How can I query unique columns, conditionally choosing which row based on another column's value?

Question

I have the following setup:

date | event | hits

Jan  |   A   |  0
Jan  |   B   |  2
Jan  |   C   |  0
Feb  |   A   |  4
Feb  |   B   |  0
Feb  |   C   |  0

And I'm looking for a query that returns:

• Events are unique (only return [Jan | B | x] or [Feb | B | x] , not both)
• Prioritize hits (so [Jan | B | 2] beats [Feb | B | 0] )
• Secondarily prioritize most recent date ( [Feb | C | 0] beats [Jan | C | 0] )

I would want the query for the above table to return:

Feb | A | 4
Jan | B | 2
Feb | C | 0

Answer 1

SELECT 
(select edate from stuff sx where s.event=sx.`Event`  ORDER BY sx.hits DESC,sx.edate DESC limit 1) date,
s.event, (select hits from stuff sx where s.event=sx.`Event`  ORDER BY sx.hits DESC,sx.edate  DESC limit 1) hits
 FROM stuff s
GROUP BY Event
order by event

Table stuff edate field is defined as date. Not very efficient for large tables, but works.