Elo Ranking Methodology

The present study aims to estimate the strength of teams participating in tournaments, and subsequently the strength of the tournaments themselves, using the well-established Elo ranking system. The methodology employed in processing, transforming, and analyzing the gameplay data is outlined below.

1. Data Retrieval

• Initial data was sourced from tournaments.json.
  • teams: Two team IDs participating in the game.
  • winner: ID of the winning team.
  • league: ID of the league under which the game was played.
  • tournament: ID of the tournament.
  • section: Section (e.g., "Group A") of the game.
  • startDate: Start date of the game.

2. Elo Rating for Teams

• Each team received an Elo rating based on game performance.
  • Initial rating for all teams: 1000
  • Expected outcome was calculated for both teams.
  • Ratings were updated post-game.
  • K-factor values varied based on league and section.

3. Tournament Strength Calculation

• Tournament strength was calculated by averaging the Elo ratings of all participating teams.

4. Elo Rating Based on Tournament Strength

• Elo ranking was re-initiated with the tournament strength metric.
  • Teams initialized with scores from their tournament history.
  • Expected outcome recalculated for each game.
  • K-factor values varied based on tournament and section.

5. Global Rankings

• A global power ranking order was established from the Elo ratings.
  • Ranking was exported on a table on Amplify.

7. Tournament rankings and Team Rankings

• The routnament rankings and team rankings are derived from the global ranking in React.

In conclusion, this methodology offered a comprehensive analysis of team and tournament strengths, giving stakeholders an all-encompassing ranking system that encapsulates the multifaceted nature of competitive gameplay.