計算查詢中的分鍾數

Question

我正在尋找一種方法來計算我團隊中每個球員（或這里只有一個）的上場時間。 簡化后的數據庫表如下：

matchid    action    minute    player
-------------------------------------
1          subbedin  30        Pele 
2          starter             Pele
2          subbedout 50        Pele
3          subbedin  70        Pele
3          red       80        Pele
4          starter             Pele

我現在對其他統計數據的查詢：

$query = mysql_query("SELECT *,
  SUM(CASE WHEN action = 'starter' OR action = 'subbedin' THEN 1 ELSE 0 END) AS games,
  SUM(CASE WHEN action = 'goal' OR action = 'pengoal' THEN 1 ELSE 0 END) AS goals,
  SUM(CASE WHEN action = 'yellow' THEN 1 ELSE 0 END) AS yellows,
  SUM(CASE WHEN action = 'red' THEN 1 ELSE 0 END) AS reds,
  // MINS GOES HERE
FROM league2012
GROUP BY player");

對於每個 matchid，基本計算是

( 90 OR subbedout OR red ) - ( starter OR subbedin )

例如在比賽 2

subbedout (50) - starter (0) = 50

最后，該表應如下所示：

player    minutes    goals, cards, etc.
---------------------------------------
Pele      210        ...

在過去的一個小時里，我一直在學習教程，但似乎不知道該怎么做。

Answer 1

  sum
    (
      case action
        when 'subbedin'  then 90 - minute
        when 'starter'   then 90
        when 'subbedout' then minute - 90
        when 'red'       then minute - 90
      end
    ) as minutes

Answer 2

我會先計算每個球員在每場比賽中的上場時間，然后將它們加起來得到每個球員的總分。 要將獲得的結果與您正在計算的其他統計數據相結合，我認為除了以相同的方式計算其他統計數據之外別無他法，即首先是每個球員和比賽，然后是每個球員。 這就是我的意思：

SELECT
  player,
  SUM(games) AS games,
  SUM(goals) AS goals,
  SUM(yellows) AS yellows,
  SUM(reds) AS reds,
  SUM(minutesplayed) AS minutesplayed
FROM (
  SELECT
    player,
    matchid,
    SUM(CASE WHEN action IN ('starter', 'subbedin') THEN 1 ELSE 0 END) AS games,
    SUM(CASE WHEN action IN ('goal', 'pengoal') THEN 1 ELSE 0 END) AS goals,
    SUM(CASE WHEN action = 'yellow' THEN 1 ELSE 0 END) AS yellows,
    SUM(CASE WHEN action = 'red' THEN 1 ELSE 0 END) AS reds,
    IFNULL(SUM(CASE WHEN action IN ('subbedout', 'red') THEN minute END), 90)
    - IFNULL(SUM(CASE WHEN action = ('subbedin') THEN minute END), 0) AS minutesplayed
  FROM league2012
  GROUP BY
    player,
    matchid
) s
GROUP BY
  player