CASE statement not working as it should in SQL - mysql

Question

SELECT userid,
(SELECT CASE WHEN (COUNT(CASE WHEN casinowagers != 0
THEN 1
ELSE null
END)
+ COUNT(CASE WHEN depositmade_amt != 0
THEN 1
ELSE null
END)) >= 3
AND (Round(sum(totalhold - playercomps - freemoney - (depositmade_amt*.1)),2)) >= 10
THEN "VIP"
ELSE "NON-VIP"
END as VIPcheck
FROM player_activity
WHERE YEAR(txndate) = YEAR(CURRENT_DATE - INTERVAL 3 MONTH)
AND MONTH(txndate) = MONTH(CURRENT_DATE - INTERVAL 3 MONTH)
) as vipMonthStatus,
(COUNT(CASE WHEN casinowagers != 0
THEN 1
ELSE null
END)
+ COUNT(CASE WHEN depositmade_amt != 0
THEN 1
ELSE null
END)) as activityCount,
(Round(sum(totalhold - playercomps - freemoney - (depositmade_amt*.1)),2)) as Value,
FROM player_activity
WHERE userid = 2023410
GROUP BY year(txndate),month(txndate)
LIMIT 1000
So basically the vipMonth status is always returning as "VIP". However, for the month of May (because of the below "where" statement), it should be "non-vip" because only 1 deposit and 1 wager were made.
What gives?
( WHERE YEAR(txndate) = YEAR(CURRENT_DATE - INTERVAL 3 MONTH)
AND MONTH(txndate) = MONTH(CURRENT_DATE - INTERVAL 3 MONTH) )

Answer 1

You need to tie the player_activity table in the vipMonthStatus subquery to the player_activity table in the main query.
First alias the "inner" player_activity and the "outer" player_activity tables so that you can tell them apart. Something like in_player_activity and out_player_activity.
Then you need to add an additional WHERE statement to the vipMonthStatus subquery to restrict the records in the inner table to the associated records in the outer table based on userid . It would end up looking like this:
FROM player_activity in_player_activity
WHERE YEAR(in_player_activity.txndate) = YEAR(CURRENT_DATE - INTERVAL 3 MONTH)
AND MONTH(in_player_activity.txndate) = MONTH(CURRENT_DATE - INTERVAL 3 MONTH)
AND in_player_activity.userid = out_player_activity.userid
Keep in mind now that you have introduced table aliases, you will need to add the alias prefix to all of the other column names. Personally I would use something more concise like opa and ipa.