查找具有特定条件的最大值并检查同一组是否有更高/更低或任何值

Question

I am trying to get game... and max score for user and information is there higher/lower value(score of other players) than user's max value. I will explain situation.

User should get his games and max score for each game WHERE game is active. That works fine with JOIN and GROUP BY but I have no idea how to get information is there more scores and are they higher or lower than users max score.

Table 1: scores

+----+--------+------+--------+-------+
| id | gameID | game | player | score |
+----+--------+------+--------+-------+
|  1 |      1 | pang | lea    |    50 |
|  2 |      1 | pang | lea    |    60 |
|  3 |      1 | pang | paola  |    70 |
|  4 |      2 | pong | lea    |   100 |
|  5 |      2 | pong | paola  |    90 |
+----+--------+------+--------+-------+

Table 2: games

+----+------+--------+
| id | name | active |
+----+------+--------+
|  1 | pang | yes    |
|  2 | pong | yes    |
|  3 | pung | yes    |
+----+------+--------+

Code:

$loggedUser = 'lea';
$sql = 
    "
SELECT s.gameID
     , s.game
     , COUNT(s.id) c
     , MAX(s.score) Max
  FROM scores s
 RIGHT 
  JOIN games g
    ON s.gameID = g.id
 WHERE g.active = 'yes' 
   AND s.player = '$loggedUser'
 GROUP 
    BY s.gameID 
 ORDER 
    BY c
";

Output should look like this:

Hello Lea, your top scores:

pang played x 2, max score 60 pts (top score is 70)....

pong played x 1, 100 pts , you have higher score....

Answer 1

In MySQL 8.0, you can use window functions to:

• rank the scores of a player for each game: ROW_NUMBER() OVER(PARTITION BY player, game ORDER BY score DESC)
• count how many times a player played each game: COUNT(*) OVER(PARTITION BY player, game)
• compute the top score of a given game amongst all players : MAX(score) OVER(PARTITION BY game ORDER BY score DESC)

Consider the following query:

SELECT game_name, cnt games_played, score max_score, top_score
FROM (
    SELECT
        g.name game_name,
        s.player,
        s.score,
        ROW_NUMBER() OVER(PARTITION BY player, game ORDER BY score DESC) rn,
        COUNT(*) OVER(PARTITION BY player, game) cnt,
        MAX(score) OVER(PARTITION BY game ORDER BY score DESC) top_score
    FROM scores s
    INNER JOIN games g ON g.id = s.gameID AND g.active = 'yes'
) x WHERE rn = 1 AND player = 'lea';

In this db fiddle with your sample data, the query yields:

| game_name | games_played | max_score | top_score |
| --------- | ------------ | --------- | --------- |
| pang      | 2            | 60        | 70        |
| pong      | 1            | 100       | 100       |

---

In earlier versions of MySQL, one solution would be to use an aggregated subquery to compute the top score for each game amongst all players, and the JOIN it with the main aggregated query:

SELECT g.name game_name, COUNT(*) games_played, MAX(s.score) max_score, ts.top_score
FROM  scores s
INNER JOIN games g 
    ON g.id = s.gameID AND g.active = 'yes'
INNER JOIN (SELECT gameID, MAX(score) top_score FROM scores GROUP BY gameID) ts 
    ON s.gameID = ts.gameID
WHERE s.player = 'lea'
GROUP BY s.player, g.id, g.name;

Demo on DB Fiddle .