一次查询中的 MySQL 足球统计数据

Question

我有一个看起来像这样的桌子“游戏”：

桌面游戏：

id (int)
home_team (varchar)
away_team (varchar)
home_goals (int)
away_goals (int)

对于特定匹配，我需要获取以下内容：

主队：

home wins
home tie
home loss
home scored
home conceded

因为我还需要知道上述所有实体的总数（胜利、失败、得分等），所以我基本上需要与上述相同，但是当球队在客场比赛时，所以：

away wins
away tie
away loss
away scored
away conceded

客队：

away wins
away tie
away loss
away scored
away conceded

因为我还需要知道以上所有实体的总数（赢、输、得分等），所以我基本上需要和上面一样，但是当球队在主场比赛时，所以：

home wins
home tie
home loss
home scored
home conceded

最好的办法是在一个查询中获得所有这些信息。恐怕这会很困难。所以我可以想象我需要进行多重查询。 目前我有 2 个查询只是为了获取主场球队的数据：

select
count(case when home_goals > away_goals then 1 else null end) as state_win
,count(case when home_goals < away_goals then 1 else null end) as state_loss
,count(case when home_goals = away_goals then 1 else null end) as state_tie
,sum(home_goals) as state_scored
,sum(away_goals) as state_conceded
from
game 
where 
home_team = 'chelsea'

select
count(case when home_goals < away_goals then 1 else null end) as other_win
,count(case when home_goals > away_goals then 1 else null end) as other_loss
,count(case when home_goals = away_goals then 1 else null end) as other_tie
,sum(away_goals) as other_scored
,sum(home_goals) as other_conceded
from
game 
where 
away_team = 'chelsea'

我的计划是将查询 1 中的统计信息与查询 2 相加，以获得总数。 有没有办法只做一次这个查询？我尝试了 union，但我不确定它是正确的方法。

Answer 1

您可以尝试联合在一个查询中结合主客场比赛的统计数据。由于列的名称/语义不同，简单的联合将不起作用。

select
'home' as place
,count(case when home_goals > away_goals then 1 else null end) as win
,count(case when home_goals < away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(home_goals) as scored
,sum(away_goals) as conceded
from
game 
where 
home_team = 'chelsea'
union
select
'away' as place
,count(case when home_goals < away_goals then 1 else null end) as win
,count(case when home_goals > away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(away_goals) as scored
,sum(home_goals) as conceded
from
game 
where 
away_team = 'chelsea'

Ant 使用这个长查询来获取您指定的确切格式的结果：

select 
a.win as state_win
,a.loss as state_loss
,a.tie as state_tie
,a.scored as state_scored
,a.conceded as state_conceded
,h.win as other_win
,h.loss as other_loss
,h.tie as other_tie
,h.scored as other_scored
,h.conceded as other_conceded
from 
(select
'home' as place
,count(case when home_goals > away_goals then 1 else null end) as win
,count(case when home_goals < away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(home_goals) as scored
,sum(away_goals) as conceded
from
game 
where 
home_team = 'chelsea'
union
select
'away' as place
,count(case when home_goals < away_goals then 1 else null end) as win
,count(case when home_goals > away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(away_goals) as scored
,sum(home_goals) as conceded
from
game 
where 
away_team = 'chelsea') h,
(select
'home' as place
,count(case when home_goals > away_goals then 1 else null end) as win
,count(case when home_goals < away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(home_goals) as scored
,sum(away_goals) as conceded
from
game 
where 
home_team = 'chelsea'
union
select
'away' as place
,count(case when home_goals < away_goals then 1 else null end) as win
,count(case when home_goals > away_goals then 1 else null end) as loss
,count(case when home_goals = away_goals then 1 else null end) as tie
,sum(away_goals) as scored
,sum(home_goals) as conceded
from
game 
where 
away_team = 'chelsea') a
where h.place='home' and a.place='away'