Jump to content

TEST: impact of retraining on CA


Recommended Posts

Following a discussion with some members of the forum on what would be the impact of retraining on the CA of a player I realized that there were opinions on the subject but little data.

So I decided to conduct a little test to clarify the situation.

The starting hypothesis was the following: when players gain attributes it will cost them CA points, that cost is not the same for all the players and is driven by their positions. As an example it is more expensive for a defender to train tacking than it would be for a striker because that attribute is considered by the game to be useful for the position while it is considered to be much less useful in the case of the striker.

There was therefore a theoretical possibility to exploit that by retraining a player for a position very different of his. For instance you take a striker you teach him tackling which is pretty much a free attribute for him and then you turn him into a defender.

For that test you are going to take a player put him in a position different than his with the editor while erasing his natural position to simulate the perfect training that we could have achieved (like a natural defender with striker stats) and then we are going to reintroduce a position familiarity in the striker role up to natural.

I took one of the striker from my AJAX save. Based on his attributes when he is playing as a natural striker with no other roles he has a CA of 125. When I erased the striker familiarity to 1 and instead puts him as a natural defender his CA is 103. There is no magic here this is just the application of the different weights for the different attributes and it shows that our starting hypothesis is valid. 

Now this is what happens when I start retraining him as a striker

image.png.ffb340bcf746e0a24328058f5a049bef.png

What you see in the defender/strikers columns are the position familiarity for the role.

The RCA is what the game would see as the new CA, if that value is above the CA of the player we have 2 options, either the player is maxed out in terms of PA and he will has to lose attributes to compensate until he reaches back his CA, if he is not maxed out it will cost him CA in term of developement and will not be able to reach the same level of attributes than he would have had he not been retrained.

The delta is the impact on the RCA of 1 point increase in the similarity when we retrain him.

I also did another simulation with another player with better stats and a natural CA of 144 as a striker instead of the 125 that we have in the first set of data to see if the move are consistent for different player ranges.

What we can see is that retraining someone up until 8 is free but then some costs start to appear for the better subject when we reach 9. 

Then the cost of retraining is not linear with the bulk of the cost happening between 10 and 16 familiarity which means that is basically always worth it in a more normal scenario to retrain someone from accomplished to natural in a position as the cost for that upgrade is almost free.

We can also see that retraining someone with 2 opposite positions seems to take to worse of the 2 weights and not the average. It can be shown that when our subject was a pure defender he was CA103, a pure striker CA125 and when he was natural in both he was CA133. So retraining someone to an opposite position is never worth it from an optimisation point of view.

Also when you scout and you see players able to play different positions (even if they cant play them well) it will have an impact in terms of development vs someone that can only play 1. The more opposite the position the worse the impact will be. So for instance an advanced midfielder that can play as a striker will cost you virtually nothing, while a player being able to play as a DC and all the midfielder roles will cost a lot.

So the test clearly goes in the direction that some people had hinted to in their comments but at least now we have data to back it up :)

 

Edited by xavinwonderland
Link to post
Share on other sites

I'm still not entirely sure what it is you're wanting to prove or demonstrate @xavinwonderland across a few threads now you have had snippets from myself (Stoke Researcher, meaning I get access to some wonderful tools and information about attributes and positions) along with highly knowledgeable individuals like Herne and Enigmatic who have had a wealth of experiences that help inform their posts. Two strikers who are 6'2", 75kg and 130CA are capable of being incredibly different from each other. 

Most of us won't deal in specific numbers because honestly there's no benefit to it. Almost everything is so incredibly situational and will vary from player to player in the game that you can't establish a hard and fast rule.

-----

Your point about an advanced midfielder and striker having less impact is potentially nonsense because it very much depends on the player in question.

Which brings me onto the main point, these breakpoints your seeing changes are for players with these particular spread of attributes. You could try a different 125 CA forward and a different 144 CA forward and get different results. Because the CA really doesn't mean that much. 

Doing this in a sterile environment without the in-game corrections along the way removes some of the natural elements of progression or lower level adjustment that happens during retraining. So it looks more contrasting here and more standout than it would do if organically done through the game. 

It comes back to the aspect of what is it you're actually wanting to know or understand? If you straight up ask the questions the odds are members of the community can provide you with an explanation of how it works in the game. If the element of the game itself keeps being stripped out, like with the other test you linked to with the janky CA experiment which incorrectly concluded nothing but a small selection of attributes matter then the conclusions ultimately don't have any basis in reality.

Link to post
Share on other sites

Some observations:

1) Wouldn't it be better to study the impact on more likely retraining scenarios, such as a Striker to AMC?  Surely testing should be done with some reality in mind because nobody is going to retrain a striker as a defender.

2) Why are you looking at CA?  It's meaningless and is not a measure of how "good" a player is.  It doesn't factor in all attributes, nor does it account for things such as form, morale, fitness or how well a player is able to play in any particular tactic.  There was someone recently asking for help with a tactic, however when looking closer there wasn't anything much wrong with the tactic but his player choice - despite players having high CAs - in key areas was relatively poor and didn't really fit and combine well with the system.

3) You may argue that more positions = more CA eating thus a player's attributes may not grow as much if he only had one position.  Well being able to play in multiple positions doesn't seem to have done Messi any harm, or indeed any player who can play in multiple positions.  Isn't restricting players to play in one position potentially damaging to your club anyway because you then lack variety if you need to adapt?  For example playing a player out of position would (probably) have a greater impact on his ability to play effectively in your team than any assumed CA impact by retraining him.

4) How do you account for age or other natural attribute declines which are not age related?  It's perfectly possible to have a net zero (or even reduced) impact on CA during a player's natural development cycle if you introduce a new position because attributes can naturally decline anyway, which would be more than enough to "free up" CA for any new position's attribute calculations.  As you yourself say, retraining to a nearby position (lets say striker to AMC again) "will cost you virtually nothing" (although that statement itself is rather limiting as no two players are the same) so combine that with potential natural attribute changes and hopefully you can see how little CA actually means to us.

5) You don't min/max (which you mention in your other thread and I guess is what you are still trying to do) by looking at CA.  You min/max by carefully studying your system and developing specific player attributes, roles and positions accordingly.

6) You are drawing conclusions based on your very short and limited test.  Just like you did when telling us that pretty much all attributes are useless because some random on reddit ran a similarly limited (and nonsensical) test.  Don't draw conclusions based on such limited or irrelevant testing.

So yes, in your perfect microcosm you would see such movements as you describe.  But the actual in game world isn't anything like that so to draw such conclusions as you are I'm afraid is incorrect.  It's not as easy as 1+1=2.

I really don't mean to come across as harsh.  I understand you are trying to investigate and share something you would like to nail down and categorise.  But in reality it just isn't that straight forward.  As a researcher @santy001 is able to do things with a player within certain restrictions which would make our heads spin.  I remember him telling me about how he put together Marco Arnautovic's attributes to reflect the player he was at the time which was honestly pretty mind blowing.

So TL;DR - for your own peace of mind please stop looking at CA.  It's meaningless and is hidden for a reason.  Only SI know how it's calculated (despite other people's claims) and pretty much anything else you see written or view on YouTube about it stems from supposition and assumptions trying to make up for a lack of knowledge and is usually just plain wrong.

Link to post
Share on other sites

I am not trying to prove or demonstrate anything just trying to understand how the game performs certain tasks and for that we have to look under the hood. Some people need to understand how the calculations are performed and I am one of them.

When you want to understand something very complex you need to look at a simplification of what the reality this is why we are building models in economy or in weather forecasting. That model is not the reality but will allow you to draw conclusions. The more generic the conclusion is the more likely it is to be correct, the more specific it is the more cautious you should be with the result as it could be because of datamining even over very large samples. If you look at weather forecast for instance if based on a model we conclude something like winters are on average colder than summers it is quite likely to be the case very often. If we were to draw the conclusion that on average the 22nd or December is a more rainy day than the 23rd of December we would have to be very prudent with the result as there is a chance that it is just a random coincidence from the observations being made.

By choosing a more extreme test scenario we can see the trends better than by looking at fairly similar things (it is easier to spot a blue ball in a pack of red balls than picking a light red one in that same pack)

So clearly the test in an extreme limited scenario however I believe than the generic conclusions are likely to be valid even over a much larger sample size:

- retraining for an opposite position is most likely going to be a losing proposal in terms of player development (so unless you are desperate donc retrain a striker into a defender)

- the bulk of the retraining cost happens between 10 and 16 natural position

- the more positions the player knows, the higher the cost

- the cost is not an average but rather a worse off scenario where the game takes the highest weight for each of the known positions past a certain level of mastery. So the more different the weights the greater the impact will be.

If anybody wants to test these hypothesis further they are welcome but Im pretty sure that they will remain true most of the time (like with the weather example)

I dont really agree that CA is useless even if it is cleary not the main factor of how good a player is, some 140CA players will be better than some 160CA ones because they have their attributes at the right place or the right hidden attributes like consistency. However everything else being equal CA is important and once again the goal is not to find a formula for the right player/retraining but just understanding how the game performs certain calculations and in that CA is very useful.

Restricting buying players able too play many positions is (I think) the optimal play past a certain stage as what you lose is flexibility and what you gain is better (more specialized) player. I said past a certain stage because after a few seasons it is very easy to have a very large squad depth with a B team as much as good as the A team so once you reach that stage the flexibility gained is useless pretty much as you will never use it. For example I have benfica save where I am in my 8th season, won the CL multiple times and my A team is probably playing 50% of the game at most given than I need to play my youngsters on top of my B team.

 

Link to post
Share on other sites

Interesting to see the taper kicks in below positional familiarity 14 which is where I've generally noticed it playing around with editors.

That said, the differences between defender and striker are pretty extreme in terms of weightings.  If you attempt a more typical retraining (AML to ML) the gaps will be much smaller, as the weightings are quite similar. So your CA130 AML might be a CA133 M/AML. Is that worth worrying about? Probably not.

Generally if you're retraining a player it's because they'll be good in a similar position, not because their skillset is completely unsuited to any of the positions they currently play. The theoretical loopholes got closed a while back.

 

34 minutes ago, xavinwonderland said:

Restricting buying players able too play many positions is (I think) the optimal play past a certain stage as what you lose is flexibility and what you gain is better (more specialized) player. I said past a certain stage because after a few seasons it is very easy to have a very large squad depth with a B team as much as good as the A team so once you reach that stage the flexibility gained is useless pretty much as you will never use it. For example I have benfica save where I am in my 8th season, won the CL multiple times and my A team is probably playing 50% of the game at most given than I need to play my youngsters on top of my B team.

Unless you've created a very specific challenge for yourself, there's no limit to the CA of a player you can buy, so yes, a M/AMRC/ST might have a CA 15 points higher than a ST with similar attributes, but that's definitely not a reason not to buy him (just might be a reason why the scout is more excited by him than you are when you look at his attributes). He won't sign for me or he costs too much are reasons not to buy, but that applies to some non-versatile CA140 players at least as much as some versatile CA155 players

 

Link to post
Share on other sites

I'll try an analogy for you because I'm not sure you're understanding or even reading things.

Car manufacturers advertise fuel consumption figures.  However when we drive their cars we tend to get very different fuel consumption figures.  Further, 2 different people could drive the exact same car and get different fuel consumption from both the manufacturer's claims and each other.  Why?  Because the manufacturers are testing in a lab whereas we are driving in real world conditions and no two drivers drive in the same manner.

What you are trying to do is lab testing.  But that is very different from actually playing the game and developing players in game (in the "real" world).  Indeed you could have 2 players exactly the same but they will likely develop differently over time because of all the various factors involved - none of which has got anything to do with CA.

Does that help?

I'll also add this to your comment:

1 hour ago, xavinwonderland said:

I dont really agree that CA is useless even if it is cleary not the main factor of how good a player is, some 140CA players will be better than some 160CA ones because they have their attributes at the right place or the right hidden attributes like consistency. However everything else being equal CA is important and once again the goal is not to find a formula for the right player/retraining but just understanding how the game performs certain calculations and in that CA is very useful.

You seem to contradict yourself here.  You start by saying a 140CA player could be better than a 160CA player (which is absolutely correct) but then still say CA is important/very useful.  How?  You said yourself a lower CA player can be better than a higher CA player so how exactly is CA important?  By understanding certain calculations?  Which calculations?  What do you think CA is even used for?  One thing it is used for is to compare against PA to ensure a player doesn't exponentially develop beyond their potential.  How does that help you?  That little knowledge still doesn't help you in playing the game because - by your own admission - it still doesn't tell you how good a player is.  Players can still develop even if their CA has reached their PA because their attributes overall can change.   We're never going to know everything that CA does in the game's code because only SI know and they aren't going to tell us - and nor should they.  It has zero impact on our gameplay.

But you may say "aha!  How can you say it has zero impact when you also don't know what CA is used for".  Because thousands of people quite happily play the game without knowing CA even exists, let alone what it's used for.

So I'll ask a rhetorical question - now that you know CA is used as a measure to stop players exponentially progressing beyond their PA ceiling, how specifically does that make you a better player of the game?  I've known that for years and can honestly say it has zero impact on my development/transfer/tactical/any decisions.  Do you see what I mean now?  Sure we all might like to know the ins and outs of the game's coding but there really is no need to know because it doesn't affect our gameplay.  So when you say:

1 hour ago, xavinwonderland said:

Some people need to understand how the calculations are performed and I am one of them.

Then no, I'm afraid you don't "need" to know at all because, as we keep saying, it has zero impact on our gameplay.  You might like to know, but there really is no need.  All we actually need to know is the better the building blocks we put in place (coaches, facilities, tactics, training and so on) the more likely we are to improve our chances of success.  It's up to us then as managers to decide how to go about doing that.  If we knew the coding the game would be nothing more than a lego set with instructions to follow.

My honest advice to you is to drop this whole thing.  You're never going to crack the code and CA is a completely useless value to us anyway - you've even said as much yourself.  Attributes are important, not CA.  Focus on those :thup:.

Link to post
Share on other sites

The limit to the CA is what your scouts could find and what your money can afford but otherwise I agree.

The Holy grail of the experiment would have been to find a situation where retraining is either free or much less punitive because it would have allowed to de facto overcome the PA of a player. If your defender maxed out had a PA of 130 and CA of 130 but given his stats and the weight of those it would have been equivalent to 150CA for a striker. If retraining was free it would have been possible to make a 150CA player out of a 130PA player. 

There is no loophole there so kudos to the devs on that. 

For retraining costs kicking time I believe that 9 is not even the floor. It should be possible to find scenarios even more extreme where it starts to have an impact even before like maybe 6 or 7.

Link to post
Share on other sites

24 minutes ago, xavinwonderland said:

If your defender maxed out had a PA of 130 and CA of 130 but given his stats and the weight of those it would have been equivalent to 150CA for a striker. If retraining was free it would have been possible to make a 150CA player out of a 130PA player. 

With the exception that the game would equal out the attributes over a period of time. Since the player would be above his CA, in a manner of speaking, the attributes would slowly fall down until they are suitable for the CA. This levelling out happens all the time with player around their max CA. If you train a max CA player to get better in a specific attribute, it will increase, but at the cost of other attributes. The same would happen in your case, I imagine. So that means you will have a player over-statted for a while, before he turns into a more versatile player with fewer outlying attributes.

Link to post
Share on other sites

41 minutes ago, XaW said:

With the exception that the game would equal out the attributes over a period of time. Since the player would be above his CA, in a manner of speaking, the attributes would slowly fall down until they are suitable for the CA. This levelling out happens all the time with player around their max CA. If you train a max CA player to get better in a specific attribute, it will increase, but at the cost of other attributes. The same would happen in your case, I imagine. So that means you will have a player over-statted for a while, before he turns into a more versatile player with fewer outlying attributes.

Yeh and just to expand on this a little, when attributes increase and/or positions change affecting weightings (effectively pushing a player beyond max CA) other attributes across the board would tend to decrease to compensate.  So instead of one or two attributes reducing by 1 or 2 points, 20 might decrease by just 0.1 or 0.2 to compensate,  Thus any "negative" impact felt would by extremely small.  We can see that in game when we see a player with loads and loads of orange (not red) arrows pointing diagonally down across most of their attributes.

(It's not the only reason we see such arrows, but certainly one of them).

Another reason why "lab" testing is very different to "real world" testing.

Link to post
Share on other sites

Of course, I find the tests interesting, but it's not something that has ever bothered me. If I am looking to retrain a player, it's because his attributes suit what I need for the position.

Sometimes, it's also because he's 2nd or 3rd choice at his current position and with retraining he'll be 1st choice in the new position, while still getting better and getting more play time (to improve more) than where he was playing before. Other times, I have a weak spot and I have a perfect target in mind with the right attributes, but he doesn't play the position. So either I retrain this player, pay more for someone else or get someone worse.

When retraining, it's never something extreme either. An AMC as an AML/AMR or ST. A Defensive Midfielder as a Fullback or Wingback. A winger to an attacking fullback/winger. This also means that the impact of retraining will be less.

There are so many advantages that any disadvantage just doesn't matter to me.

Link to post
Share on other sites

Il y a 1 heure, herne79 a dit :

I'll try an analogy for you because I'm not sure you're understanding or even reading things.

Car manufacturers advertise fuel consumption figures.  However when we drive their cars we tend to get very different fuel consumption figures.  Further, 2 different people could drive the exact same car and get different fuel consumption from both the manufacturer's claims and each other.  Why?  Because the manufacturers are testing in a lab whereas we are driving in real world conditions and no two drivers drive in the same manner.

What you are trying to do is lab testing.  But that is very different from actually playing the game and developing players in game (in the "real" world).  Indeed you could have 2 players exactly the same but they will likely develop differently over time because of all the various factors involved - none of which has got anything to do with CA.

Does that help?

I'll also add this to your comment:

You seem to contradict yourself here.  You start by saying a 140CA player could be better than a 160CA player (which is absolutely correct) but then still say CA is important/very useful.  How?  You said yourself a lower CA player can be better than a higher CA player so how exactly is CA important?  By understanding certain calculations?  Which calculations?  What do you think CA is even used for?  One thing it is used for is to compare against PA to ensure a player doesn't exponentially develop beyond their potential.  How does that help you?  That little knowledge still doesn't help you in playing the game because - by your own admission - it still doesn't tell you how good a player is.  Players can still develop even if their CA has reached their PA because their attributes overall can change.   We're never going to know everything that CA does in the game's code because only SI know and they aren't going to tell us - and nor should they.  It has zero impact on our gameplay.

But you may say "aha!  How can you say it has zero impact when you also don't know what CA is used for".  Because thousands of people quite happily play the game without knowing CA even exists, let alone what it's used for.

So I'll ask a rhetorical question - now that you know CA is used as a measure to stop players exponentially progressing beyond their PA ceiling, how specifically does that make you a better player of the game?  I've known that for years and can honestly say it has zero impact on my development/transfer/tactical/any decisions.  Do you see what I mean now?  Sure we all might like to know the ins and outs of the game's coding but there really is no need to know because it doesn't affect our gameplay.  So when you say:

Then no, I'm afraid you don't "need" to know at all because, as we keep saying, it has zero impact on our gameplay.  You might like to know, but there really is no need.  All we actually need to know is the better the building blocks we put in place (coaches, facilities, tactics, training and so on) the more likely we are to improve our chances of success.  It's up to us then as managers to decide how to go about doing that.  If we knew the coding the game would be nothing more than a lego set with instructions to follow.

My honest advice to you is to drop this whole thing.  You're never going to crack the code and CA is a completely useless value to us anyway - you've even said as much yourself.  Attributes are important, not CA.  Focus on those :thup:.

I dont think we are talking about the same things. When you are talking about the car example you are talking about specific situations not generic ones. So yes if we both own the same car we will both have a different consumption and also a different one than what was advertized by the manufacturer. The lab will not be able to tell you how much you as a person will consume as it depends on many factors. Obviously the more details you give them the more accurate prediction they will be able to make.

However the lab will be able to formulate generic rules which would apply to everyone. For instance if you drive your empty car alone vs the same car loaded with 4 obese people and a trunck/boot full of bricks the later will have a very negative impact on consumption. Likewise if you drive much faster than usual because you are late for work it will also negatively impact your consumption. These generic rules can be found (and are well known) in every car test lab.

Then I think that there is a misunderstanding of what I say regarding CA:

I said that some 140CA players could be better than some 160CA one (and we both agree that this is correct)

However it doesnt mean that CA is not important. In fact everything else being equal it's probably one of the most important driver of performance. Take 1000 players with a 120CA and 1000 players with a 160CA, you will have more better players in the 160CA group than you would have in the 120CA group. This is the reason why Mbappé is not a 120CA player and why we all look for 5 stars youngsters and not 2 stars one. So some lower CA players can be better than some higher CA players but:

- this is the exception and not the rule

- the higher the gap between the 2CA the less likely it will be true (it's very easy to find a 155CA player better than a 160CA player, it is virtually impossible to find a 110CA player better than a 180CA player).

As stated the goal is not to make any point but just try to understand what is under the hood as I need to know from an intellectual perspective. Most people know how to use a microwave oven (even my 5 years old) without having any understanding of how it works and this is completely fine. Some people (most engineers will fall into that category) will need to understand intellectually how it works and at no point in time in the process they believe that they will become better users of a microwave oven. Call it intellectual curiosity?

 

Link to post
Share on other sites

Posted (edited)
il y a 23 minutes, HUNT3R a dit :

Of course, I find the tests interesting, but it's not something that has ever bothered me. If I am looking to retrain a player, it's because his attributes suit what I need for the position.

Sometimes, it's also because he's 2nd or 3rd choice at his current position and with retraining he'll be 1st choice in the new position, while still getting better and getting more play time (to improve more) than where he was playing before. Other times, I have a weak spot and I have a perfect target in mind with the right attributes, but he doesn't play the position. So either I retrain this player, pay more for someone else or get someone worse.

When retraining, it's never something extreme either. An AMC as an AML/AMR or ST. A Defensive Midfielder as a Fullback or Wingback. A winger to an attacking fullback/winger. This also means that the impact of retraining will be less.

There are so many advantages that any disadvantage just doesn't matter to me.

And that is completely fine. Im retraining players all the time myself. I was just curious as to what the cost of that retraining would be. Going forward it also reinforces a gut feeling that I had for a long time which is that players knowing a lot of positions typically dont end up being as good as more focused players. So it helps me scouting faster as I would really need a very good reason to sign a player like that now as this versatility has a probably a fairly high price tag associated to it (much higher than what I thought until now). Obviously this is my personal conclusion and some other players might consider that this guy is a must sign.

image.png.ec35fac22165e7fbdff7e163434dc4d7.png

Edited by xavinwonderland
Link to post
Share on other sites

Il y a 1 heure, XaW a dit :

With the exception that the game would equal out the attributes over a period of time. Since the player would be above his CA, in a manner of speaking, the attributes would slowly fall down until they are suitable for the CA. This levelling out happens all the time with player around their max CA. If you train a max CA player to get better in a specific attribute, it will increase, but at the cost of other attributes. The same would happen in your case, I imagine. So that means you will have a player over-statted for a while, before he turns into a more versatile player with fewer outlying attributes.

It's very hard to say what would happen if the game was working differently because precisely it would be working differently. At the moment what seems to happen is that when a player gets above the max CA because of the new position, the game will reajust the attributes based on the new weights in order to bring it back within his CA/PA which effectively acts as a ceiling. If retraining was free it would be possible that this rebalancing would not happen as technically the player would still be a 130CA defender based on his original weights so there would be no need to rebalance anything as the game would ignore the fact that he is now a better striker than he was a defender. But in any case this is not how to game works, rebalancing takes the most punitive weightings and as a consequence the game will bring the stats of the player in line with their ceiling.

Link to post
Share on other sites

il y a 56 minutes, herne79 a dit :

Yeh and just to expand on this a little, when attributes increase and/or positions change affecting weightings (effectively pushing a player beyond max CA) other attributes across the board would tend to decrease to compensate.  So instead of one or two attributes reducing by 1 or 2 points, 20 might decrease by just 0.1 or 0.2 to compensate,  Thus any "negative" impact felt would by extremely small.  We can see that in game when we see a player with loads and loads of orange (not red) arrows pointing diagonally down across most of their attributes.

(It's not the only reason we see such arrows, but certainly one of them).

Another reason why "lab" testing is very different to "real world" testing.

I think that you are touching on a different (even if valid point), the player profile only shows us round numbers but we do not know what the match engine uses. So if the game balances a new position by reducing some stats by 0.2 points we do not see it in the player profile but it might have a real reduction impact match engine wise.

Link to post
Share on other sites

1 hour ago, xavinwonderland said:

However it doesnt mean that CA is not important. In fact everything else being equal it's probably one of the most important driver of performance

CA doesn't drive performance.  Attributes drive CA and thus (partially) drive performance.  CA is just a sum of some of those attributes and imho opinion all of the most important attributes are not included at all (Determination and none of the hidden attributes).  Further, nothing else which contributes to performance such as form, morale, tiredness, tactical familiarity, team cohesion and so on are included in CA.  So how can CA drive performance?

1 hour ago, xavinwonderland said:

Take 1000 players with a 120CA and 1000 players with a 160CA, you will have more better players in the 160CA group than you would have in the 120CA group.

Quite possibly yes, although I think the figures you use are too far apart to be meaningful, but I understand the point.  However:

- Attributes are the drivers of CA, so focus on the things we can see not the things we can't.

- Not all attributes are included in the CA calculation, nor lots of other factors I mentioned previously.

The bottom line is some people like to focus on CA (once they find it).  And if that's how they want to play the game then no problem.  It's a convenient little number which may provide an indication that player x might be better than player y, rather than delving into all attributes, reading scout/coach reports and considering the likely fit of those things into their current tactical set up.  It's a quick and dirty way of going about things and sometimes that's all people want.  But if you really want to min/max as you say (which is where I'm coming from) then the way to do that is to do it the long way - not the quick and dirty way.

Anyway, one more thing then I'll leave you in peace lol:

1 hour ago, xavinwonderland said:

I think that you are touching on a different (even if valid point), the player profile only shows us round numbers but we do not know what the match engine uses.

It uses attributes (and lots of other things), that's why different players (due to their attributes) but the same CA can have a different impact on matches.

1 hour ago, xavinwonderland said:

Call it intellectual curiosity?

Agreed :thup:.

Link to post
Share on other sites

Posted (edited)
Il y a 2 heures, herne79 a dit :

CA doesn't drive performance.  Attributes drive CA and thus (partially) drive performance.  CA is just a sum of some of those attributes and imho opinion all of the most important attributes are not included at all (Determination and none of the hidden attributes).  Further, nothing else which contributes to performance such as form, morale, tiredness, tactical familiarity, team cohesion and so on are included in CA.  So how can CA drive performance?

I agree with the first part that CA drives performance through the attributes not obviously in itself as CA is just a number and I doubt that it is used in the match engine. 

However I disagree with the second points all the factors that you mention are indeed driver of performance but they are independent of CA. So you can manage morale or tactical familiarity for a 160CA player just as well as you could manage it for a 120CA player so you might as well get more attributes. 

I think CA in itself is  irrelevant because once again this is just a number and we cant see it. So very often talking about CA is just a simple way of talking about the sum of attribute points which is very easy to get. 

So I a way knowing CA is not even needed. It is easy enough to export into excel all the attributes of the up to 500 guys on the short list and do the sum. It will have a very strong correlation with what the CA would be. 

Ultimately I think (and this is my philosophy so other people could have a different view) scouting is a key aspect of the game as once again everything else being equal the better team will win most of the time (not all the time).

So scouting becomes for me a game of finding the needle in the haystack and sorting out the largest amount of data as easily as possible to narrow down the choices. Doing the sum of attributes with some weighting does a very nice job with that. I understand that I have a very brute force approach to scouting and that other players might like more finesse 😊

That being said I can usually identify the vast majority of the top regen in the world in a couple of seasons without any third party help so it is working well. But that would be for a topic on how to hardcore scout 😁

Edited by xavinwonderland
Link to post
Share on other sites

Il y a 2 heures, Baodan a dit :

@xavinwonderland You're a good dude and I appreciate the effort you've made here to illuminate a murky aspect of FM. All the while keeping an open mind and a respectful tone, which is something I personally fail at from time to time.

Many thanks. I always try to keep an open mind, a lot of people have very strong views about what FM is or should be but at the end of the day I think that only data matters :)

 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...