Jump to content

Current Ability and Atrributes Research


Recommended Posts

Edit 21.02.08 - If you just want a quick answer or to just view the results then click the link CLICK ME!!!!!!!

Definitions

“CA†refers to current ability (0-200)

“Max CA†refers to the generic maximum i.e. 200 CA, not the actual maximum CA (i.e the potential ability) of a player.

TATT – total attributes i.e. all the players attributes added together

Formatting

Hopefully what I've written will have come out ok but if not then any mod is welcome to put it right icon_wink.gif

Introduction

The relationship between current ability (CA) and player attributes has interested me for a while now, as dull as it sounds.

So the crux of this thread is how attributes relate to current ability. Primarily for now, the focus will be trying to determine whether for a player with CA X, there is a fixed (or fixed within upper and lower limits) minimum, maximum or average number of total attributes (and to as lesser extent whether it also might differ per position). And also to see if it is possible to accurately estimate a player’s CA (a hidden stat) by using a player’s total attributes (TATT)

For this I will have to go to the dark side and either use the editor or download FMM to check the current abilities, but it shall be in the name of research.

So, for example, if 10 players all have a CA of around 150, do they all have the same/a similar number of TATT?

Do they all have not less than X and/or not more than Y? What is the average number of total attributes?

And then further, comparing a player like Ronaldinho who has a very high CA with a lower ability player? I assume he will have a greater number of total attributes, but it will be interesting to explore this, especially to try and find out if there is a maximum number (out of 720) that a player can have (assuming they have 200 or close CA). I assume that there is indeed a theoretical maximum number (to stop players having 20s in everything for realism) but it might be interesting to try to establish what that number is.

Also, can we work out the current ability of a player just by looking at their TATT?

Predications

You could go on to assess the full total of attributes by including the hidden attributes in the analysis but I never use FMM normally in my games and don’t want to explore that avenue.

All figures are rounded up unless otherwise stated.

Obviously the maximum number of attributes is 720 (36 attributes with a max of 20 in each theoretically).

Also, I will be obviously including the mental attributes like aggression and determination that might not necessarily be influenced by CA but hopefully this won’t make such a big impact, and at least it will be consistent in approach.

GK are perhaps somewhat of a special case, so comparisons with outfield players might not be accurate.

I should say from the outset that I do not having anything in mind to prove or disprove in terms of improving my experience of FM. I am doing this purely out of interest and well, a pursuit of knowledge and exploration of something that interests me. It is not intended to hopefully produce some in-game advantage or to highlight any flaws in the game. Hopefully it will do neither. Obviously player A could have a higher TATT than player B, but I might still prefer player B because for the position I want to play him in, he has higher attribute values in those areas that I feel I want for that position. This is in the same way that player A might have a higher CA than player B but for those reasons and others, I might still pick player B. This research doesn’t intend to prove to the contrary.

Let the research commence

Ok well I thought I’d start with my Middlesbrough squad and work from there – this should hopefully also give us a range of positions, CAs and ages. (Players are listed in their ‘natural’ position(s) (regardless of whether I’m playing them there are not and all values are in May at the end of the first season).

GK – Almunia – CA 152 – TATT 430

DR – Young – CA 135 – TATT 448

DC – Huth – CA 152 – TATT 510

DLC – Pogatetz – CA 144 – TATT 488

DL – Taylor – CA – 134 – TATT 496

ML – Arca – CA 142 – TATT 475

DMC – Shawky – CA 138 – TATT 481

MC – Cattermole – CA 132 – TATT 477

MC/DMC – Rochemback – CA 147 – TATT 477

AMC – Piatti – CA 130 – TATT 422

AMC – Kapo – CA 146 – TATT 481

AMR – Torje – CA 135 – TATT 450

SC - Stancu – CA 138 - TATT 437

And then (not my players...!):

GK – Buffon – CA 185 – TATT 493

DL – Evra – CA 170 – TATT 534

DR – Alves – CA 179 – TATT 583

DC – Terry – CA 182 – TATT 534

AML – Ronaldinho – CA 188- TATT 527

AMR – Ronaldo - CA 192 – TATT 519

DMC – Mascherano – CA 175 – TATT 510

MC – Gerrard – CA 184 – TATT 549

AMC – Kaka – CA 192 – TATT 522

SC – Eto – CA 186 – TATT 522

Summary

My Boro lads

Average CA: 140.38

Average CA as a percentage of maximum: 70.19%

Average TATT: 467.08

Average TATT as a percentage of maximum: 64.87%

Average CA as a percentage of average TATT: 30.05%

Huth CA as a percentage of TATT: 29.80%

Good players:

Average CA: 184.22

Average CA as a percentage of maximum: 92.11%

Average TATT: 531.44

Average TATT as a percentage of maximum: 73.81%

Average CA as a percentage of average TATT: 34.66%

Ronaldo CA as a percentage of TATT: 36.78%

From this we can see that whilst the difference between the average CAs of the two groups was over 20%, the difference in TATT was less than 10%. This is a very interesting result in my opinion. You would assume that CA would determine what level of TATT a player could “hold†if you will.

The biggest result was between Huth (76% of max CA) and Ronaldo (96% of max CA) with a difference of some 20%. Whereas, in terms of TTAT Huth (70.83% of max TATT) compares favourably with Ronaldo (72.08% of max TATT) with a difference of less than 2%. Does this tell us anything useful? I’m not sure really. It certainly tells us that Ronaldo is a better winger than Huth is a defender, because purely from their respective CA and TATT scores can see that Ronaldo’s attributes must be highly concentrated (assuming they are concentrated in the right areas – which they are) and Huths must by definition be fairly evenly spread given his much lower CA and similar TATT score. What use highlighting this is I’m not, except to encourage users to go for specific killer scores in key attributes.

The figures for average CA as a percentage of average TATT are also interesting. This way it might be possible to roughly work out (in-game) a players approximate current ability based on their stats. Obviously this can already be done fairly roughly at present i.e. if I look at Alves, he looks like he’s going to have a high current ability from all the high stats! But things might be deceptive. For example, Huth looks like he’s probably got a higher current ability than he has in my opinion. So if we take Huth and Ronaldo as the limits here than its possible to say that by using a players TATT, we can be accurate to within plus or minus 7% as to their CA score. Let’s call it 32.5% (the midpoint between 29% and 36%). Now, even as I’m writing this this sounds the most spurious of logic and I’m conscious of the fact that even if this is true, the phrase “so what...†springs to mind....but hey ho, its there.

To test this I’m going to pick players at random and then take their TATT score (in-game) and produce an approximate current ability and then test it by using FMM. Fingers crossed.

I’m a Bolton fan so let’s go there.

Diouf

TATT: 476

Predicted CA: 155 (476 x 32.5%)

Actual CA: 160

Difference: 2.5%

Age: 27

Peak Ages: 27-32

Andranik

TATT: 461

Predicted CA: 150 (461 x 32.5%)

Actual CA: 148

Difference: 1%

Age: 25

Peak Ages: 27-32

And Gks

Jussi

TATT: 435

Predicted CA: 141 (435 x 32.5%)

Actual CA: 161

CA as a percentage of TATT: 37%

It seems this doesn’t hold water for GKs! This is fairly understandable I think as you’re not comparing like for like between GKs and outfield players. Let’s just try 37% on some other GKs and see what happens......

Almunia

TATT: 430

Predicted CA: 159 (430 x 37%)

Actual CA: 152

Difference: 3.5%

CA as a percentage of TATT: 35%

Let’s try it with a younger GK. My prediction is that using 37% will vastly over predict a younger goalkeeper’s CA.

Brad Jones

TATT: 390

Predicted CA: 144 (390 x 37%)

Actual CA: 125

Difference: 10.5%

Age: 26

Peak Ages: 31-35

CA as a percentage of TATT: 32.05%

And youngsters

Zeefuik

TATT: 434

Predicted CA: 141 (434 x 32.5%)

Actual CA: 102

Difference: 19.5%

CA as a percentage of TATT: 23.50%

Age 17

Peak Ages: 26-31

It doesn't seem to work in this example of a youngster either. However, I think this can be explained (see below).

Hmm.... there are some patterns to be found but so far its only “normal aged outfield players†and “older-aged†outfield players and “normal aged GKs†that we can see a pattern with.

So far so good. Let’s see if it still holds true for players past their prime:

Campo

TATT: 431

Predicted CA: 140

Actual CA: 108

Difference: 16%!!

CA as a percentage of TATT: 25.06% - a bench mark applicable to older players perhaps?

Speed

TATT: 422

Predicted CA: 137 (422 x 32.5%) or;

106 (422 x provisional 25% “older players†benchmark)

Actual CA: 100

Difference from “normal†players formula: 18.5%

Difference from “older†players formula: 3% - this is more acceptable

CA as a percentage of TATT: 23.7%

Giggs

TATT: 485

Predicted CA: 121 (485 x 25%)

Actual CA: 136

Difference: 7.5% - just about acceptable.

CA as a percentage of TATT: 28.04%

I think that obviously different players start to decline at different rates and Giggs is just going into decline on my game now so I think if a player has gone into decline the slightly lower “older players rate†should apply (god it sounds like tax law!). It looks like the older and/or more declined a player has become then the lower the rate you need to apply. The closer the player is to still being a “normal player†that you’d play week in week out, it seems that you should err progressively higher to the 32.5% benchmark for normal players.

Lets try another player.

Hamman (not as old as Speed at 34 (same as Giggs) but has gone into decline on the game in my opinion by this point)

TATT: 414

Predicted CA: 116 (414 x 28% same as Giggs)

Actual CA: 114

Difference: 1%

CA as a percentage of TATT: 27.54%

I am almost losing the will to live as I continue this journey – despite what the above might suggest, I really hate maths and never touched since leaving school – so could some other people chip in with some tests please!? Lol.

Conclusions - could it be that:

Youngest - Younger is approx. 23% - 28%

Older to Oldest is approx. 28% - 23%

Players within “normal†age range (but either only just into peak or not yet into peak) approx 32.5%

Players just into peak and into peaks its 33% - X%

I also think that these figures will be influenced by the determination (and possible other mental stats too) not just in terms of numerically, but also in terms of how hard that person will train, perhaps distorting CA disproportionately to TATT. Determination might also affect the rate of decline as a player gets older which could distort the formula slightly relative to age, and similarly with youngsters. But hopefully if people are willing we can do further testing on the above four types of players.

Based on the above I decided to do one last test:

Anelka

TATT: 470

Predicted CA: 165 (470 x 35% - as here’s into his peak in a comparable way to Jussi but is less determined so won’t train as hard– see below)

Actual CA: 165

Difference: 0%! (Well technically the predicted CA figure was 164.5 but I’m rounding all figures up for ease).

Further Research

Could someone test these out as I’ve been at this for hours, and have the new patch downloaded and haven’t even played a game with it yet!! (I know, poor me...).

Maths disclaimer

I’ve tried my best to ensure all workings out are accurate but if the odd one isn’t then it is purely through losing the will to live whilst doing this and simple human error. As I said I wasn’t really trying to prove anything in particular for my own ends or to discredit SI so I’ve got no reason to purposely distort any figures.

General disclaimer

I don’t say that all of the above is amazing research or even that it brings anything interesting to light, that’s for people reading it to decide. I also don’t claim that the above theories and conclusions are 100% watertight. As I’ve said, I’d like more people to test similar things to what I have just done as we need a bigger sample.

A big thank you in bold for getting to the bottom.

Link to post
Share on other sites

  • Replies 741
  • Created
  • Last Reply

Law Man if you really want to know the relationship between CA and attributes you could create a model making some base assumptions about the weightings applied to different attributes (the accuracy of the model would largely be influenced by the accuracy of this assumption) and solving the simultaneous equations. Someone who studies Maths and has access to the right software could do this pretty easily, or if you were dedicated to it you could do it the old fashioned tedious way of stepping through the model.

The basic model for outfield players (where * indicates 'multiplied by')

CA = (w1*A1) + (w2*A2) + (w3*A3) + ........... + (w35*A35) + (w36*A36)

where

CA = Current Ability

wn = weighting n applied to attribute An for n = (0,1,2,3,..........,35,36)

An = attribute n for n = (0,1,2,3,.........,35,36)

1. Assumption 1: Weightings are a numeric value between 0 and 1 and are not in themselves functions of other variables

2. Assumption 2: Weightings are applied based on natural position only

3. Assumption 3: Only visible attributes are incorporated in the model

You could add another assumption about certain variables being independent of CA. Not knowing precisely which attributes are part of the CA calculation will create potential inaccuracies but the model could still serve as a decent rough approximation.

These assumptions allow for simplicity but at the same time this simplicity could lead to inaccuracies.

So using the assumptions you take 36 players for each natural position and solve the simultaneous equations for the weighting values. Then you have a model for the relationship between CA and attributes. With the right software this could be done easily but doing it by hand would go way beyond tedious, and this in itself could lead to mistakes.

Link to post
Share on other sites

Originally posted by isuckatfm:

Law Man if you really want to know the relationship between CA and attributes you could create a model making some base assumptions about the weightings applied to different attributes (the accuracy of the model would largely be influenced by the accuracy of this assumption) and solving the simultaneous equations. Someone who studies Maths and has access to the right software could do this pretty easily, or if you were dedicated to it you could do it the old fashioned tedious way of stepping through the model.

The basic model for outfield players (where * indicates 'multiplied by')

CA = (w1*A1) + (w2*A2) + (w3*A3) + ........... + (w35*A35) + (w36*A36)

where

CA = Current Ability

wn = weighting n applied to attribute An for n = (0,1,2,3,..........,35,36)

An = attribute n for n = (0,1,2,3,.........,35,36)

1. Assumption 1: Weightings are a numeric value between 0 and 1 and are not in themselves functions of other variables

2. Assumption 2: Weightings are applied based on natural position only

3. Assumption 3: Only visible attributes are incorporated in the model

You could add another assumption about certain variables being independent of CA. Not knowing precisely which attributes are part of the CA calculation will create potential inaccuracies but the model could still serve as a decent rough approximation.

These assumptions allow for simplicity but at the same time this simplicity could lead to inaccuracies.

So using the assumptions you take 36 players for each natural position and solve the simultaneous equations for the weighting values. Then you have a model for the relationship between CA and attributes. With the right software this could be done easily but doing it by hand would go way beyond tedious, and this in itself could lead to mistakes.

Go for it icon_biggrin.gif My brain really doesn't work that way so if yours does then great, lets try and do some further research icon_smile.gif

Link to post
Share on other sites

Go for it My brain really doesn't work that way so if yours does then great, lets try and do some further research

I tried to do it before out of curiosity but the tedium lead me to stop pretty early in my attempt. Unless I'm getting paid I prefer not to spend my leisure time solving simultaneous equations. But I thought from the effort you put into your post I might be able to cajol you into doing it. I guess I was wrong icon_biggrin.gif

Maybe an FM playing Maths student with access to software for solving simultaneous equations might drift into this thread and do it the easy way icon_wink.gif.

Link to post
Share on other sites

Hehe well good effort anyway icon_biggrin.gif Sadly I have neither the software nor the expertise - I'm very much a words man...... In fact, I used to hate simultaneous equations in maths at school!

But at least you flagged the opportunity up, so if there's anyone out there who fancies it....... icon_wink.gif

Link to post
Share on other sites

If someone can tell me how to pull all attributes for a bunch of players, I could run a simple regression to determine those weights, and maybe even tests which attributes are more significant for each position. The problem is that it's too time consuming to pull data player by player. Though I am pretty sure it could be done relatively easy.

Link to post
Share on other sites

If someone can tell me how to pull all attributes for a bunch of players

I don't know how to do this, but hopefully someone more knowledgeable than I might be able to access the database somehow icon_smile.gif

I could run a simple regression to determine those weights, and maybe even tests which attributes are more significant for each position.

What's a "regression"? I am massively out of my depth with anything maths based! icon_biggrin.gif

The problem is that it's too time consuming to pull data player by player. Though I am pretty sure it could be done relatively easy.

Completely agree it would be too time consuming doing it 'by hand' so to speak player by player and I wouldn't encourage anyone to be mad enough to attempt to do so. Its important that we get as big as sample as possible.

Link to post
Share on other sites

If someone can tell me how to pull all attributes for a bunch of players, I could run a simple regression to determine those weights, and maybe even tests which attributes are more significant for each position. The problem is that it's too time consuming to pull data player by player. Though I am pretty sure it could be done relatively easy.

Pulling data is a key issue especially if you were to attempt to use regression. Also another issue with regression is that it's necessary that each of the variables have a similar statistical distribution to reduce the errors which I don't know if that would be the case. Then again it's been a few years since I studied this stuff so I could be completely wrong as my memory seems to be deteriorating as I get older icon_wink.gif.

That's why I suggested that simple model. If the assumptions hold true then theoretically all you need is 36 players with the same natural position to solve the set of equations for that position. With the right software solving these equations would be a breeze. Doing it by hand is tedious to say the least. There's probably an easier way to do it using matrices rather than by a stepping sequence but as with regression it's been years since I did that stuff and it's all just a hazy blur at this point in time.

If the mood strikes I might give it a go but I wouldn't hold my breath.

Completely agree it would be too time consuming doing it 'by hand' so to speak player by player and I wouldn't encourage anyone to be mad enough to attempt to do so. Its important that we get as big as sample as possible.

If the program kolobok intends to use takes excel as input, and we can get enough contributors it might be possible to get a sample of reasonable size. It won't exactly be within the desired sample size to produce a high level of confidence but it could give a rough idea of the relationship.

The model produced could be significant to training aswell as it could be used to see how you could redistribute attribute points within a given CA by adjusting training sliders.

Link to post
Share on other sites

If the mood strikes I might give it a go but I wouldn't hold my breath.

Quick, do it before I run out of oxygen! icon_biggrin.gif

Then again it's been a few years since I studied this stuff so I could be completely wrong as my memory seems to be deteriorating as I get older

That's still infinitely ahead of where I am as all I did was GCSE maths!! Emre trying to stand up in the Atlantic would be less out of his depth than me currently! lol.

The model produced could be significant to training aswell as it could be used to see how you could redistribute attribute points within a given CA by adjusting training sliders.

Good point, I hadn't thought of that.

I really hope we can get this off the ground. I feel like writing a comedy sketch - if only we pull together Kobolok's programme, your maths skills, some software to pull the attributes on mass and mug of a coffee, we'd be set!

Link to post
Share on other sites

I really hope we can get this off the ground. I feel like writing a comedy sketch - if only we pull together Kobolok's programme, your maths skills, some software to pull the attributes on mass and mug of a coffee, we'd be set!

The thing about Maths jargon is I know the language so for someone with less exposure to it might think I'm a genius. But someone who knows about it may recognise that I am in fact talking through my arse (this isn't a conscious decision driven by ego but more to do with my sieve like memory and it's limited capacity. Much like Homer Simpson learning new things tends to push other stuff out icon_wink.gif). So someone with a PhD or currently studying Maths may well read what I wrote and pull me up on it.

No more then if you were to write some legalese it would probably sound to me as if you know a great deal about it, when in fact you could be talking rubbish but I'd be none the wiser (I have a vague recollection of reading in one of your posts that you're a lawyer/studying law, or am I just making that connection because of your screen name? Either way you could substitute anything you know alot about and I know knothing about in 'legalese' in the the sentence and the point would still stand.)

Link to post
Share on other sites

PaulC posted this information regarding the drop some players are seeing in attributes since 8.02. His explaination and further reply seemed pertinent to this discussion.

Originally posted by PaulC:

Ok folks, I'll tell you exactly what is happening here.

If you are seeing a drop in some of your players' attributes after a short time playing 8.0.2, then its because we have modified the CA-attribute balance slightly. The algorithm we used in FM2008 was less harsh than the one in FM2007, and this was contributing to some of the flaws in the match engine, with an overload on certain types of attacking player. As a result we have modified the algorithm for 8.0.2 to be a halfway point between 7.0.2 and 8.0.1.

To clarify, its not training related, and it has no bearing on player CA progression over time. That doesnt mean progression hasnt been tweaked itself, because it has but that is a separate issue.

Cheers,

Paul

ps The reason you see the red arrows is simply that the game *thinks* training has modified the attributes, but you can ignore the initial red arrows because its down to something else.

Originally posted by PaulC:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Asmodeus:So does this mean that we can expect our younger players to increase again in the future after the initial decline (PA permitting)?

It has no effect on their ability to improve.

Originally posted by flibby7:First to say that english is not my language, but after reading this twice I still dont understand.

This means what? My best player with 18-19 in speed, acc., heading,composure and jumping droped his ratings to 16-17 and will not gain them back in those positions, but rather in some other areas or what?????????

It means the new algorithm considered his attributes too high for his CA, and adjusted them. If he improves, then his attributes in general will rise....obviously which attributes change depends on his position, his age and his training regime.

A quick rule of thumb for attribute progression:

- Mental attributes grow through career

- Technical may grow less as they are essentially "natural". Training can improve them to a degree.

- Some physical attributes ( eg jumping, strength, stamina ) can grow when a player is young.

- Some physical attributes ( eg stamina, pace, acceleration ) will start to drop when a player hits a certain age.

Each attribute is worth X CA points for each player, depending on his position(s). So for each rise, there should be a CA increase. If his CA is maxed out, each rise will see a fall elsewhere. And vice versa.

I wont give any more away than that, as its mainly common sense anyway icon_smile.gif </div></BLOCKQUOTE>

Link to post
Share on other sites

This is what PaulC emailed back. Credit to him for emailing back and looking at the thread within minutes, even when he's obviously fairly busy:

Hi Richard,

Quite busy but a cursory read suggests you aren’t accounting for attribute weightings in your calculations. All attributes are weighted according to player position(s).

Also, the attributes that are discarded from the CA control are only the tendencies. So aggression is a tendency, but determination is more of an ability. That’s how we see it anyway.

Otherwise, excellent thread J

Cheers,

Paul

So this, along with what Kolobok and isuchatfm have said, and what powermonger added from PaulC, should give us some real direction. Now all we need is someone with the skills and software and motivation to have a goood go at this icon_smile.gif

Link to post
Share on other sites

@Law Man

You can use the model I posted to get the position specific weightings. Plus what PaulC said about tendencies not being part of CA I had included initially, but I wasn't sure of which attributes are removed.

In fact removing the attributes makes finding the solution easier as we have less unknowns, the weightings w in the equation.

One other issue that would affect the accuracy of the model is that attributes are displayed between 1 and 20 but are actually stored over a greater range (according to FM Modifier it is 1 and 100). But still using 1 to 20 should give a useful model.

We can account for different positions if we assume that natural position is the key control. Then we can establish weightings for a specific natural position and the CA model for that natural position. What do you reckon to that?

So what else would you consider a tendency besides aggression (assuming hidden attributes like consistency and important matches aren't controlled by CA for simplicity). I would have assumed determination was a tendency myself, so based on what PaulC said the only ones I can think of are aggression, bravery and flair (maybe creativity). For goalkeepers I would say eccentricity as well.

Also would you consider the following scheme as an assumption with regards to weightings on attributes within the CA model:-

1. left back = right back

2. sweeper = centre half

3. left winger = right winger

4. striker = forward

1 and 3 I feel are realistic enough to assume but 2 and 4 might be a bit dodgy. What do you reckon?

Link to post
Share on other sites

What about the weighting of the physical stats, equal weight for all positions?

This is something that i'm thinking of testing would give your CA model slightly less variables if it is the case.

With the model in my posts I make the assumption that each position has equal weightings for players who have it as a natural position so that I don't have to figure out how weightings vary across positions within the same model.

So with my assumptions adding equal weighting for physical attributes independent of position to the equations would be irrelevant.

But there is more than one way to skin a cat so taking that approach and incorporating that into a model is equally feasible if you believe from looking at players in the database that this is the case. If you're having a go at this yourself and actually reach the point of having a model with the weightings assigned then by all means post it in here.

I don't intend to cross over positions except possibly for those that I think it is valid to assume they are equivalent e.g. left back = right back = full back.

Basically what I'll do (if the tedium doesn't get to me icon_wink.gif) is take a natural position, say centre half and select a number of these players equivalent to the number of attributes in the model. Then I'll use the attributes and CAs of these players to figure out the weightings. In this way it will identify weightings that are specific to natural positions. If as you suggested physical weightings are independent of natural position then this should fall out of the equations as part of the process.

Link to post
Share on other sites

You can use the model I posted to get the position specific weightings.

Nice one icon14.gif

One other issue that would affect the accuracy of the model is that attributes are displayed between 1 and 20 but are actually stored over a greater range (according to FM Modifier it is 1 and 100). But still using 1 to 20 should give a useful model.

I agree with you, I still think it will be useful.

e can account for different positions if we assume that natural position is the key control. Then we can establish weightings for a specific natural position and the CA model for that natural position. What do you reckon to that?

As far as I understand it.....that sounds good icon_biggrin.gif

PaulC said that determination is "more of an ability" than a tendency (however bizarre that sounds they seem to classify it that way).

I think it would be logical to classify most of the mental attributes as 'tendencies' but we'd just be guessing. I could always email PaulC again in a few days when he's less busy and ask.

Of the mental attributes I agree with you on aggression, bravery and flair (but I think creativity would be an "ability). I also think that work rate and teamwork could be "tendencies" too. What do you think?

Yes completely agree with left back/right back and left winger/right winger. I think the sweeper and centre half and striker and forward would still apply to a large extent. There might be some attributes that would differ but alot of it would probably apply, especailly with forwards as that's exactly the same position, so may be less so with sweeper as that's a different position.

Also, what about wingbacks, DMCs, MCs, AMCs, and AMRs/AMLs and GKs?

I might give it a lash anyway and then someone can test it in the patched version.

Please do! icon_smile.gif If I had the knowledge or expertise then I would, and I feel kind of bad starting a thread at a really basic level and other people doing the hard technical stuff but as I said I just don't have the maths for it.

Thanks for your contributions isuckatfm, and also everyone else who has chipped in so far icon14.gif

Link to post
Share on other sites

I don't intend to cross over positions except possibly for those that I think it is valid to assume they are equivalent e.g. left back = right back = full back.

I think this is the right way to do it.

Basically what I'll do (if the tedium doesn't get to me Wink) is take a natural position, say centre half and select a number of these players equivalent to the number of attributes in the model. Then I'll use the attributes and CAs of these players to figure out the weightings. In this way it will identify weightings that are specific to natural positions. If as you suggested physical weightings are independent of natural position then this should fall out of the equations as part of the process.

I think I sort of understand that now icon_smile.gif Can't wait to see what you come up with mate!

Link to post
Share on other sites

Originally posted by trebor1185:

What about the weighting of the physical stats, equal weight for all positions?

This is something that i'm thinking of testing would give your CA model slightly less variables if it is the case.

I would say they should apply equally, but that's based on nothing more than my personal opinion. This is another thing that I could add to my list to ask PaulC in a few days.

Link to post
Share on other sites

I think it would be logical to classify most of the mental attributes as 'tendencies' but we'd just be guessing. I could always email PaulC again in a few days when he's less busy and ask.

TBH I can't see him revealing that. Kind of like asking Coca Cola for part of the recipe.

Of the mental attributes I agree with you on aggression, bravery and flair (but I think creativity would be an "ability). I also think that work rate and teamwork could be "tendencies" too. What do you think?

I always thought that as well but the fact that PaulC views determination as an ability shows a difference between my interpretation and SIs. I was thinking of using the training graph to maybe figure it out. I might go with the worst case scenario and remove the ones we agree on. So aggression, flair and bravery for now.

Also, what about wingbacks, DMCs, MCs, AMCs, and AMRs/AMLs and GKs

That little statement right there encapsulates the tedium. Let's say we remove aggression, flair and bravery from the model. That leaves 33 visible attributes for each position. For each position I then have to select 33 players and write down their attributes and CA. I then have to solve 33 simultaneous equations for each position:-

FB, CB/SW, DM, WB L/R, M C, ML/R, AM L/R, AM C, F/ST

With a computer program doing the calculations even extracting all the variables would be a mind numbing process. So I might give it a go for one position and test how accurate it is. If it isn't up to much I'll post to say so and won't bother with the other positions as the lack of accuracy will indicate a fundamental flaw in the model.

Please do! If I had the knowledge or expertise then I would, and I feel kind of bad starting a thread at a really basic level and other people doing the hard technical stuff but as I said I just don't have the maths for it.

Reading over the thread I felt the opposite, that I'd hijacked your thread. So if you want to keep doing what you're doing and I'll open a new thread if and when I do it that's fine. But as i said it might take me a while as I'm a little ring rusty so it might be a couple of weeks. I'll repost in this thread if and when I do to let you know although it'll probably be buried under the barrage of threads that I reckon might be coming as people's tactics stop working on 8.02.

Link to post
Share on other sites

That is A LOT of tedium.... lol icon_biggrin.gif

Yes I reckon go with what you said and just do the one, and then post in here to let us know the results! icon_smile.gif For the one you do though, I reckon it would be best to do one of the "perfect" ones e.g MC rather than F/ST.

Cheers.

Link to post
Share on other sites

Originally posted by isuckatfm:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Yes I reckon go with what you said and just do the one, and then post in here to let us know the results! For the one you do though, I reckon it would be best to do one of the "perfect" ones e.g MC rather than F/ST.

I'll go with MC icon14.gif </div></BLOCKQUOTE>

Cool icon14.gif I await the results of your furious simultneous equation solving......! icon_smile.gif

Link to post
Share on other sites

Originally posted by Cleon:

My head hurts after reading this thread icon_biggrin.gif

Hehe you're about as "mathsie" as I am then pal! As I said further up, Emre trying to stand up in the Atlantic would be less out of his depth....lol.

But if isuckatfm and Kobolok can do the hard stuff we might find out some useful things icon_smile.gif

Link to post
Share on other sites

Just been doing a few small tests with the data editor, which has given some interesting results.

What I did was create a player who had CA + PA of 1, then gave him 20 for every attribute as well as set him as 20 for every position except for GK.

It appears that several attributes hold no/little weighting. The "test player" still had 20 for the 4 set piece attributes, 20 for aggression, determination, flair, and natural fitness. He had 13 for acceleration, agility, balance and pace. all the remaining attributes were 3.

Does anyone else think it's strange how the set piece attributes appear to have no/little weighting in relation to CA vs TATT?

But to be honest there useless without the other mental/technical attributes anyway.

Link to post
Share on other sites

But if isuckatfm and Kobolok can do the hard stuff we might find out some useful things
Hehe you're about as "mathsie" as I am then pal! As I said further up, Emre trying to stand up in the Atlantic would be less out of his depth....lol.

But if isuckatfm and Kobolok can do the hard stuff we might find out some useful things

Ok I did some simple tests using data Law_Man provided in the initial post. Since the data don't have specific attribute values I could not estimate an importance of each attribute. I also couldn't do analysis by positions as 2-4 players per position obviously not enough. However I did some trick (I am statistical analyst / modeler IRL, so trust me icon_wink.gif) to see if relationship between attributes sum and CA varies position from position.

So here some results. First, overal the relationship is clear and loud (no surprise given PaulC response icon_biggrin.gif)

Actual CA Predicted CA Difference Predicted CA Difference

(no position) from CA % (position) from CA (%)

Piatti 130 122 6% 120 8%

Kapo 146 150 3% 151 4%

Kaka 192 170 11% 173 10%

Ronaldinho 188 172 8% 183 3%

Torje 135 136 0% 141 5%

Ronaldo 192 169 12% 178 7%

Huth 152 164 8% 161 6%

Terry 182 176 3% 173 5%

Taylor 134 158 18% 145 8%

Evra 170 176 3% 165 3%

Pogatetz 144 154 7% 141 2%

Shawky 138 150 9% 149 8%

Mascherano 175 164 6% 164 6%

Young 135 135 0% 119 12%

Alves 179 199 11% 191 7%

Cattermole 132 148 12% 149 13%

Gerrard 184 183 1% 188 2%

Rochemback 147 148 1% 149 2%

Arca 142 147 4% 155 9%

Eto 186 170 9% 185 1%

Stancu 138 129 6% 139 1%

Average 158 158 7% 158 6%

The table above does not show clear evidence if the position affects the relationship, which is fine as I would expect specific attributes weights should vary from position to position, while total attribute sum shouldn't vary too much. However, some additional statistical evidence show that position plays important role so it's definitely worth to test it by attributes and positions. The differences between actual and predicted CA are most likely due to hidden atributes that are part of the equation but not accounted for. It opens up an opportunity to figure out something about hidden atributes without using so-called "cheating" tools while playing the game.

Another interesting thing would be to investigate relationship between CA, current attributes, and PA. I don't use FM scout or something like this to figure out anything about players (I feel it would take some fun out of the game), but I would definitely like to have better assessment of player's potential and hidden attributes.

Actually, I figured out how get the list of attributes for big list of players (thanks Cleon for teaching me how to create text files!). You simply go to player's search screen in FM, set filter as you want to, choose "view" attributes (physical, technical, etc.) then go to options and set "Print Screen" option as a text file. Yep, it would require a few files to be created for each list, but it's faster than writing them down one by one. So if FMM shows CA as part of player's attributes, I may download it and maybe do some additional analysis next week. If anyone knows better way to get bunch of players with attribues, CA and PA - let me know.

Link to post
Share on other sites

Oops, the table looked nice when I paste it, but now it's messed up. If Cleon could sort it out, that would be great. Otherwise here is the key:

- first figure after name is actual CA;

- second figure is Predicted PA when position is not taken into consideration;

- third figure is the differnce between first and second number in %;

- fourth number is Predicted PA when position is taken into consideration;

- fifth number is the differnce between first and fourth number in %;

Hope it helps.

Link to post
Share on other sites

Originally posted by trebor1185:

Just been doing a few small tests with the data editor, which has given some interesting results.

What I did was create a player who had CA + PA of 1, then gave him 20 for every attribute as well as set him as 20 for every position except for GK.

It appears that several attributes hold no/little weighting. The "test player" still had 20 for the 4 set piece attributes, 20 for aggression, determination, flair, and natural fitness. He had 13 for acceleration, agility, balance and pace. all the remaining attributes were 3.

Does anyone else think it's strange how the set piece attributes appear to have no/little weighting in relation to CA vs TATT?

But to be honest there useless without the other mental/technical attributes anyway.

Great work mate, very interesting!

I think it can be explained like this in terms of nature/nurture:

Some attributes are innate i.e. you have them or don't, as if you were born with them. For example, some people are determined, some aren't. Some people are ambitious some aren't. Some people are brave some aren't. You can't really alter these attributes (through training). Thus they remain regardless of CA.

Most physical attributes are probably part genetic and part training - for example you can always get quicker and stronger with the right training, and at the same time, some players are natural well built (e.g. Rooney) and some players are naturally very quick (e.g. Henry).

The rest of the stats i.e. mainly technical stats are all predominantly a result of nurture i.e. training. You learn to pass well, you learn to dribble you learn to cross well, you learn to take a good free kick.

So what is very interesting is that the set pieces attributes stayed the same! Beckham practiced every day after training to get better and better so this seems slightly illogical. However, you could try and explain it in the sense that during the process of becoming a player, all midfielders at a club go through the same training sessions and therefore the same exercises in training and all play matches, yet some of them seem to have a sort innate ability for set pieces. Granted that ability is grounded on non-innate skills i.e. the skills of technique and shooting or crossing that have been nurtured. So perhaps that explains it a little.

Personally I think it should be calculated this way:

corners - technique plus crossing plus decisions

direct free kicks - technique plus long shots plus composure plus decisions plus hidden pressure stat plus a little bit of finishing and also the "big games" stat too if it is in fact a big game.

indirect free kick - technique plus crossing/passing plus creativity plus decisions

KUTGW trebor, and post any other findings you make icon_smile.gif

Link to post
Share on other sites

Originally posted by kolobok:

Oops, the table looked nice when I paste it, but now it's messed up. If Cleon could sort it out, that would be great. Otherwise here is the key:

- first figure after name is actual CA;

- second figure is Predicted CA when position is not taken into consideration;

- third figure is the differnce between first and second number in %;

- fourth number is Predicted CA when position is taken into consideration;

- fifth number is the differnce between first and fourth number in %;

Hope it helps.

Great work Kobolok icon_smile.gif I've just skimmed it and am about to read it slowly now!

Just a minor point but shouldn't it be CA (above)?

Link to post
Share on other sites

The table above does not show clear evidence if the position affects the relationship, which is fine as I would expect specific attributes weights should vary from position to position, while total attribute sum shouldn't vary too much. However, some additional statistical evidence show that position plays important role so it's definitely worth to test it by attributes and positions. The differences between actual and predicted CA are most likely due to hidden atributes that are part of the equation but not accounted for. It opens up an opportunity to figure out something about hidden atributes without using so-called "cheating" tools while playing the game.

Another interesting thing would be to investigate relationship between CA, current attributes, and PA. I don't use FM scout or something like this to figure out anything about players (I feel it would take some fun out of the game), but I would definitely like to have better assessment of player's potential and hidden attributes.

Actually, I figured out how get the list of attributes for big list of players (thanks Cleon for teaching me how to create text files!). You simply go to player's search screen in FM, set filter as you want to, choose "view" attributes (physical, technical, etc.) then go to options and set "Print Screen" option as a text file. Yep, it would require a few files to be created for each list, but it's faster than writing them down one by one. So if FMM shows CA as part of player's attributes, I may download it and maybe do some additional analysis next week. If anyone knows better way to get bunch of players with attribues, CA and PA - let me know.

Good stuff Kobolok icon_smile.gif About half-way through my little experiments the though occurred to me about trying to add in PA to CA and TATT relationship but my lack of maths skills meant I didn't know where to start!

Feel to do any further experiments you think necessary and post future results here for us all icon_smile.gif

Link to post
Share on other sites

Sorry, I've just realised I've been spelling your username incorrectly! Apologies for that. It shall be Kolobok from now on!

icon_biggrin.gif Who cares? Kolobok is character from Russian kids story, so he is definitely not offeneded, especially since fox ate him at the end. icon_biggrin.gif

Link to post
Share on other sites

Originally posted by kolobok:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Sorry, I've just realised I've been spelling your username incorrectly! Apologies for that. It shall be Kolobok from now on!

icon_biggrin.gif Who cares? Kolobok is character from Russian kids story, so he is definitely not offeneded, especially since fox ate him at the end. icon_biggrin.gif </div></BLOCKQUOTE>

Ok cool, was just saying in case you took offence icon_biggrin.gif

By the way, here's the link to the 8.01 version of FMM:

FMM 2.23

Link to post
Share on other sites

originally posted by trebor1185:-

Does anyone else think it's strange how the set piece attributes appear to have no/little weighting in relation to CA vs TATT?

But to be honest there useless without the other mental/technical attributes anyway.

It's interesting you should note that as I was wondering the same myself and whether or not weightings for set pieces vary from position to position. My theory on it anyway is that using the model I have if there is no positional variance then there should be some consistency across the weightings for different positions.

originally posted by kolobok:-

Ok I did some simple tests using data Law_Man provided in the initial post. Since the data don't have specific attribute values I could not estimate an importance of each attribute. I also couldn't do analysis by positions as 2-4 players per position obviously not enough. However I did some trick (I am statistical analyst / modeler IRL, so trust me ) to see if relationship between attributes sum and CA varies position from position.

First off kolobok since you do this for a living throw your manners out the window. If anything I post about the maths is completely wrong let me know since I'm working off the little bits that remain in my brain of things I studied 10 years ago but never actually used in practice. Plus I don't want to plug through it when my fundamental methods are completely arseways.

Not to derail the thread but out of curiosity, what do you work on?

Also does using matrices to solve sets of equations sound familiar to you? There's a part of my brain that's telling me there's an easier way to solve those equations using matrices rather than step by step substitution. But I don't know if it's my memory mixing things up. I've been on the net trying to find a specific link but can't nail it down.

That's good stuff in your analysis. It really helps to see some numbers that appear to back up the assumptions of position affecting weightings. Out of intellectual curiosity what was your 'little trick'?

Another issue which could screw up my method and came to mind as I was picking MCs for using in the equations. What if my assumption of 'natural position' as the controller is flawed and it is in fact a weighted combination of all positional ratings? I might just stick with my assumption and tests the results. If they're completely wrong then that may point to it being false (or me not really knowing what I'm doing icon_biggrin.gif)

I have to say extracting the players attributes (even using the print screen method) has reminded me of what a low threshold I have for repetitive processes. Check natural position, Print screen to file, save as html, import to excel, rearrange........ I made it as far as 17 players and lost patience.

originally posted by kolobok:-

Another interesting thing would be to investigate relationship between CA, current attributes, and PA. I don't use FM scout or something like this to figure out anything about players (I feel it would take some fun out of the game), but I would definitely like to have better assessment of player's potential and hidden attributes.

On 07 I did look at this, not in any strict mathematical manner but on that version some hidden attributes appeared to be separated from CA. Also having looked at it on 07 using Genie Scout in that model there was no obvious link between CA and PA, except at the point of generation players with higher PA would often have higher CA.

On 08 I don't know as no Genie scout makes looking at these things too time consuming. I could be wrong but I don't believe that it is possible to do what I think you would like to be able to do:-

'Look at a players attributes from which you can estimate CA, then using both estimate PA without relying on coach reports'

Link to post
Share on other sites

originally posted by Law Man:-

Do you have results for the 17 players you tested? I don't think there's anything wrong with any type of results, positive or negative, they'll be useful to some extent

I'm not looking for general statistical things just yet. I'm looking to figure out the weightings.

In maths terms I have a set of one equation for CA involving 33 unknowns, the weightings w. So to solve for these I need 33 equations thus 33 players attributes plus CA to plug into those equations to figure out the weights. Basically 17 players is not enough to figure out the weights.

I don't know how much you remember about simultaneous equations from your school days, but you might remember finding x in terms of y to find the values for x and y. That's essentially what I have to do except I have to find all 33 weights rather than just x and y.

I hope that didn't come across as condescending icon_smile.gif

Link to post
Share on other sites

Originally posted by isuckatfm:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> originally posted by Law Man:-

Do you have results for the 17 players you tested? I don't think there's anything wrong with any type of results, positive or negative, they'll be useful to some extent

I'm not looking for general statistical things just yet. I'm looking to figure out the weightings.

In maths terms I have a set of one equation for CA involving 33 unknowns, the weightings w. So to solve for these I need 33 equations thus 33 players attributes plus CA to plug into those equations to figure out the weights. Basically 17 players is not enough to figure out the weights.

I don't know how much you remember about simultaneous equations from your school days, but you might remember finding x in terms of y to find the values for x and y. That's essentially what I have to do except I have to find all 33 weights rather than just x and y.

I hope that didn't come across as condescending icon_smile.gif </div></BLOCKQUOTE>

Not condescending at all, was just at my level! icon_smile.gif But glad you explained it like that as I get it now, and I remember doing simultaneous equations at school now, must have had to do them for GCSE, I actually got an A in maths, but god knows how! That's the state of the British education system for you....

Anyways, back to top - the point being that you've got to convince yourself to do the other 16 players whilst attempting to not lose the will to live.... icon_biggrin.gif

Link to post
Share on other sites

Back when FM Scout was around, I exported all the players from the database filtered by position and ran it through a spreadsheet to work out the dominant attributes for each position. Unfortunately I don't think I have those spreadsheets or the data anymore and as far as I'm aware there is no tool to export player data now icon_frown.gif If there was I could do the same thing again.

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...