Beyond Elo: Improving Matchmaking Using the Team Player Effect

Nico Elbert, Victor Klockmann, Fabian Kosse, Alicia von Schenk, Nikolai Stein & Christoph M. Flath, 2026,
IEEE Transactions on Games,
[Published Online: 28 April, 2026]

Abstract:
Effective matchmaking in online multiplayer games is essential for ensuring balanced competition and player satisfaction. Traditional rating systems primarily assess individual skill but overlook social and coordination-based factors that influence team performance. This study adapts the team player effect (TPE)-a residual-based metric capturing a player’s consistent contribution to teamwork beyond technical skill-and evaluates its usefulness in ranked team play. Building on prior work that introduced a TPE-based approach to matchmaking, the present analysis leverages a large-scale dataset from Age of Empires II: Definitive Edition. Incorporating TPE and team familiarity significantly improves match outcome prediction over Elo-based benchmarks, with the influence of TPE increasing in larger teams. This extended manuscript adds two contributions. First, the behavioral meaning of the proposed imbalance measure is validated by showing that lower predicted imbalance is associated with longer realized match duration. Second, the operational implications of TPE-informed predictions are assessed in a controlled matchmaking simulation, demonstrating reductions in waiting times and improvements in realized match imbalance under standard imbalance constraints. These results suggest that coordination-related metrics can enhance both predictive accuracy and system-level performance in team-based matchmaking.