Attribute Weighting?

A few versions ago, a lot of research was done on the weighting of various attributes by position to find out how best to optimise training. I forget who did the research, but does anyone else remember it and do they know if the following table is relevant on FM10?

For those of you who do not want to know how CA is/was calculated please close this thread now.

Oh well the table messed up, but those of you who know what I am on about should recognise it.

Fixed the table...

Att/ Posn	GK	DR/L	WBR/L	SW	DC	DMC	MC	AMC	MR/L	AMR/L	ST	Training category
Acceleration	3	4	4	4	4	4	4	5	5	6	6	Aerobic
Agility		4	3	3	3	3	3	3	3	3	3	3	Aerobic
Balance		2	2	2	2	2	2	2	2	2	2	2	Aerobic
Inj Prone	0	0	0	0	0	0	0	0	0	0	0	
Jumping		1	2	1	3	4	1	1	1	1	1	4	Aerobic
Natural Fitness	1	0	0	0	0	0	0	0	0	0	0	Strength
Pace		2	4	4	4	4	4	4	5	5	6	6	Aerobic
Stamina		1	3	4	3	3	3	3	3	3	3	2	Strength
Strength	3	3	2	3	4	3	3	3	2	2	4	Strength
Corners		0	1	1	1	1	1	1	1	1	1	1	Set pieces
Crossing	0	2	3	1	1	1	1	1	4	4	2	Set pieces
Dribbling	0	1	2	1	1	2	2	3	3	4	3	Ball control
Finishing	0	1	1	1	1	2	2	3	2	2	4	Shooting
First touch	1	2	3	3	2	3	3	3	3	3	4	Ball control
Free kicks	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Heading		1	2	1	2	4	1	1	1	1	1	4	Ball control
Long shots	0	1	1	1	1	3	3	3	2	2	2	Shooting
Long throws	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Marking		0	3	2	2	4	1	1	1	1	1	1	Defending
Passing		1	2	3	3	2	4	4	4	3	2	2	Attacking
Penalties	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Tackling	0	4	3	2	4	4	3	2	2	2	1	Defending
Technique	1	2	3	3	1	3	3	3	3	3	3	Ball control
Versatility	0	0	0	0	0	0	0	0	0	0	0	
Aggression	0	0	0	0	0	0	0	0	0	0	0	
Anticipation	2	3	3	4	3	3	3	3	3	3	3	Tactics
Bravery		4	2	1	2	2	1	1	1	1	1	1	
Composure	2	2	2	3	2	2	3	3	3	3	4	Shooting
Concentration	4	4	3	4	4	3	2	2	2	2	2	Defending
Consistency	0	0	0	0	0	0	0	0	0	0	0	
Decisions	4	4	3	4	4	3	3	3	2	2	2	Tactics
Dirtiness	0	0	0	0	0	0	0	0	0	0	0	
Flair		0	0	0	0	0	0	0	0	0	0	0	Ball control
Imp Matches	0	0	0	0	0	0	0	0	0	0	0	
Iinfluence	2	1	1	2	2	1	1	1	1	1	1	
Off the ball	0	1	2	1	1	1	2	3	2	2	4	Tactics
Positioning	4	4	3	4	4	3	2	2	1	1	2	Tactics
Teamwork	2	2	2	1	1	2	2	2	2	2	1	Tactics
Creativity	1	2	2	2	1	3	4	4	3	3	2	Attacking
Workrate	1	2	2	1	2	4	3	3	3	3	2	Strength
Aerial ability	4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Command of area	4	0	0	0	0	0	0	0	0	0	0	Tactics
Communication	4	0	0	0	0	0	0	0	0	0	0	Tactics
Eccentricity	0	0	0	0	0	0	0	0	0	0	0	
Handling	5	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Kicking		4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
One on ones	4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Reflexes	5	0	0	0	0	0	0	0	0	0	0	Aerobic
Rushing out	0	0	0	0	0	0	0	0	0	0	0	Tactics
Tend to punch	0	0	0	0	0	0	0	0	0	0	0	
Throwing	3	0	0	0	0	0	0	0	0	0	0	Goalkeeping

Read:

0 - is free attribute for this position

1 - least weight for position ... 6 - heaviest weight for position.

Hope this helps you pinpoint the formulas. Note that the degree of proficiency in a position reflects on the CA. Thus, a 20 in ST with 20 in AMC will actually reflect 50% of the above weights for ST and 50% for AMC.

You also need the "weaker foot" weighting. Still, great work!

You also need the "weaker foot" weighting.

Is the weaker foot still valid on FM10?

I was led to believe that the weighting on that ha changed in some way to reduce its impact, then again I may of just imagined that.

It is still weighted, yes. Not as much as it was for FM09. CA calculations therefore require it.

Is it a viable way to judge players when buying/selling or when selecting a lineup ?

When I play with Real Madrid I don't look at stats because I know who's a good player and who isn't, but when I manage other teams it's very hard to judge who's better, especially when I manage low level clubs, like conference or D1B clubs.

I'd love a spreadsheet with some funky formula to help me give a score to my players (Average just isn't good enough)

how was this table done?

You should take this to the Tactics and Training forum. The likes of Sfraser and prozone have been working hard to create an optimised training system.

hey can anyone please tell me how to use this table.

hey can anyone please tell me how to use this table.

This table was for a previous version and is fairly difficult to explain, but I will do my best (I wish I could find the original research and who did it).

Basically it was found on previous incarnations that every attribute was not worth an equal amount of CA and some attributes cost zero or very little CA for certain positions, so take one column, in this example we will use DC (Central Defenders).

By going own this column you will see that the following attributes cost NO CA:

Inj Prone, Natural Fitness, Versatility, Aggression, Dirtiness, Flair and Imp Matches (of those we can only increase Flair with Ball Control training)

Likewise all the attributes marked with a 5 or 6, cost a lot of CA to raise and are generally useful to the positions paid.

The number is effectively how costly in terms of CA it is to raise each ability, so when you have an older player who has reached his peak, you can design a training schedule which takes advantage of the low cost of free attributes to keep him useful.

Now do not get confused and assume a value of 6 in a column means it costs 6 points of CA to raise and ability, it does not work like that, there is a lot of math behind the abilities, including the fact that even though we see them as a value of 1-20, they are actually stored in the game as values of 1-100.

Like I said, it is pretty complex an it has been at least a year since I last looked at the table my self, I just wanted to see if it was still valid so I could design some training schedules for my first team players who have reached 27+ years old.

Hey The Lambs

Thanks for that table. It's great. Going to do some experimentation based on it.

There was this thread from a while back that might interest you...

http://community.sigames.com/showthread.php?t=75694

Thanks

Not wanting to be a pain, but any chance of this in a .ods/.xsl document? (ie. spreadsheet)

Don't have time to do it myself atm, so if someone already has, would be handy

Cheers

The above would be useful, but could you please remeber where the original research if possible? Yeah good work though :thup

By going own this column you will see that the following attributes cost NO CA:

Inj Prone, Natural Fitness, Versatility, Aggression, Dirtiness, Flair and Imp Matches (of those we can only increase Flair with Ball Control training)

Likewise all the attributes marked with a 5 or 6, cost a lot of CA to raise and are generally useful to the positions paid.

The number is effectively how costly in terms of CA it is to raise each ability, so when you have an older player who has reached his peak, you can design a training schedule which takes advantage of the low cost of free attributes to keep him useful.

It's a great table and great work, but the above is not entireally accurate.

Attributes with 0 CA Weight are not "Free" but independant of the CA/PA system for attribute change. This means rather than costing no CA to change, instead they cannot be changed at all other than by a small number of specific gameplay events.

For all attributes weighted 1 to 6 in the table, they change according to CA quantities and CA attribute behaviour mechanics, so can be trained and changed that way, can increase/decrease as CA increases/decreases and so on. Attributes weighted at zero are a completely different type of attribute that behave in different ways.

They are first of all generally more "behaviour/personality defining" such as Determination/Aggression/Flair/Natural Fitness etc. that have a very big say in playstyle and behaviour. Or they are hidden attributes that define strengths/weaknesses in key areas of a player such as Injury Proneness, Important Matches and so on.

These attributes are incredibly hard to change because they define huge aspects of a players individuality. They all change not through training, but by ingame events such as Mentoring, or playing more regularly/big performances, or disciplinary measures or by getting serious and longterm injuries.

Two things regularly overlooked/misunderstood is that there is a select number of relatively static yet personality/behaviour/playstyle defining attributes. These attributes represent "natural talent" or "natural gifts" etc. For example Carrick will never, ever reach the same level of Playmaking ability as Xavi because he has such a low Flair attribute and prefers not to attempt the flamboyant. These are key "natural personality/playstyle" defining attributes, and players that naturally lack certain key attributes will never be "the greatest ever" in certain roles, no matter their PA etc.

The second thing overlooked, and it is quite hard to spot, is that FM has a rather deep "event driven development" system for players. Players that achieve certain things, fail in certain things, consistently perform certain things etc. will improve and decline in the related attributes. A youngster that plays more regularly, puts in good performances, scores important goals will increase his Influence and become a more important player in the eyes of other players. He will improve his Consistency/Important Matches etc.

Likewise a Free Kick specialist that regularly takes and scores from Free Kicks, but does not do any set piece training, will still improve in his Free Kick attribute. Someone made captain will likewise improve in his Influence. Someone that gets a broken leg will become more injury prone, and so on.

These "event based changes" do not happen regularly, but they do happen and effectively chart a players career. You have limited direct control over them, and cannot change these attributes like you can the rest of the "Ability Attributes".

SFraser

Big thank you for the insight I just got from reading your post. I was unaware of the event based changes, and I'd never thought of free attributes in the way you defined them. Thanks for sharing your knowledge and insights.

Don't mean to bump an old thread, but why do 0-weight attributes, if not influenced by training still have a pertinent training schedule?

I know this is an old thread but i'm not understanding this.

So 1 accelaration is 6 CA points for a striker?

Well then if he has 20 in that area surely he is taking up 120 CA and pretty much fully using his CA up???!?

Nevermind: Just read a previous post, so how much more CA would it cost to take a strikers finishing from 17 to 19?

Or his composure to 15 to 18?

If anyone is interest I created training schedules using this data.

http://community.sigames.com/showthread.php/264658-Training-Attribute-Weighting-Ultimate-Training-Schedules-Released

I know this is a very old thread, but seems to be the only one who really can help what I am experiencing.

I edited a player in FMRTE to have a higher potential and some better stats. The stats dropped after couple of weeks, to reflect the CA that I didn't change, that's what I still understand. But when I look at his potential stats (Genie Scout), they are a lot worse than for players with the same position and roughly same PA. Look at the 2 pictures from Genie Scout.

Anyone knows what's the deal here?

I know this is a very old thread, but seems to be the only one who really can help what I am experiencing.

I edited a player in FMRTE to have a higher potential and some better stats. The stats dropped after couple of weeks, to reflect the CA that I didn't change, that's what I still understand. But when I look at his potential stats (Genie Scout), they are a lot worse than for players with the same position and roughly same PA. Look at the 2 pictures from Genie Scout.

Anyone knows what's the deal here?

Frank Kelly is 2 footed, the other chap isn't. 2 footed players generally look worse stat wise but play much better than the stats suggest.

I already tried changing the weaker foot to 1, this led to approximately +1 for each stat. Still means there is a big difference between the players. Also there are a lot of two-footed players with great stats (eg sneijder), so I don't think all this difference is because of the two-footedness.

Some weights going as high at 16 or 17 now I think, there are more factors for weightings now than previously I think too. Retraining a player to a new position may have no change on weightings or it may have a dramatic change on weightings, all depending on his natural positions and whatnot.

That table at the top is depressing reading, for some reason SI insist in every version on making raw Speed (pace/accel) pretty much two of the most important traits for most positions except GKs.

Not only is this utterly unrealistic in a real world sense (plenty of slow or not to mobile players are fantastic footballers - someone like Xavi springs to mind or Zidane) but it also makes the match engine very one dimensional, and it has been for years. There should be no 'weighting' as such, it should be dynamic. Some Strikers (for instance) are good cause they are fast, some cause they have fantastic movement, some cause they are like battering rams and so on...

In FM of those three strikers equally talented in their respective areas, the speed merchant would be most effective (given identical mental and hidden attributes) and thats a shame imo.

It means that an increase in speed comes at a higher premium. This doesn't necessitate that slower players can't be effective or useful, it just means for example if you were to train a player intensively to work on his pace it would use up more of his CA than if you did his technique for example. When players become very fast they often (not always) lose out in other aspects of their game, this is what the weighting system tries to balance out. A player will struggle to have 18+ pace and the touch and technique of a player like Xavi or Zidane.

Like Avelives said, if that's true that would be a shame. How about Messi and Ronaldo then, they are fast + excellent technique. In my opinion physical stats should even not cost CA points, it should just be determined by its beginning stats, and grow slowly depending on training.

I also have the feeling that the stats of a player are not fully down to CA. Since some players have way better stats but same CA. Maybe the starting stats of a player/regen also have an influence. Like two of my regens had same CA (90), but the stats for one of them were way better, like +3 average for each important attribute. I edited the potential to to the same level, seeing if in the end they would end up with around the same stats. It came out that the one with the higher beginning stats became way better, also +3 for each important attribute.

So maybe the stats that a player is given when created in beginning also have an influence on eventual stats (when same CA), or is this just crazy talk??

Sneijder has good attributes because of very high ca, and he is has only 16 at weaker foot if I remember well. I have made some tests, and difference between 1 and 20 at weaker foot at striker or midfielder means around 1 point less at technical-mental skills and 1-2 points (!) less at physical skills. Looking at http://community.sigames.com/showthread.php/254710-Your-most-prolific-striker?highlight=prolific+striker thread, you may see that one footed forwards are way better.

Like Avelives said, if that's true that would be a shame. How about Messi and Ronaldo then, they are fast + excellent technique. In my opinion physical stats should even not cost CA points, it should just be determined by its beginning stats, and grow slowly depending on training.

I also have the feeling that the stats of a player are not fully down to CA. Since some players have way better stats but same CA. Maybe the starting stats of a player/regen also have an influence. Like two of my regens had same CA (90), but the stats for one of them were way better, like +3 average for each important attribute. I edited the potential to to the same level, seeing if in the end they would end up with around the same stats. It came out that the one with the higher beginning stats became way better, also +3 for each important attribute.

So maybe the stats that a player is given when created in beginning also have an influence on eventual stats (when same CA), or is this just crazy talk??

I don't know. But when I edited my Adam le Fondre to some crazy god player with almost all attribute 20 but only 190 CA. His stats never dropped for 2 seasons. But for those generated wonderkids, they dropped 2 points at least about 2 months after being created.

I did not see Determination on the chart. Does anyone know if it is weighed or not?

Nope, it is not.

You may have increased Kelly physical attributes. Note that physical attributes don't cost CA points per se, the incremental costs.

However, Natural Fitness incremental doesn't cost CA.

is the table still the same for FM 12? has anyone notices any changes?

The table was never 100% accurate, however it was a good guide, and I imagine it still is.

You may have increased Kelly physical attributes. Note that physical attributes don't cost CA points per se, the incremental costs.

However, Natural Fitness incremental doesn't cost CA.

I edited Lukaku's physical stats using database editor and all his other stats dropped in game straight after starting.

How exactly is this table done? Anybody know how would 2012 version look like?

Fixed the table...

Att/ Posn	GK	DR/L	WBR/L	SW	DC	DMC	MC	AMC	MR/L	AMR/L	ST	Training category
Acceleration	3	4	4	4	4	4	4	5	5	6	6	Aerobic
Agility		4	3	3	3	3	3	3	3	3	3	3	Aerobic
Balance		2	2	2	2	2	2	2	2	2	2	2	Aerobic
Inj Prone	0	0	0	0	0	0	0	0	0	0	0	
Jumping		1	2	1	3	4	1	1	1	1	1	4	Aerobic
Natural Fitness	1	0	0	0	0	0	0	0	0	0	0	Strength
Pace		2	4	4	4	4	4	4	5	5	6	6	Aerobic
Stamina		1	3	4	3	3	3	3	3	3	3	2	Strength
Strength	3	3	2	3	4	3	3	3	2	2	4	Strength
Corners		0	1	1	1	1	1	1	1	1	1	1	Set pieces
Crossing	0	2	3	1	1	1	1	1	4	4	2	Set pieces
Dribbling	0	1	2	1	1	2	2	3	3	4	3	Ball control
Finishing	0	1	1	1	1	2	2	3	2	2	4	Shooting
First touch	1	2	3	3	2	3	3	3	3	3	4	Ball control
Free kicks	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Heading		1	2	1	2	4	1	1	1	1	1	4	Ball control
Long shots	0	1	1	1	1	3	3	3	2	2	2	Shooting
Long throws	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Marking		0	3	2	2	4	1	1	1	1	1	1	Defending
Passing		1	2	3	3	2	4	4	4	3	2	2	Attacking
Penalties	0	1	1	1	1	1	1	1	1	1	1	Set pieces
Tackling	0	4	3	2	4	4	3	2	2	2	1	Defending
Technique	1	2	3	3	1	3	3	3	3	3	3	Ball control
Versatility	0	0	0	0	0	0	0	0	0	0	0	
Aggression	0	0	0	0	0	0	0	0	0	0	0	
Anticipation	2	3	3	4	3	3	3	3	3	3	3	Tactics
Bravery		4	2	1	2	2	1	1	1	1	1	1	
Composure	2	2	2	3	2	2	3	3	3	3	4	Shooting
Concentration	4	4	3	4	4	3	2	2	2	2	2	Defending
Consistency	0	0	0	0	0	0	0	0	0	0	0	
Decisions	4	4	3	4	4	3	3	3	2	2	2	Tactics
Dirtiness	0	0	0	0	0	0	0	0	0	0	0	
Flair		0	0	0	0	0	0	0	0	0	0	0	Ball control
Imp Matches	0	0	0	0	0	0	0	0	0	0	0	
Iinfluence	2	1	1	2	2	1	1	1	1	1	1	
Off the ball	0	1	2	1	1	1	2	3	2	2	4	Tactics
Positioning	4	4	3	4	4	3	2	2	1	1	2	Tactics
Teamwork	2	2	2	1	1	2	2	2	2	2	1	Tactics
Creativity	1	2	2	2	1	3	4	4	3	3	2	Attacking
Workrate	1	2	2	1	2	4	3	3	3	3	2	Strength
Aerial ability	4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Command of area	4	0	0	0	0	0	0	0	0	0	0	Tactics
Communication	4	0	0	0	0	0	0	0	0	0	0	Tactics
Eccentricity	0	0	0	0	0	0	0	0	0	0	0	
Handling	5	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Kicking		4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
One on ones	4	0	0	0	0	0	0	0	0	0	0	Goalkeeping
Reflexes	5	0	0	0	0	0	0	0	0	0	0	Aerobic
Rushing out	0	0	0	0	0	0	0	0	0	0	0	Tactics
Tend to punch	0	0	0	0	0	0	0	0	0	0	0	
Throwing	3	0	0	0	0	0	0	0	0	0	0	Goalkeeping

Read:

0 - is free attribute for this position

1 - least weight for position ... 6 - heaviest weight for position.

Hope this helps you pinpoint the formulas. Note that the degree of proficiency in a position reflects on the CA. Thus, a 20 in ST with 20 in AMC will actually reflect 50% of the above weights for ST and 50% for AMC.

thanks a lot...

very usefull...

Is this table still valid for FM2014?

Is this table still valid for FM2014?

Probably not... The training system has been revised and the "categories" have been removed, in favour of the role-based training (which is, IMO, a step backwards in terms of players development freedom, sacrificed on the altar of "realism").

In FM14 you can't design a custom training schedule anymore, but all you can do is picking a role for your players to train in... Too bad you can only guess know which attributes will be trained.

This is particularly disappointing when you have a heavily unbalanced player, say, a careless DC or a weakass MC... With the old system you could try to "fix" him by giving him a specific training routine focusing on his weaknesses.

Currently you just go for the one-size-fits-all schedule hoping for the best, but you won't ever be able to close the gap a bit, not unless you go for one attribute at a time IF all the required ones are available for individual focus.