Lower is Better? The Effects of Local Latencies on Competitive First-Person Shooter Game Players

Title of the Paper

Lower is Better? The Effects of Local Latencies on Competitive First-Person Shooter Game Players

Paper Information

• Research Domain: Performance and experience evaluation in computer games, impact of system latency on players
• Keywords: FPS, esports, latency, user study, CS:GO, quality of experience, player performance, online gaming, game interaction
• Publication Date: 2021
• Conference Name: CHI Conference on Human Factors in Computing Systems (CHI ’21)
• Location: Yokohama, Japan

Research Background and Problem Statement

• Identified Issues or Challenges:

  1. The specific impact of reduced system latency on esports players' performance remains unclear, particularly within the low latency range (below 125ms).
  2. Existing studies mostly focus on network latency, with limited research on the effects of local system latency on overall game performance.
  3. Differences in latency adaptation between high-level and low-level players are known, but whether these differences extend to competitive players is still uncertain.

• Significance: Low-latency gaming systems can enhance player performance and quality of experience, which is crucial for winning in esports scenarios. This study provides developers with concrete data for optimizing system configurations and assists players in making informed equipment purchase decisions.

• Motivation and Related Work: Previous research has primarily explored the effects of network latency, input device latency, and game configurations on player performance. However, there is limited investigation into the impact of low latency on FPS players in competitive esports settings, and a lack of comprehensive experimental data in real gaming environments.

Proposed Solution

• Proposed Approach: Conduct a user study in a real FPS gaming environment (CS:GO) to measure the impact of local system latency (25 to 125ms) on player performance and quality of experience. Introduce control variables (e.g., weapon types) to analyze latency effects on specific tasks.

• Innovations:

  1. The study focuses on local system latency (not network latency), eliminating confounding factors related to network conditions to purely test the relationship between machine performance and player experience.
  2. Participants conducted experiments in real home settings using high-performance, customized gaming setups, providing a reference for user study designs during the pandemic.
  3. Analyzes the linear relationship between latency and user performance (accuracy and scores) as well as quality of experience (QoE).

• Implementation Steps:

  1. Selection of FPS game: CS:GO was chosen as the test platform, with game settings modified via configuration files to ensure consistent experimental conditions.
  2. Baseline system latency measurement: Minimum system latency (average 25ms) was measured using high-frame-rate cameras.
  3. Experimental design: Five latency conditions (25, 50, 75, 100, 125ms) were set, with additional latency simulated using custom software to mimic different gaming environments.
  4. User study: 43 participants with extensive CS:GO experience were recruited, and their performance (accuracy and scores) and QoE subjective survey data were recorded.
  5. Data analysis: Statistical methods were used to evaluate the impact of latency on different weapon types and overall gaming experience.

Research Findings

• Specific Results Achieved:

  1. Player performance improved linearly with reduced latency: For every 10ms reduction in latency, accuracy increased by approximately 0.6% for AK-47 and 0.9% for AWP, while scores rose by about 1.1 to 1.2 points.
  2. Quality of experience (QoE) also improved linearly with reduced latency: For every 10ms reduction, QoE increased by approximately 0.15 points (out of a total score of 5).

• Advantages Compared to Existing Solutions:

  • The study tested the impact of latency across different weapon types, revealing that latency has a more significant effect on weapons requiring high accuracy (e.g., AWP).
  • Detailed analysis within the low latency range provides guidance for optimizing high-end and mid-range gaming systems.

• Experimental or Evaluation Results:

  1. Linear fitting of the data showed strong correlations, with performance and QoE significantly improving as latency decreased.
  2. High-level players benefited more than low-level players in terms of kill counts and accuracy.
  3. Latency also had a noticeable impact on players' interactive behaviors, such as shooting frequency and character movement frequency.

• Limitations and Future Directions:

  1. The experiment focused on individual players rather than team strategies, leaving team collaboration scenarios unexplored.
  2. Environmental factors such as lighting or chair comfort in home settings may introduce noise.
  3. Female participants and non-professional players were not included, leaving gender and skill-level differences unexamined.
  4. Future research could extend to other game genres (e.g., MOBA and RTS) and device types (e.g., controllers and virtual reality). Additionally, exploring advanced latency compensation techniques for team-based games and broader skill-level players could be valuable.

Conclusion

This study analyzed extensive experimental data from CS:GO players, demonstrating the significant impact of local system latency on competitive game players' performance and quality of experience, particularly within the low latency range where linear improvement was observed. The findings provide critical guidance for developing more efficient esports systems and optimizing player equipment configurations, while establishing a new experimental design paradigm for latency-related research across different domains.