An empirical investigation into the effects of sleep loss on esports performance

Esports (competitive, organised video game play) has risen from obscurity to rival and/or surpass many traditional sports in terms of popularity, viewership, and earnings. As a result, human factors are beginning to be explored in the context of esports, ultimately with the same goals that are pertinent in much traditional sport research; to augment performance, or minimise performance loss. One human factor which has drawn attention within esports literature and practice is sleep. This is largely due to the substantial cognitive demands of esports, combined with the wealth of research linking sleep loss to impeded cognitive performance. Nonetheless, the relationship between sleep loss and esports performance has not been formally investigated to date; this current thesis aims to address this gap in scientific knowledge. Chapter two systematically explores the current scientific literature on how acute sleep restriction impacts the cognitive performance specifically for individuals who engage in cognitively demanding tasks with critical or safety-critical outcomes in their occupation or area of expertise (Elite Cognitive Performers). This chapter finds simple cognitive tasks to be most susceptible to sleep loss induced performance hindrance, however performance on complex tasks demanding cognitive flexibility (e.g. task-switching, a cognitive ability deemed highly relevant to esports) also appears potentially sensitive to sleep loss. Chapter three examines the test-retest reliability and presence of practice effects for a shortened version of the Category Switch Task, a task-switching paradigm with unpredictable switches, which allows for the assessment of cognitive performance on a complex task with and without cognitive flexibility demands. Chapter four provides an introduction to the esport Rocket League, which is the target esport within the current thesis. Chapter five outlines the identification of performance and rank indicators in the esport Rocket League through use of machine learning methods on a large dataset of in-game data. Performance indicators outlined are metrics targeted within later exploratory analysis on sleep loss and its impact on in-game Rocket League performance. Chapter six outlines key methodological details about the sleep measurement methods and analytical approach used in the subsequent chapter. It includes a bespoke simple imputation approach to deal with missing actigraphy-derived sleep data, which I show to outperform other simple imputation approaches. Chapter seven outlines a study exploring how experimentally induced total sleep deprivation impacts the cognitive and in-game performance of esport players. Cognitive tasks include the Psychomotor Vigilance Task and Category Switch Task, and the esport targeted was Rocket League; chosen due to various properties lending itself strongly to experimental research, as well as access to performance indicators (from chapter five) allowing for analytical depth. I find the overall in-game performance of Rocket League players to not change following ~29 hours of total sleep deprivation, despite increases in sleepiness, and decreases in alertness, motivation, and cognitive performance, immediately prior to esport play. Further exploratory analysis suggests that sleep deprived players may have adopted a simpler or safer (or both) playstyle. Chapter eight combines the findings of chapter seven with expert opinion from professional players, coaches, and analysts, to explore this playstyle change. In this chapter, I find that simpler and safer playstyles are very much analogous within Rocket League, helping to contextualise my findings with previous sleep loss and decision making literature. Collectively, the chapters within the current thesis provide novel insights into how sleep loss impacts in-game performance within esports, providing further evidence and discourse toward the topic of performance optimisation in esports.