Calculation with GROUP and SUBSELECT not giving right value

I have an issue to get a properly calculated value from my query... I have 3 tables:

Table match:

Table match2player:

Table set:

The following query is throwing the output underneath:

SELECT 
	`match2player`.`idPlayer` AS 'Player',
    COUNT(DISTINCT `match`.`id`) AS 'Matches played',

	SUM(IF(`match2player`.`party`=
	(SELECT set1.`partySetWon` FROM `tennisdb`.`set` set1 WHERE set1.`idMatch` = `set`.`idMatch` GROUP BY set1.`partySetWon` ORDER BY COUNT(set1.`partySetWon`) DESC LIMIT 1)
	,1,0)) AS 'Matches Won',
    
    SUM(IF(`match2player`.`party`=`set`.`partySetWon`, 1, 0)) AS 'Sets Won',
    SUM(IF(`match2player`.`party`<>`set`.`partySetWon`, 1, 0)) AS 'Sets lost'
FROM `tennisdb`.`set`
LEFT JOIN `tennisdb`.`match` ON `match`.`id` = `set`.`idMatch`
LEFT JOIN `tennisdb`.`match2player` ON `match2player`.`idMatch` = `set`.`idMatch`
WHERE `match`.`idEvent` = 6 AND `match`.`status` IN ('completed','awarded')
GROUP BY `match2player`.`idPlayer`
ORDER BY Player;

Player,"Matches played","Matches Won","Sets Won","Sets lost"
8,2,0,1,4
32,2,3,3,3
52,1,0,0,2
53,1,3,2,1
54,2,5,4,1
55,2,3,3,3
56,1,0,1,2
57,1,2,2,0

The column "Matches won" is giving wrong value (too high) and I have no idea how to solve... I played with further groupings, windows, etc., none was delivering expected result...

The subselect shall identify which party won the majority of sets, and resulting the match. This works okay when I on top group over the matches (and not only the player). But I am not looking for a breakdown of individual matches, I want to get the stats over all matches...

Analyze matches in inner query and then sum sets, matches won and lost:

select idplayer, sum(case when setwon > setlost then 1 else 0 end) mwon,
       sum(case when setwon < setlost then 1 else 0 end) mlost,
       sum(setwon) swon, sum(setlost) slost
from (
  select mp.idmatch, idplayer, 
         sum(case party when partysetwon then 1 else 0 end) setwon,  
         sum(case party when partysetwon then 0 else 1 end) setlost  
  from match2player mp 
  join `set` s on s.idmatch = mp.idmatch 
  group by mp.idmatch, idplayer) grp
group by idplayer order by idplayer

dbfiddle demo