小弟在做HackerRank的題目時
遇到這題
https://www.hackerrank.com/challenges/challenges/problem
我是寫出解法了
寫法如下

select c1.hacker_id as h_id,name,count(c1.hacker_id) as cnt from Hackers h1
join Challenges c1 using(hacker_id)
group by c1.hacker_id,name
having count(c1.hacker_id)=(
    select max(challenges_created) 
    from (
        select count(hacker_id) as challenges_created 
        from Challenges
        group by hacker_id
    ) g
)
union
select c2.hacker_id as h_id,name,count(c2.hacker_id) as cnt from Hackers h2
join Challenges c2 using(hacker_id)
group by c2.hacker_id,name
having cnt not in (
    select challenges_created
    from (
        select count(hacker_id) as challenges_created 
        from Challenges
        group by hacker_id
    ) g1
    group by challenges_created 
    having count(challenges_created) > 1
)
order by cnt DESC, h_id

可是有一點我不明白
我原本卡在union的第二個query
我那裏的having 原本寫的是 count(c2.hacker_id)
但他跳出來的結果一直會是重複的值(題目要找不重複的)
所以就會錯
我不懂為什麼我把alias接在having後面就可以了

請教各位大大 是別名在having 上的差別 或是 跟count有關?