Is CA(Current Ability)/PA(Potential Ability) flawed??

[BoxToBox]

@TheMadSheep - Fine, but the point was, it usually takes until a player 'bursts' onto the scene to get his PA modified. Then again though, watching Van Gaal's United v Liverpool makes me half wonder whether positions should be eliminated from the game too. Ashley Young was pretty good in the WB position, doubt I'd try him there in FM though!

I do wish the positions deal was more flexible like that.

[BiggusD]

For the record, I think the CA/PA system works reasonably well and don't think it is flawed per se.

The problem is, as have been mentioned in this thread already, that we KNOW the PA (roughly). We shouldn't. We should only be able to see how good a player is currently, and based on age, personality and actual development over time (i.e six months), the scout would recommend him on these grounds.

The Curious Case of Felix Encinar

I have a youngster on my youth team, and he is currently loaned out to Málaga in the BBVA, where he is seemingly an important piece in the puzzle for the AI manager. His name is Felix Encinar, and this is him:

As you can see, as a 19-year-old he is already an entirely alright BBVA player (assistant says Decent BBVA), and that should indicate that he still has got 5 years with which to develop further. He could become really good if he continues his hard work.

Except, there is no hope for that;

He will maybe get 5-6 +1's spread around the key attributes for his position, and then that's it. If I couldn't see those two stars, that would have been fine. He didn't have it in him after all and I'd just move him on for an alright career in the BBVA, which I'm happy to do. However, I DO see those two stars. I have seen them all along! They used to be 2,5 stars but anyhow I always knew that he would never become good enough to challenge those I already have in his position. There was never any chance of him getting +3 everywhere from this point onwards (which would be necessary). Yet, everything points to him being well ahead of many of his u19 teammates currently, as he as always been. So what's stopping him?

Having a PA in the system is fine; letting us see it, however accurate, is not.

[PhroX]

BiggusD - Yep. Even if that is his potential, given his current ability for his age, your scouts/other staff should think it likely he'd go on to be a good player for you. He should be one of those players you have high hopes for but don't make it, rather than of one those that you know from the begining will never end up good enough for you.

[Loki679]

For the record, I think the CA/PA system works reasonably well and don't think it is flawed per se.

The problem is, as have been mentioned in this thread already, that we KNOW the PA (roughly). We shouldn't. We should only be able to see how good a player is currently, and based on age, personality and actual development over time (i.e six months), the scout would recommend him on these grounds.

The Curious Case of Felix Encinar

I have a youngster on my youth team, and he is currently loaned out to Málaga in the BBVA, where he is seemingly an important piece in the puzzle for the AI manager. His name is Felix Encinar, and this is him:

As you can see, as a 19-year-old he is already an entirely alright BBVA player (assistant says Decent BBVA), and that should indicate that he still has got 5 years with which to develop further. He could become really good if he continues his hard work.

Except, there is no hope for that;

He will maybe get 5-6 +1's spread around the key attributes for his position, and then that's it. If I couldn't see those two stars, that would have been fine. He didn't have it in him after all and I'd just move him on for an alright career in the BBVA, which I'm happy to do. However, I DO see those two stars. I have seen them all along! They used to be 2,5 stars but anyhow I always knew that he would never become good enough to challenge those I already have in his position. There was never any chance of him getting +3 everywhere from this point onwards (which would be necessary). Yet, everything points to him being well ahead of many of his u19 teammates currently, as he as always been. So what's stopping him?

Having a PA in the system is fine; letting us see it, however accurate, is not.

Great example. Have you checked what his actual CA/PA are just out of interest?

[Cougar2010]

For the record, I think the CA/PA system works reasonably well and don't think it is flawed per se.

The problem is, as have been mentioned in this thread already, that we KNOW the PA (roughly). We shouldn't. We should only be able to see how good a player is currently, and based on age, personality and actual development over time (i.e six months), the scout would recommend him on these grounds.

The Curious Case of Felix Encinar

I have a youngster on my youth team, and he is currently loaned out to Málaga in the BBVA, where he is seemingly an important piece in the puzzle for the AI manager. His name is Felix Encinar, and this is him:

As you can see, as a 19-year-old he is already an entirely alright BBVA player (assistant says Decent BBVA), and that should indicate that he still has got 5 years with which to develop further. He could become really good if he continues his hard work.

Except, there is no hope for that;

ybe get 5-6 +1's spread around the key attributes for his position, and then that's it. If I couldn't see those two stars, that would have been fine. He didn't have it in him after all and I'd just move him on for an alright career in the BBVA, which I'm happy to do. However, I DO see those two stars. I have seen them all along! They used to be 2,5 stars but anyhow I always knew that he would never become good enough to challenge those I already have in his position. There was never any chance of him getting +3 everywhere from this point onwards (which would be necessary). Yet, everything points to him being well ahead of many of his u19 teammates currently, as he as always been. So what's stopping him?

Having a PA in the system is fine; letting us see it, however accurate, is not.

Its possible that your staff are right and he has 2* potential.

Its also possible that your staff are wrong and he may continue to improve over the next five years. The only way to know this though is to keep playing him and see when his development stops.

[BiggusD]

@ Cougar and Loki;

My assistant has 20/20 jcpa and he is rarely wrong, not over several years and in that direction. I've got several *3.5+ potential youngsters who I am equally sceptical towards. The success rate of tutoring is too small for that. I don't normally check editor data but in this case I was 100% certain of the player's lack of potential and wouldn't have kept him even if it turned out to be 175... That would be cheating

[PhroX]

He should be getting a PPA in excess of 180, definitely not working as expected. The only reason I can see for him being given such a low PA star rating would be if the lowest CA rating in your senior squad is 190.

So does PPA include things like being good for his age and having a high reputation regardless of actual PA?

[BiggusD]

He should be getting a PPA in excess of 180, definitely not working as expected. The only reason I can see for him being given such a low PA star rating would be if the lowest CA rating in your senior squad is 190.

I don't know what the CA of my squad is. I've probably got the best AMC in the world, though. That could affect it. I know that 2.5 star = leading player and 3 star = star player. The league is rated 2nd behind Premier League.

[sosasoser]

Out of curiosity, have you tried to sell him to see who is interested? I had a very similar player and my ass man rated his potential to be about the same. I don't want to be tempted, so I don't have the editor and thus don't know what his true PA was. To test the waters, I offered him up to clubs for £5 million. Surprisingly, I received a ton of interest at that price, in the neighborhood of 30 clubs were interested. The player was sold to Bayern. I found that odd as if my 20 judging potential ass man was correct, then the player would be a no hoper for Bayern.

As Alex said there was a flaw in Encinar's PPA, I'm wondering if the flaw only exists for the player's current club and if AI clubs see the player correctly as having a high PPA?

[BiggusD]

There wasn't a problem loaning him out as far as I can remember. I set to autoaccept, in which case he chose the highest-rep club right?

[sosasoser]

There wasn't a problem loaning him out as far as I can remember. I set to autoaccept, in which case he chose the highest-rep club right?

Not necessarily. The loan system is a bit of a mystery to me. I've had players turn down being loaned out to teams such as Arsenal and Liverpool where they would have been cover for the first team to go to lesser clubs where they would be valuable first team members. Other players don't seem to care about playing time and want the highest reputation clubs. Still other good young players get no one interested in loaning them, presumably because the player wouldn't be interested in going to anyone who would want him.

[jujigatame]

Alex, can you please explain the "role ability" aspect a little more? This was one of the more baffling part of FM14 for me, the way my assman would recommend certain roles in extremely disproportionate amounts, even when other roles seemed far more appropriate for a given attribute set. Most of my strikers (even speedsters with atrocious crossing, and big/strong TM types) were recommended as AFs, just about every wide defender was recommended as a CWB, and DLP and B2B was frequently recommended for midfielders who seemed much better suited to other roles. Do you know why this is? Is it something that will be worked on for FM15 or is it something that is working as intended?

[BoxToBox]

Alex, can you please explain the "role ability" aspect a little more? This was one of the more baffling part of FM14 for me, the way my assman would recommend certain roles in extremely disproportionate amounts, even when other roles seemed far more appropriate for a given attribute set. Most of my strikers (even speedsters with atrocious crossing, and big/strong TM types) were recommended as AFs, just about every wide defender was recommended as a CWB, and DLP and B2B was frequently recommended for midfielders who seemed much better suited to other roles. Do you know why this is? Is it something that will be worked on for FM15 or is it something that is working as intended?

Almost every midfielder according to assman recommendation, are at their worst playing as a plain CM, which is odd.

[jujigatame]

Almost every midfielder according to assman recommendation, are at their worst playing as a plain CM, which is odd.

That's exactly why I ask. For the last couple of FM iterations, I've strongly felt that a lot of complaints would be alleviated if the assman didn't throw so much obviously problematic advice at you. Roles like B2B and CWB should only be recommended for top players, the fact that they're basically the default is baffling and makes you doubt the whole system.

[lil ole me]

I managed to get Barrow from the non league north to the championship over a number of years and i still had a young striker who is only valued at a good league 2 player but everytime i put him into the side due to injury or as a sub, he comes up with the goods....he has 9 goals in 14 games so far...if they produce the goods for my side then i dont care what star ratings they have.

[Super Bladesman]

@TheMadSheep - Fine, but the point was, it usually takes until a player 'bursts' onto the scene to get his PA modified. Then again though, watching Van Gaal's United v Liverpool makes me half wonder whether positions should be eliminated from the game too. Ashley Young was pretty good in the WB position, doubt I'd try him there in FM though!

Totally agree with the this - the concept of 'positions' in the FM sense is almost obsolete now. Certainly for any offensive strata, anyway.

If a player has the attributes and the positional awareness for a position they should be capable of doing a job there. FM has the data, it just needs to shift the emphasis away from the old-fashioned position and onto the roles/duties a bit more.

[Ohdear]

Getting back to the idea of taking a punt on a lower league talent - obviously due to the nature of the impression of players during research there are not going to be huge numbers of lower league players being billed as potential international superstars, but once you start to get into the regen market do lower league clubs sometimes get great potential talents through their youth intake?

I.e. If a regen at a league 1 club looks like he could be the business is he worth a punt or not?

[forameuss]

Totally agree with the this - the concept of 'positions' in the FM sense is almost obsolete now. Certainly for any offensive strata, anyway.

If a player has the attributes and the positional awareness for a position they should be capable of doing a job there. FM has the data, it just needs to shift the emphasis away from the old-fashioned position and onto the roles/duties a bit more.

But then that would be an absolute sea change in how things are done. I'm not saying it's the wrong way to go, but that seems to me like several years development. The ME would change as we know it, and even if that happened, it sounds like one of those things that's easy for us to discuss and model, but not so easy for the ME to do so.

Who knows though, they could already be working on the next generation of the ME, to be released in several versions time. Maybe this is part of it. But by then football could have moved on again, who knows.

[BiggusD]

He should be getting a PPA in excess of 180, definitely not working as expected. The only reason I can see for him being given such a low PA star rating would be if the lowest CA rating in your senior squad is 190.

Could the fact that he is loaned out have anything to do with the unusual star rating accuracy?

[themadsheep2001]

That's because both camps are right.

Generally people pick and choose who to develop and tutor and force through to reach their max potential. Dare I say, some will be checking these guys to ensure they have high PA in the first place.

Everyone else, gets thrown on the backburner and they either become useful backups or get sold (i.e. they don't develop).

And since plenty of players seem to aim for the 5star 180+PA wonderkids then it's no surprise that the majority of players don't develop.

On the AI side of things, players rarely develop well imo.

@TheMadSheep - Fine, but the point was, it usually takes until a player 'bursts' onto the scene to get his PA modified. Then again though, watching Van Gaal's United v Liverpool makes me half wonder whether positions should be eliminated from the game too. Ashley Young was pretty good in the WB position, doubt I'd try him there in FM though!

Actually, they will get it for the months proceeding well. Researchers dont just wait for the first team, they will be watching the youth and reserve teams too. For example, if James Wilson gets adjusted up for United, it wont be for the goals at the end of the season, but the way he had been tearing through the Under 18s and Under 21s too for the whole season. That goes for the likes of James, Keane, Lingard as well. Researchers take a longer view than people often give them credit for.

[eastley]

Removed because of posting CA/PA..

[PhroX]

Almost every midfielder according to assman recommendation, are at their worst playing as a plain CM, which is odd.

I always like the full backs who's best role is Limited FB and second best is Complete WB, with normal FB and WB coming in thrid and fourth...

[Stevenwonderboy]

If a player have a CA of 180 and another player have a CA of 160 it do not have to mean that the player who have 180 in CA will be better. It all comes down to how his attributes are distributed.

In this context I would like to mention that players PA is not permenant. By improving the players via training it is possible to exceed their PA. This goes the otherway around to, which mean that your players PA can go down aswell.

Hope this sort something out.

[Cougar2010]

If a player have a CA of 180 and another player have a CA of 160 it do not have to mean that the player who have 180 in CA will be better. It all comes down to how his attributes are distributed.

In this context I would like to mention that players PA is not permenant. By improving the players via training it is possible to exceed their PA. This goes the otherway around to, which mean that your players PA can go down aswell.

Hope this sort something out.

Just to clarify PA as a fixed number is set when a player is created. 99% of the time this never, ever changes although there is a rumour that a serious injury can reduce it (I don't think this has ever been proven or disproven though).

I think what Steven is referring to is that a bad spread of attributes or hidden attributes can lead to a player being worse than his PA suggests whereas a good spread of attributes or hidden attributes can lead to a player being better than his PA suggests.

The stars shown by the staff members are not PA, they are the staff member's opinion of the players PA (Generally referred to as PPA) and this can alter and change over time.

[Stevenwonderboy]

Excactly! Although it is possible for players to increase their PA. Im so sad that i did not take a god damn screenshot of it.

I've tried several times where the etc. the PA is at 180 and then some years after, due to traning etc. I see the PA have been increased. The amount the PA increases are ofc different for player to player.

I will for sure remember to take a screenshot next time!

[Ackter]

PA never increases (or shouldn't) so it's either a bug or you've got something mixed up.

[Cougar2010]

Excactly! Although it is possible for players to increase their PA. Im so sad that i did not take a god damn screenshot of it.

I've tried several times where the etc. the PA is at 180 and then some years after, due to traning etc. I see the PA have been increased. The amount the PA increases are ofc different for player to player.

I will for sure remember to take a screenshot next time!

PA can never, ever increase. The star ratings can but they are not PA.

Given you can't see PA within the game how did you know the numbers? If you used 3rd party software its possible there were bugs/errors with it.

[Stevenwonderboy]

I've used Genie Scout. It have only occured with regens and not the "real" players.

[Loki679]

The SI ingame editor shows PA as far as I know (I don't have it but I remember in another thread someone checked the PA of a player with it I think).

edit: It was this thread BiggusD checked his player with it.

[aderow]

I remember a few years ago the problem people had with the game was a lack of late bloomers and the lack of players capable of playing at the top that come from the lower leagues (regen or otherwise).

I understand why though. At least for real players. Their CA/PA is determined by the researchers and is specific to a particular game. In my experience, I haven't seen 'late bloomers' in my saves. Tbf, that's due to me playing with Manchester United; I generally work with top youth talent. But I have had my scouts and coaches wrongly determine a player's potential. And I have found an absolute stud in the lower leagues before.

I had a player who was written off by my staff so I paid no attention to him. I loaned him out for 3 seasons when he was old enough. 5 years on, he became a top class FB starting for me and getting called up to England.

The player I found in the lower leagues was a striker. He was playing for Bury and was scoring 40 goals a season. He played so well I called him up to the England U19s and he kept scoring. When I became U21s coach, I called him up again. And he kept on scoring. When I became the coach of the senior side, I called him up again. And he scored on his debut. I eventually took him to the world cup and he scored 3 goals including the goal that sent us to the final. And all of this happened while he was still at Bury! He didn't sign for a bigger club until over a season later.

Scouts/coaches make mistakes. There are a ton of factors that influence development. I think if people start to focus less on the numbers and focus more on players (more specifically what they think makes a good player), they will have less issues with the current system.

Heck the last player to make it to the first team in my United save in FM10 was a kid with a PA of 146. And he looked like he'd been playing with the first team squad for years.

[jamez_b]

If a player have a CA of 180 and another player have a CA of 160 it do not have to mean that the player who have 180 in CA will be better. It all comes down to how his attributes are distributed.

Wasn't this though disproved by the post earlier where they took high CA players and put low stats in the key attributes? Showed that the CA seem to matter more how a player performed than the key attributes.

[Wonky]

Wasn't this though disproved by the post earlier where they took high CA players and put low stats in the key attributes? Showed that the CA seem to matter more how a player performed than the key attributes.

I remember that post. I'd love to see more research on this.

[gianmixx]

Wasn't this though disproved by the post earlier where they took high CA players and put low stats in the key attributes? Showed that the CA seem to matter more how a player performed than the key attributes.

if confirmed, then the actual pa/ca system is totally flawed.

Btw

CA and PA must not INFLUENCE other stats like Reputation, Value, Match rate, etc because in REAL noone can be sure about CA/PA of a player, can only speculate on it!

Im an old FM player, and I can understand very easy if a player have an high or low PA without any editor, because the behaviour of the game spot it easily.

[Wonky]

The player I found in the lower leagues was a striker. He was playing for Bury and was scoring 40 goals a season. He played so well I called him up to the England U19s and he kept scoring. When I became U21s coach, I called him up again. And he kept on scoring. When I became the coach of the senior side, I called him up again. And he scored on his debut. I eventually took him to the world cup and he scored 3 goals including the goal that sent us to the final. And all of this happened while he was still at Bury! He didn't sign for a bigger club until over a season later.

I would love that to happen in real life. :-) Great story.

[magicmastermind124]

I will always remember my half-star to 1 star striker who would go on to score 25 goals a season, once being the top scorer in the league. His partner, a 5 star striker, would score over 20 but never more than the 1 star player.

[jamez_b]

Had Kozak on a previous iteration -- I think I've barely had a forward who had this high a header completion rate on the level I was playing then. It's up to you to utilize that.

That said, the attribute system on FM can make for weird results if you start (extreme) experimenting. You'll find that heavily modifying key attributes you associate to bookings and tacklings won't alter much.

That low finishing and composure doesn't keep players from being top scorers, nor having very high SOT ratios (60% and up)

Or that key attributes you would associate to passing don't massively let players down at all, columns left to right: appearances, pass success ratio, first touch, passing, technique, decisions.

Additionally, ulta low work rate and determination won't offset a player's running statistics (distances), even players with "1"s in both can make top of the league, as your tactical setup has a far greater bearing on that (roles/duties that encourage lots of runs, generally aggressive attacking tactics, look for overlap encouragements etc.). Things to bear in mind are that: Can't find the post but I remember Paul Collyer, the chief coder of the match code, stating that FM would assume any footballer taken from the database to be "decent footballers". edit: I found it via google, heh. http://community.sigames.com/showthread.php/283877-FM2012-difficulty.?p=7379345&viewfull=1#post7379345

There is also a post from Marc Vaughan I found by accident when searching for something, who I think is now working mainly on the handheld iterations, and who in 2004 said: "I tend to develop AI by making a 'perfect' attack/defense and then introduce human errors - this has two advantages, (1) easy tuning/fixing of problems (ie. you 'know' that any failures are 'bugs' and not simply a player performing poorly), (2) you can introduce human characteristics and (hopefully) ensure that failures are done in a realistic manner." If the match sim of the main game is done in this way, big if, this was ten years ago, this means that the occasional error would be more likely to happen with lower attributes, not a generally all low performance. At key moments there could be a "fail check" that makes a pass go astray, rather than a player frequently failing to pass the ball.

Still you might be wondering if the researchers are given guidelines that would make them insert the data which is suitable to produce realistic behavior, if possible, within the match sim. To an extent, the attributes aren't as straight forward as they seem, at least some. Later in the game this could also beg the question whether the attributes are weighted in this heavily -- or whether they're still weighted against the overall CA in a second or first step to determine an overall level of performance long before individual low attributes could make some damage. But then this is experimenting -- in the game you'll generally see players topping the passing/assists/heading/scoring charts you generally would expect to do. Yet some individual attributes don't have the influence you would expect them to have in isolation. Players with a general level of poor attributes (which inevitably means a low CA) perform horribly, that is understood, whilst those that still have a high overall CA appear still decent passers despite their individual attributes for the area being very very low. I don't know whether that is a revelation, it certainly is but a limited and flawed extreme experiment. But it might make you think the next time you try to assess a player based on his individual visible attributes alone when technically poor players can still make the best passers of any top flight. (Everything run on full detail match simulation).

It also begs the question to what extent you'd be better adviced to look at your scout's PPA and CA ratings, and whether some attributes are really that important (why look at technique when keeping possession can be done with any type of player?). Going one step further, you might ask what is genuine and what is mere "cosmetics" to make the game appear more complex and in-depth than is. But those are ones for SI to tackle who no doubt won't give anything really away (as they never do). ;-)

The post was made by Svenc earlier in this thred.

[Ackter]

Tbf, even with those attributes in FM14 I'd still expect him to be a leading goal scorer. Perfect pace, acceleration, off the ball, balance, anticipation, bravery, decisions, flair, dribbling, first touch, long shots and technique equals a hell of a lot of good chances every match.

He's not playing well because he has CA, he's playing well because he's perfect in so many important ways.

[jamez_b]

Ah my bad i hadn't thought to look a Messi's age and that it's leadership now and not influence. Apologies and ignore me wasn't clear from the original post.

[aderow]

I would love that to happen in real life. :-) Great story.

It really was. I was crushed when my hard drive corrupted and I lost my save (as well as everything else) before I got the chance to sign him.

[Stevenwonderboy]

Stats are more important than CA/PA.

It has so much to say whether their attributes are distributed so it fit the players playing role.

For instance if you have a good striker who have a CA at 180. Then lets say he got 19 in tackling. These 19 points of stats could have been used better as he is a striker.

The players stats should fit the role they are playing. So yeah stats are more inportant than CA, it all comes down to how they are distrubuted and how well they fit the players role and position.

[rand21648]

Stats are more important than CA/PA.

For instance if you have a good striker who have a CA at 180. Then lets say he got 19 in tackling. These 19 points of stats could have been used better as he is a striker.

Actually, the difference between 1 and 20 in tackling for a pure striker in FM14 is about 2 CA points - similar to 1 point (out of 20) for his weaker foot and probably less than 1 point of acceleration. It's practically more or less a free attribute for them.

------------------------------------------------------------

Continuing my experimenting with abusing the system with positional attribute weightings to create low CA players with monstrous key attributes, I have now managed to use the editor to create a CA/PA 85 player who became the top scorer of the English Premier League with 29 goals in 36 games in my first test season with him.

[santy001]

Attributes in the right place can make an enormous difference, and it's nowhere near as effective as it used to be in terms of shifting weightings with various positions - but it still plays a role. There are a lot of subtle differences that just notching a player up a few rating points in a position re-balances weightings by quite a bit.

The CA/PA system is a lot more complex than most players give it credit for, it's not perfect - I've seen some suggest there should be a CA/PA for technical skills, mental abilities and physical ability but to a large extent we can do that already - we can spend a large amount of a players CA on physical attributes or technical attributes if that is where their strength lies.

[Timych]

Continuing my experimenting with abusing the system with positional attribute weightings to create low CA players with monstrous key attributes, I have now managed to use the editor to create a CA/PA 85 player who became the top scorer of the English Premier League with 29 goals in 36 games in my first test season with him.

Can you post screens of him in-game and in-editor? I wanted to make a test with 90 ca/pa team to see if it could win something.

[rand21648]

Can you post screens of him in-game and in-editor? I wanted to make a test with 90 ca/pa team to see if it could win something.

His weak foot is rated 12, positions are: SW (20), CD (20), STC (12). Versatility, Consistency and Important Matches are maxed, as are all personality attributes except controversy. Injury proneness and dirtiness are 1.

The attribute and position distribution is not that thought out and certainly far from optimal. Actually it was a clear mistake to have a positional value for SW added because without it his RCA (the CA corresponding to the attributes) is just 71. With more carefully considered attributes and a tactic made by someone like Cleon (I played him as advanved forward in the game's default 4-4-2 tactic with no adaptions whatsoever) I think the results could be even more impressive by far.

Perhaps this abuse of the attribute weighting system is a bit silly because neither the game nor the data researchers creates pathological players like this, but I do think there is something slightly hilarious with having a CA 85 (or 71) central defender who is the most lethal striker of the EPL...

I have also played around a little with creating a similar monster central defender. Here I find that it seems advantageous to set postions in the AM strata as natural. STC seems to be slightly worse, because Jumping and Heading have high weights for strikers.