The reason why I think the current attributes system needs to change.

[a5864003]

When the production team sets the players' attributes, I don't know if they are using the editor we are using or if they have their own tools. Currently, the players' attributes are expressed on a 1-20 scale. Of course, we know that there are decimal capabilities such as 15.2 or 16.7 between 1-20. But if they use the editor we use, all the numbers below the decimal point are meaningless. This is because the editor can only input integers. That will reduce the discrimination of attributes between players. Messi's dribbling ability is 20. Suppose there are 10 players who show better dribbles than him right now. It's unrealistic, but they have a better dribbling success rate than Messi, and Messi's dribbling skills remain. They should also be given a dribble ability of 20. Then, what will Messi do with his abilities? It will probably turn into 19. If so, all players' dribbling abilities will also be lowered by 1. Then, what should I do with a player whose dribbling ability was 1? A player with a dribbling ability of 2 will have the same dribbling ability as a single player. Of course, this problem occurs even when the scale is increased. For example, even if 0-80 is used, such as baseball, there are players with the lowest ability. However, in accepting such changes, 0-80 is more receptive than 0-20. It will be more acceptable if it is on a larger scale rather than 0-80. Therefore, 0-20 must be changed to a larger scale.
If the producers use a different editor than us, please make it editable in the editor we use.

[Matada]

0-20 is only for display, attributes are from 0 to 200, issue is in editor, not in this

Edited November 17, 2021 by Matada

[santy001]

As researchers we have a vastly superior set-up than the pre-game editor, but we still only enter attributes as single, whole, integers. 

The decimalisation of attributes in the background allows for player progression without it needing to be as janky and jarring as it would be with instant whole integer leaps up or down. No one needs to be able to know a player has a 13.2 for one attribute and another has a 13.4 because it makes no demonstrable difference.

Dribbling forms of course a very important part of dribbling with the ball and yet its not the only part. The best can ride the challenges and jostles along the way, the very best can perhaps avoid them altogether in some circumstances. This comes from a host of different attributes working in conjunction. 

How are you supposed to represent the difference between a dribbling of 75 and 76 in a scale of 1-80? In a match a lot of the time you'd struggle to tell the difference between a 10 and 15 because specialisation will only reveal itself in certain situations. There are so many situations in football matches where being the better player/s doesn't really mean all that much (see Manchester United for more details) which is why sometimes very highly rated players can be so underwhelming to use. If you don't play in a way that gives the specialisation a chance to shine you'll find a much, much "lesser" player can perform just as well or maybe even better if they have a few attributes higher elsewhere. 

[HUNT3R]

3 hours ago, santy001 said:

Dribbling forms of course a very important part of dribbling with the ball and yet its not the only part. The best can ride the challenges and jostles along the way, the very best can perhaps avoid them altogether in some circumstances. This comes from a host of different attributes working in conjunction. 

I'd agree with this. Different combinations of Acceleration, Agility, Balance and Strength, together with Dribbling, can have produce different types of player with a different set of skills. I'm sure having 'smarter' (Decisions and Anticipation, especially) dribblers can add another dimension to a player too.

[speck]

How far down the pyramid does the 1-20 attribute scale allow you go while still having somewhat accurate gameplay modeling? I've seen downloads of custom databases down to level 11 or so, but I don't know if they're able to differentiate (in gameplay, quality of players, etc.) from say level 8 and level 11.

[santy001]

The scale of 1-20 is solely for judging players by a professional football standard so a 1 is meant to be quite a bit beyond what a pubs team player is capable of. I've never looked at them too closely as that level doesn't interest me, but accurately representing those players is always going to pose a challenge because of the nature of football at that level and the fact the attributes weren't really designed with that in mind.

[speck]

16 hours ago, santy001 said:

The scale of 1-20 is solely for judging players by a professional football standard so a 1 is meant to be quite a bit beyond what a pubs team player is capable of. I've never looked at them too closely as that level doesn't interest me, but accurately representing those players is always going to pose a challenge because of the nature of football at that level and the fact the attributes weren't really designed with that in mind.

Thanks, I asked a similar question about ten years ago and received a similar answer. I wasn't sure if FM might have been updated to handle it better, since I've been seeing people create those much lower level league databases. The addition of women's leagues piqued my interest, since I've been wondering if that might lead to some changes with the existing attributes system.  Of course I guess the women could just get their own 1-20 rating system and have it be completely separate from the men's game.

[StewartAHannah]

I am sure most of the data (hub) comes as part of the SciSports partnership agreement. I can't imagine they spend much time messing with attributes and players/staff history. SI will spend their time creating the algorithms for the AI engines and the games GUI interfaces. Except ManUtd cos they clearly mess big time with them they even made Ole a good manager !! 

[Michael Sant]

Changes to the attributes are something that has been raised from within the community before. As with all ideas, this is something that could be rejected or after further development could go on to become a part of the FM series in the future.