Collective Intelligence Outperforms Individual Talent: A Case Study in League of Legends

Angelo Josey Caldeira, Sajan Maharjan, Srijoni Majumdar, Evangelos Pournaras · 2025

Read Collective Intelligence Outperforms Individual Talent: A Case Study in League of Legends by Angelo Josey Caldeira, Sajan Maharjan, Srijoni Majumdar, Evangelos Pournaras (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-2506-02706.txt.

Introduction

Angelo Caldeira, Sajan Maharjan, Srijoni Majumdar, and Evangelos Pournaras test whether a well-knit team of middling players beats a squad of stars in League of Legends. The arena is a multiplayer online battle game, where real-time choices, resource trades, and shared information collide. Machine learning and statistics sift large match datasets built for the study. Teamwork is captured both by game metrics and by graph measures of how players interact.

Cooperative groups with modest individual skill outperformed more talented but loosely linked rivals. The topological measures caught cooperative patterns that box scores miss. The authors pitch the game as a clean lab for collective intelligence.

Catalog record

CREATORAngelo Josey Caldeira, Sajan Maharjan, Srijoni Majumdar, Evangelos Pournaras
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-02706

Browse the full library · About the Tower of Babel

Full text

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

Collective Intelligence Outperforms Individual Talent: A Case Study in League of Legends Angelo Josey Caldeira, Sajan Maharjan, Srijoni Majumdar, Evangelos Pournaras Abstract 1 1 footnotetext: Corresponding author: Sajan Maharjan, School of Computing, University of Leeds, Leeds, UK, E-mail: [email protected] Gaming environments are popular testbeds for studying human interactions and behaviors in complex artificial intelligence systems. Particularly, in multiplayer online battle arena (MOBA) games, individuals collaborate in virtual environments of high realism that involves real-time strategic decision-making and trade-offs on resource management, information collection and sharing, team synergy and collective dynamics. This paper explores whether collective intelligence, emerging from cooperative behaviours exhibited by a group of individuals, who are not necessarily skillful but effectively engage in collaborative problem-solving tasks, exceeds individual intelligence observed within skillful individuals. This is shown via a case study in League of Legends , using machine learning algorithms and statistical methods applied to large-scale data collected for the same purpose. By modelling systematically game-specific metrics but also new game-agnostic topological and graph spectra measures of cooperative interactions, we demonstrate compelling insights about the superior performance of collective intelligence. 1 Introduction The recent advancements in computing technology have evolved the way individuals socialize and interact; online gaming platforms being one of the most prominent mediums for social interaction and cooperation. Zhong reports that in China there are 147 million people playing online games, thus representing new forms of community, social interaction, and collaboration [ 1 ] . These online gaming environments are popular testbeds for studying human interactions and behaviours in complex artificial intelligence systems. The simulation of diverse scenarios ranging from strategic decision-making to adaptive behaviours within online gaming environments allows the testing of complex real-world mechanisms under controlled low-cost conditions. Furthermore, the virtual worlds of such online games are populated with autonomous artificial agents and other human players, thus offering opportunities for interesting interactions that guide players and enhance social dynamics [ 2 ] , resulting in a collective intelligence derived from the synergistic cooperation of humans and computing agents. Collective intelligence is the general ability of a particular group to perform well across a wide range of different tasks [ 3 ] . In the field of computer science, it involves the study of how groups of one person and one computer interact (human-computer interaction) as well as how larger groups of people interact with computing technologies to cooperate and make intelligent decisions, for instance, in the development of collaborative software such as Wikipedia [ 4 ] . Collective intelligence is prominently demonstrated in many online gaming environments among which multiplayer online battle arena (MOBA) games serve as a notable example. MOBA games are a subgenre of real-time strategy games in which two teams, each consisting of 5 individual players, compete against each other with each player controlling a single character (called a champion ), while strategy revolves around individual character development and cooperative team play in combat [ 5 ] . Riot Games’ League of Legends is the most popular MOBA game played worldwide with over 150 million monthly active players [ 6 ] . League of Legends features 169 uniquely playable champions with their own set of skills, that relies on an individual player’s micro-management ability, paired with a strong understanding of champion mechanics along with the collective effort of a team to think strategically and showcase team-work abilities. The game is structured around a map called Summoner’s Rift , and the primary objective of the game is to destroy the opponent team’s base, a tower structure called the Nexus , while the game progresses with champion development via earning gold, gaining experience, crafting items and scoring kills. The collaborative dynamics observed among randomly assigned players within a team, each controlling a distinct champion with unique sets of abilities and working towards the shared objective of winning the game, potentially exemplifies the presence of collective intelligence in MOBA games. In fact, exchange of in-games chat messages , signals and assists to secure game objectives elicits a quantitative measure of collective intelligence. Leavitt et al. show that non-verbal communication in the form of in-game pings improves team strategy and performance [ 7 ] . On the other hand, individual players can excel in executing their champion mechanics and abilities, thus displaying high levels of individual intelligence. Using in-game data to model various metrics of individual and collective intelligence, this paper aims to outline the importance of such individual as well as team-based collective intelligence metrics in influencing the outcome of a game. That is, it tests the hypothesis H1: collective intelligence outperforms individual intelligence of the players in winning a game . Additionally, Riot Games maintains a two-tiered in-game ranking system (called tier and division ) that establishes a hierarchy of players based on their recent in-game performances. Using such information, this paper also tests another hypothesis, that is, H2: Highly-skilled players exhibit higher levels of collective intelligence compared to less-skilled players . The main contributions of this work are – (i) a framework to evaluate individual and collective intelligence of players in League of Legends using in-game metrics and machine learning algorithms, and (ii) the availability of dataset and source code that can be used to reproduce the results. 2 Related Work Collective intelligence can be observed in human-computer interaction, collective-decision making and collaborative software development; in the decentralized collective learning of self-managed economies [ 8 ] , swarm intelligence in the large-scale coordination of drones [ 9 ] , etc. Woolley et al. first provided the evidence for a collective intelligence factor, c in human groups, suggesting that c is not correlated with the average or maximum individual intelligence of the group members, but rather it emerges from the way group members interact [ 10 ] . Krause et al. conclude that the potential benefits of human swarm intelligence depend on the type of problem, while advocating for diversity in groups, and that high-performing individuals can be outcompeted by a similar size group of low-performing individuals [ 11 ] . Likewise, Santos et al. argue that social diversity catalyzes cooperative behaviour [ 12 ] . The sports and gaming industries have integrated collective intelligence paradigms to solve problems and generate insights beyond the individual capabilities of a single player. Duch et al. quantified the contribution of individual players towards the overall team performance during the 2008 European Cup football tournament via network graphs that modeled the flow of passes among players in a team [ 13 ] . Grund confirmed that centrality of interaction between players leads to a decreased team performance by analyzing soccer passes of teams in the English Premier League [ 14 ] . Szabo et al. investigated team behavioral dynamics in escape room settings and reported that more effective teams tend to solve problems via coordinated, balanced communication patterns, maintaining a dynamic alternation between different types of tasks [ 15 ] . In the realm of digital gaming, 42 Entertainment launched I Love Bees , an alternate reality game used for the marketing of Halo 2 in which over 600,000 players collaborated in solving puzzles related to the game prior to its release [ 16 ] . Gonzalez-Pardo et al. conducted an empirical study describing how collective intelligence

Read the complete text