Skill-based matchmaking aims to ensure fair play in competitive games. However, recent research indicates this approach may have an unintended downside, as matches between similarly skilled players can lead to sequences of defeats that discourage continued participation.

The study, featured in Management Science, suggests that effective matchmaking should consider more than just player ability. It should factor in responses to recent results and overall patterns. Examining 5.4 million Lichess games revealed that refined systems raised player involvement by 4% to 6% over traditional skill-based methods.

The conventional skill-based matchmaking framework now appears insufficient for services focused on sustaining user activity.

Why fair matches can backfire

SBMM connects players of comparable skill to produce even contests. In theory, this seems ideal. No one enjoys repeated defeats against superior opponents, and little is gained from dominating novices.

The issue lies in patterns over time. One defeat is normal, yet multiple consecutive losses may leave a player feeling stuck in a cycle of disappointment.

This behavioral element is often overlooked by standard systems. Every outcome influences whether a user starts another match, pauses, or exits the game entirely.

How much engagement is at stake

The Lichess findings underscore the impact. Across 5.4 million games, the improved approach raised engagement by 4% to 6% compared to typical skill-based matchmaking. In modeled cases, improvements reached up to 50%.

Players would notice only minor shifts. An enhanced system would incorporate recent results into its calculations and focus on the full session rather than ratings alone.

For developers, these gains matter. With the global gaming sector expected to reach nearly $188 billion yearly per the Global Games Market Report, even small retention boosts translate to significant value.

When retention efforts go too far

Advanced matchmaking does not allow unchecked manipulation of matches. It introduces concerns about trust, since competitive fairness and retention aims may conflict as systems grow more complex.

This conflict intensifies with pay-to-win elements. The research noted that purchased advantages can sometimes increase involvement by altering skill distributions, yet this is not always beneficial. Higher engagement does not guarantee a superior experience.

The risk for companies lies in opaque systems that erode player confidence. Moving past pure SBMM requires demonstrating that matches still honor fair competition. Future matchmaking must encourage play without appearing to control outcomes against users.