Random Problems with Player Role Rating Calculations and a Possible Cause

[Surrational0]

I think I may understand why player roles %'s are miscalculated but only sometimes for some players (FMRTE 22.4.1).  This happens only for roles when a player has more than one position that could have that same role.  For instance, the Roaming Playmaker role could be a MC or a DM.

 

I have seen MC players that should be roughly a ~65% Roaming Playmaker and show a correct 65% Mezzala role score.

• If the player shows very low Roaming Playmaker score (3%) instead of the correct score they will have DM ability of 1 and an MC ability of 20
• If the player shows low Roaming Playmaker score (~55%) they will have DM ability of 16 and an MC ability of 20
• If the player shows the correct Roaming Playmaker score (~65%) they will have DM ability of 20 and an MC ability of 20, sometimes, but sometimes they might have a lower DM ability even 1!

 

What causes the problem?  We can get a hint when looking at it in the opposite way.  Sometimes players with the DM of 1 and MC of 20 will still show the correct Roaming Playmaker score sometimes they will have a wrong score.  But a player with 20 DM and 20 MC always shows the correct role score. 

 

Likely, then, the Roaming Playmaker score is calculated twice in the code: once for the DM and once for the MC position.  Then when the resulting data is merged for the interface a random score from the two is chosen by the datatable code.  Maybe whichever was (randomly) inserted second?

 

This weirdness is especially noticeable for the Wingback role.  Which makes sense with the above theory because a Wingback role has 4 associated positions DL/DR/WBL/WBR so it has more chances to mess with the calculation.

[culturedleftfoot]

@Surrational0 thanks for the above input👍