I want to create a prediction table that lists each of the 10 teams playing every other team at home and away. How can I generate this table?

Create a temporary table or CTE that holds the team list, then cross‑join it with itself to get all combinations, filtering out self‑matches. Finally, insert the result into a prediction table. Example in PostgreSQL:

-- Create prediction table
CREATE TABLE predictions (
  home_team_id INT,
  away_team_id INT,
  predicted_score_home INT,
  predicted_score_away INT
);

-- Generate all home‑away pairs
INSERT INTO predictions (home_team_id, away_team_id)
SELECT t1.team_id, t2.team_id
FROM teams t1
JOIN teams t2 ON t1.team_id <> t2.team_id;

If you need both directions (home and away) for each pair, just run the same insert again or use UNION ALL. The resulting table will have 10 × 9 = 90 rows. Complexity is O(n²).