{
    "action": "final",
    "content": "## Distinct Value Counts for Categorical Columns\n\nThe dataset contains 21 categorical columns with varying cardinality:\n\n### High-Cardinality Columns (>100 distinct values)\n- **team1_players**: 1,225 distinct values — comma-separated player lists for team 1\n- **team2_players**: 1,223 distinct values — comma-separated player lists for team 2\n- **player_of_match**: 322 distinct values — individual players awarded man-of-the-match\n\n### Medium-Cardinality Columns (20-100 distinct values)\n- **tv_umpire**: 85 distinct values\n- **reserve_umpire**: 82 distinct values\n- **umpire1**: 76 distinct values\n- **umpire2**: 71 distinct values\n- **venue**: 60 distinct values — stadiums where matches were played\n- **city**: 38 distinct values — cities hosting matches\n- **match_referee**: 32 distinct values\n\n### Low-Cardinality Columns (2-20 distinct values)\n- **winner**: 20 distinct values — teams that won matches\n- **toss_winner**: 19 distinct values\n- **team2**: 19 distinct values\n- **team1**: 19 distinct values\n- **season**: 19 distinct values — IPL seasons from 2007/08 to 2026\n- **result_type**: 3 distinct values (complete, tie, etc.)\n- **toss_decision**: 2 distinct values (bat or field)\n\n### Single-Value Columns (constants)\n- **team_type**: 1 distinct value (club)\n- **match_type**: 1 distinct value (T20)\n- **gender**: 1 distinct value (male)\n- **event_name**: 1 distinct value (Indian Premier League)\n\nThe query counted distinct values across all 1,243 matches in the dataset. The player roster columns (team1_players, team2_players) have the highest cardinality because they contain comma-separated lists of 11+ players per match, creating unique combinations across different match lineups."
}