill · 2025

Read Catching Stray Balls: Football, fandom, and the impact on digital discourse by Mark J. Hill (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-01642.txt.

Introduction

Hill studies how football matches shape talk on Reddit, in a paper from two thousand twenty-five. He works with millions of posts from dozens of online groups. The core claim is that real events move mood across the web, from group to group, not just inside fan spaces.

Hill finds that bad mood tracks with harsh or toxic words. Match results shift both mood and how much people post. Mood can also spill into groups that have nothing to do with the match. The work treats online spaces as linked mood systems. Though football is the case here, the same spread may mark many other kinds of groups.

Catalog record

CREATORMark J. Hill
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-01642

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Catching Stray Balls: Football, fandom, and the impact on digital discourse Mark J. Hill Abstract This paper examines how emotional responses to football matches influence online discourse across digital spaces on Reddit. By analysing millions of posts from dozens of subreddits, it demonstrates that real-world events trigger sentiment shifts that move across communities. It shows that negative sentiment correlates with problematic language; match outcomes directly influence sentiment and posting habits; sentiment can transfer to unrelated communities; and offers insights into the content of this shifting discourse. These findings reveal how digital spaces function not as isolated environments, but as interconnected emotional ecosystems vulnerable to cross-domain contagion triggered by real-world events, contributing to our understanding of the propagation of online toxicity. While football is used as a case-study to computationally measure affective causes and movements, these patterns have implications for understanding online communities broadly. 1 Introduction Football fans are often negatively presented in popular culture due to connotations of hooliganism and far-right allegiances Awan and Zempi 2023 ; London Assembly 2023 . In response, football clubs (FCs), leagues, and fans themselves have moved to exclude those who exhibit these behaviours from matches, and instead foster tolerant atmospheres through initiatives like Kick It Out and the Rainbow Laces campaign. However, while stadiums have become more inclusive, toxicity remains online Murray 2021 ; Kassam 2024 . It is essential to emphasise that problematic supporters — both online and offline — represent a small minority of fans, most of whom engage positively with the sport and its communities Miranda et al. 2024 . Nevertheless, the visibility and impact of this minority warrants attention. Additionally, even among well-intentioned supporters, the tribal nature of sports can occasionally lead to conflict as personal identities become entwined with team allegiances and historic rivalries Sandvoss 2004 ; Porat 2010 ; Parry 2012 ; Cleland et al. 2018 . This complex landscape, with multiple fan identities interacting, has prompted growing scholarly interest in online football communities Rowe et al. 2010 ; Nuttall 2018 ; Woods and Ludvigsen 2022 ; Kaden et al. 2023 . However, much of this research treats online spaces as isolated environments, failing to capture how football discourse can emerge and move across digital contexts. This paper takes a broader view by examining the affectional dynamics of football supporter communities and their movements across online spaces. Specifically, by analysing millions of posts on Reddit it, first, shows that negative sentiment correlates with problematic language; second, match outcomes influence posting habits and sentiment; third, sentiment can transfer to unrelated communities; and fourth, offers insights into the content of these discourses. In doing this it demonstrates a direct relationship between real-world events and online discourse, offering new insights into the propagation and dynamics of online content. 2 Related Work In addition to the investigations into football fandom referenced above, the affective character of football on fans is well established both offline Kerr et al. 2005 and online Wang and Lu 2023 , with even clubs leveraging social media to foster emotional attachment Marques et al. 2018 . While the dominant emotion identified in studies of online fans is negativity, which has been shown to lead to disengagement Kaden et al. 2023 and hostility Fenton et al. 2024 , negativity is not uniformly corrosive — fans of underperforming teams may embrace suffering as part of loyalty Newson et al. 2021 . That is, negative emotions in football can be both alienating and identity-forming, making it a tricky concept to define from the fan’s perspective. Nonetheless, it is common within football communities, and this research looks at how these emotions may impact digital communities. It has been shown that users encountering hostile online discourse experience negative psychological consequences Braghieri et al. 2022 , while disengagement from toxic environments improves well-being Allcott et al. 2020 . However, just as negativity from a fan’s perspective is difficult to conceptualise, so too is framing online negativity which takes various, and often poorly defined, forms, including hate speech, trolling, and incivility Antoci et al. 2016 ; Anderson and Huntington 2017 ; Matamoros-Fernández and Farkas 2021 . In this context "toxicity" has emerged as an umbrella term. Hanscom et al. define toxicity as "interactions directed at an entity designed to be inflammatory" ( Hanscom et al. 2024 ) , with Recuero highlighting how it is enabled by platform affordances ( Recuero 2024 ) . The latter point is of particular interest as we examine how one affordance — inter-subreddit mobility — facilitates emotional spillover across online spaces and influences discourse Papacharissi 2014 . Social media’s networked nature, combined with platform design choices, amplify emotional content Milli et al. 2025 ; Kramer et al. 2014 . Studies have shown that those exposed to emotionally charged social media (both positive and negative) are themselves more likely to express similar sentiments online Ferrara and Yang 2015 ; Brady et al. 2017 . Additionally, negative content has been shown to spread more frequently and faster than positive content Tsugawa and Ohsaki 2015 . This spread has been framed in various ways as "emotional contagion" Goldenberg and Gross 2020 , risking feedback loops that encourage and reinforce incivility Wulczyn et al. 2017 . It is the emergence of, and exposure to, negativity, that this paper investigates. Computational detection of negative and harmful discourse is extensively studied Schmidt and Wiegand 2017 ; Saleem et al. 2017 ; Ayo et al. 2020 ; Jahan and Oussalah 2023 ; Torregrosa et al. 2023 . However, as a subject there are significant challenges, including language ambiguity, contextual dependencies, and data biases Davidson et al. 2017 ; Sap et al. 2019 ; Pavlopoulos et al. 2020 ; Vidgen and Derczynski 2020 . 1 1 1 Despite limitations, tools such as Google Jigsaw’s Perspective AI are being deployed Lees et al. 2022 . Within this complex computational and conceptual landscape, this approach foregrounds the multifaceted nature through which emotions emerge and are expressed online. That is, rather than proposing a method to identify harmful discourse, we aim to track it from a potential source. 3 Data and Methods Our analysis examines 62,384,329 Reddit posts taken between July 2008 and August 2024 from 41 football club subreddits. 2 2 2 Data processing and analysis scripts are available at https://github.com/markjhill/2025-catching-strays. Community sizes vary considerably, with a mean 1,521,569 and median 63,606 posts (Appendix A ). Posts were aligned with historic match results (final scores) using kick-off times plus 120-minutes (allowing for half-time breaks and added time). This covers 20,764 unique games from the top four leagues in English football, both domestic cups, the three European cup competitions, and the Community Shield (Table 1 ). Match Result Posts % Wins 6,477,964 49.6 Draws 2,690,511 20.6 Losses 3,902,686 29.9 Total 13,071,161 100 Table 1: Match-aligned posts overview. Posts were analysed using TweetNLP, a RoBERTa-based sentiment detection model Camacho-collados et al. 2022 . Outputs (categorical probability scores) were normalised into a -1 to +1 scale to aid comparisons across contexts. 3 3 3 Using S=\frac{p-n}{p+neu+n} where S is the sentiment index, p is the proportion of positive, neu is the proportion of neutral, and n is the proportion of negative content. For cross-community analyses an additional dataset of 1,151,726 posts was constructed,

Read the complete text