Youth players developing

I saw a tread earlier about youth players being rubbish (most of them), and someone said "Potential is relative". My questios it the following:

Let's say I take a rubbish player and, by surprise, I get him into my first team setup. I could actually do that in Ukraine, where there are lots of rubbish teams, so I could take the most rubbish youth pllayer and put him into the first team, because in most of the games the other players will take care of the game. Well, now, will he develop into a star or at least a medium-skilled player, though his potential was very low in the beginning of his youth career? If I play him for three years in my first team, alongside the likes of Chygrgynskiy, Pyatov, Thiago Neves, Douglas Costa, Alex Teixeira, Romelu Lukaku or Eljero Elia, will he become a real star, or he will never surpass the potential he had?

I think that is probably the difference between the way FM handles youth development, potential ability and how real life plays out. In actuality, if a team popped a youth teamer in amongst talented players and played him week in, week out, he would most likely make much more significant progress that his peers in the youth team would. In lots of cases, players look better and come on more being surrounded by more talented players

he will never surpass the potential he had?

This.

The PA is fixed from the moment they are generated. In my experience it is pretty easy for nearly every player to reach it if you make any sort of effort. If you get someone useless and throw them into the first team, they will improve quicker, but they will just peak early and stay there. They'll probably peak at about 20 as an average player at best.

Also, I've read that how well they player affects how much they develop. Bit of a catch-22 but apparently they need to get decent ratings to improve more. Obviously it is hard for a terrible player to get a good rating so just throwing in a rubbish player is not always the best idea.

IIRC, if a player starts with a PA of 120, he will never surpass that, no matter how many matches he plays at top level. Every player can only be as good as he can be (Lapalisse inspiration ).

Yeah but playing more match with them, allows their quick development to that level. Good for a mid table manager to have a utility player like that.

You could pop me into a team that contained messi, ronaldo, pele, kaka, maradona etc and I would still be a useless player. I might get to my maximum ability faster than if I merely sat on my bum and ate pies, but I still wouldn't become a great player. Each player on fm has a maximum ability, without cheating we don't know what that is, and if you play him every week he will probably get better, but he will still be limited by his potential.

Ok guys, thanks a lot for the answer

PA system needs an overhaul I think. It needs some sort of dynamicness to it.

PA system needs an overhaul I think. It needs some sort of dynamicness to it.

But why? Unless you use a cheat program you will never know what a players pa is and even having a high pa is no guarantee the player will reach that level, what would changing it add?

PA system needs an overhaul I think. It needs some sort of dynamicness to it.

Does RL need an overhaul as well then?

You could pop me into a team that contained messi, ronaldo, pele, kaka, maradona etc and I would still be a useless player. I might get to my maximum ability faster than if I merely sat on my bum and ate pies, but I still wouldn't become a great player.

No, but you will become better at your peak than if you sat on your bum. In other words, your "peak CA" is variable.

You have the potential to do much better playing with a better team than you do with a rubbish team (or by sat on your bum).

To me, this suggests PA should be variable. In fact, I go one further and say we scrap CA and PA altogether, and replace it with an attribute system unconstrained by CA (so no more silly weak foot weighting nonsense) and a talent level representing how talented a player is, the idea being PA should have a large bearing on where your peak CA should be, but that's not the end-of-all-ends - different factors affect where your peak CA will be (i.e. injuries and first-team football). What number your peak CA should be is actually not necessary - a bit like the amount of hairs you have when you have the most hair. You know you will have most of your hair as a young adult - but does it really matter how many hairs that is?

But what's the difference between 'talent level' and pa? There has to be some way of determining how good a player can be, otherwise you just buy any old Player and just train them until they become world class. Also , how is what you are describing any different to ca? Ca is merely a reflection of how good a player currently is, what ever system you bring in there will always be a 'current ability' even if it's bog given an actual score or is called something else.

But why? Unless you use a cheat program you will never know what a players pa is and even having a high pa is no guarantee the player will reach that level, what would changing it add?

You might not know his PA, but from my experience just by looking at his stats and looking at his report you can generally tell how good he is going to be. He can tell he is going to be great, okay or rubbish. How many times have you released a player a soon as he was promoted from your academy? I just don't think that's how it works in the real world.

Look at it this way, you have a decent player, you look at his report - 2.5 stars potential. You surround this player with state of the art facilities, great coaches and you tutor him, do you think he should stay at 2.5? Yes, sometimes, but not ALL of the time; like RL there should be exceptions which unfortunaltely there aren't in FM.

I'm sure plenty will disagree with me, and you're more than welcome to, but I just don't think potential should be set in stone the way it is.

But what's the difference between 'talent level' and pa? There has to be some way of determining how good a player can be, otherwise you just buy any old Player and just train them until they become world class. Also , how is what you are describing any different to ca? Ca is merely a reflection of how good a player currently is, what ever system you bring in there will always be a 'current ability' even if it's bog given an actual score or is called something else.

1) Talent level is like PA except it's not a limit of any kind. So it would be like a scale between the worst amateur player and Lionel Messi. Some players are simply more talented.

2) It doesn't mean anyone can be world-class - there may be computational limitations. For example, at the moment, it is nearly-impossible for a youngster with CA 1 to hit CA 200 even if his PA is 200. Why? Because the CA-PA difference is far too high - even the best-case scenario, where he gets the best training facilities and coaches, has no injuries and gets a rating of 10.00 in every game, bypassing the youth teams entirely, won't make him hit CA 200. It's too far.

The same applies here. If we remove a limit, it doesn't mean that every player will magically hit the limit eventually. Some players will never get that far - simple as.

The reason why I believe this system is more realistic is that where you end up is more than just talent. Talent is just one of the "ingredients". But several of these ingredients are unknown and don't stay constant - training facilities (i.e. a player moves to a better club) and first-team football (player moves, manager drops him to the youth team), for example. To me, that suggests that PA cannot be "known and fixed" if some of its ingredients are "unknown and variable".

As for CA, CA is just a weighted average and therefore computationally is unnecessary, processing power permitting.

The one flaw to saying that fixed PA is wrong is that we as users aren't supposed to know what each players PA is. Every player in real terms has a fixed ability, look at Robbie Savage for example he playes with the likes of Ryan Giggs, David Beckham and Paul Scholes yet he only turned out to be an average premiership player eventually. If you don't look at players PA there is no issue with it being fixed, because it is like this in real life.

And another point to make is that not all players will meet there potential it is only through good training that they do.

Right, so you wouldn't change anything because you just described the current system using different words. Pa is not a limiting factor unless you know what the pa is.

The one flaw to saying that fixed PA is wrong is that we as users aren't supposed to know what each players PA is. Every player in real terms has a fixed ability, look at Robbie Savage for example he playes with the likes of Ryan Giggs, David Beckham and Paul Scholes yet he only turned out to be an average premiership player eventually. If you don't look at players PA there is no issue with it being fixed, because it is like this in real life.

And another point to make is that not all players will meet there potential it is only through good training that they do.

I don't believe that "because you can't see it, there isn't a problem".

I'd also add that I don't really care what the number is - just that the fact it's fixed is a problem.

But from our perpective it's not fixed as we don't know what it is, it's only fixed if you go looking for it. Certain players are early developers, others late developers, but everyone has a maximum ability - no matter how hard rob green trains and practices he ain't gonna be as good as Joe hart. He wasn't ever goon to be as good from the minute he was born. But just like in real life we don't know where that limit is, removing that limit is simply ridiculous as potentially any good player can then be great, which is wrong.

But from our perpective it's not fixed as we don't know what it is, it's only fixed if you go looking for it. Certain players are early developers, others late developers, but everyone has a maximum ability - no matter how hard rob green trains and practices he ain't gonna be as good as Joe hart. He wasn't ever goon to be as good from the minute he was born. But just like in real life we don't know where that limit is, removing that limit is simply ridiculous as potentially any good player can then be great, which is wrong.

Who said that not having PA means that a player won't have a maximum CA?

Of course he will have a maximum. He will probably achieve it at about 27-29, or maybe later if he's a goalkeeper.

Joe Average will almost never become Lionel Messi because not everyone has a "high-enough CA" to begin with anyway - in the same way that Lionel Messi will almost never be as smart as Einstein because he isn't a mental genius.

As I said previously, there may be computational constraints. In addition, Rob Green won't develop much further as he's too old anyway.

With a talent level, you won't know the limit anyway - and neither will the game.

The game can also be designed such that you are more likely to win the lottery a thousand times than see your entire bunch of useless players turn into Lionel Messi. It leaves open such a question, of course - who knows where the next Lionel Messi will come from? After all, Messi's roots were with Newell's Old Boys, who aren't amongst the strongest teams in Argentina! Since you can never rule it out, if anything, it's more realistic to allow it to happen - but make it exceedingly rare.

Who said that not having PA means that a player won't have a maximum CA?

Of course he will have a maximum. He will probably achieve it at about 27-29, or maybe later if he's a goalkeeper.

But, but, but - that makes no sense! If he has a maximum CA then he has a defacto PA, you are just using different words for the same thing.

Also you keep saying that XXX wont become lional messi because their 'CA' wont be high enough to begin with - well that completely eliminates the chance of late developers - EG - Ian Wright - his 'CA' at 22 was very low - but by 26 he was international quality - having CA and PA can model this - what you describe is either no difference to the current system, or just plain broken from the start.

In real life is like the game, the PA is fixed but you dont know what is the PA of every player, like in FM unless you use some third edit tool.

But, but, but - that makes no sense! If he has a maximum CA then he has a defacto PA, you are just using different words for the same thing.

No. Talent is is not a limiting factor - big difference.

Talent helps determine what the peak CA will be - but it is not the only factor. Talented players will always in general do better than less-talented players, although notable exceptions exist (i.e. Fletcher and Roy Keane - perhaps not the most talented of players, but made up for it in other ways).

PA is the absolute maximum a player can be. Talent doesn't care what the maximum PA is - but it is used to determine what it will be.

Take Gran Turismo. They don't hard-code in the maximum speed for a car - they design the game such that under optimal conditions, the car acts like it is in real-life. The engine is so good the car is almost fully-realistic and under optimal conditions the maximum speeds are near-equal. The same logic applies here - the ultimate goal should be to remove the limit, and model how "maximum CA" should be obtained.

A player will clearly always have a maximum CA throughout his lifetime - CAs are never infinite. A bit like your bank balance over time - it will clearly peak somewhere, although nobody knows where (until you die). But it will peak - that's important. The reason why you probably won't become a billionaire is because the system - the world - has prevented you from achieving this status. Swap money for CA and the world for the Football Manager engine - the same thing applies here. Most players will never become elite world-class players, no matter what.

PA attempts to model this by creating the peak at the start of a player's lifetime. However, this is based on the circumstances of his beginning. If you have two 18-year-old players who are exactly the same in real-life, you can expect them to have similar career paths. But in Football Manager, if the players are at a Premier League club but one was signed from a Blue Square North side at 16, the game won't let this happen - because the former-Blue Square North player has a PA that was determined from the Blue Square North's reputation and facilities.

Put another way - if a Premier League player signed a rubbish Blue Square North player and allowed him to train with world-class facilities, players and coaches, then he would have a brighter future than if he had stayed at the Blue Square North side. He would probably still become rubbish, of course, but he would simply be more likely to have a higher "peak CA" than if he had stayed. Why is this? Because the training facilities develop players better! In other words, "peak CA" is a function of current training facilities!

So if a player moves clubs, his "peak CA" should change. If a player moves to a club with rubbish training facilities, expect his "peak CA" to be a lot lower than it was previously, because he wouldn't develop nearly-enough as before.

This is NOT PA. PA is fixed and is generated from starting circumstances.

Take a young player - say he has CA 60 but is reasonably talented. Through some analysis, we estimate his peak CA to be 90 because he is at a semi-professional club (but is one of the best youngsters to come through the club's academy in a while), and hence the training facilities are poor.

But because his talent (not estimated peak CA!) is reasonably good, he gets snapped up by a League One side. Suddenly, like a car having an improved engine, his development rate increases and the analysis estimates his peak CA to be 120 instead. Why? Because of better training facilities. He simply develops faster and his future looks a lot brighter.

Disaster. He breaks his leg and is released from his contract. He recovers and joins a Blue Square Premier side. Due to the injury and reduced training facilities, the game predicts a peak CA of 95 - because he still developed at the League One side, and hence his peak CA is going to be higher than 90 initially.

He scores a bucketload of goals and some weirdo in the Championship decides to take a gamble. He develops much better but first-team football is hard to come by. The game predicts a peak CA of 105 due to a lack of first-team football, but improved training facilities.

The manager is fired but the new manager takes a liking to the player. Suddenly he becomes a key lynchpin in the side, playing well with a new manager who has real faith in the player. His development skyrockets - his peak CA is now 130, about that of a lower-tier Premier League player. Various top sides start sniffing around.

The player falls into second-season syndrome, however, and quickly falls out of favour. The game predicts his peak CA to be around 120 - mostly because he is nearing his peak and he has less room to develop.

The player peaks with the side getting promoted to the Premier League, at a CA of 117.

If we had PA, 117 is the value that was supposed to be assigned to the player from the start as a PA. But could the game have guessed how this player's career would have turned out? No!

This system simply throws players into the game and lets them develop as managers let them. The actual peak CA is meaningless. Managers will always try to sign the most-talented players - this still exists in this system. Managers will always try to work with the best youngsters if possible, as they have a higher chance of making it - this still exists. Some youngsters will make it, some won't. The game shouldn't need to care about a predefined limit - the game will ensure that it is so well-balanced that you will not create a million Lionel Messis every season.

If I had to create a game to simulate your life, I'm not going to generate a limit for your money, lifespan, IQ and fitness at the start. I'm going to make my game so awesome that your life can be replayed again and again with vastly-different consequences, infinitely replayable. I am, of course, going to make sure that it is exceedingly unlikely that you will become the world's first quadrillionaire, because in reality this is extremely unlikely to happen. But I see no reason to hard-code this in - I'm just going to put your life into the game and watch it unfold.

Also you keep saying that XXX wont become lional messi because their 'CA' wont be high enough to begin with - well that completely eliminates the chance of late developers - EG - Ian Wright - his 'CA' at 22 was very low - but by 26 he was international quality - having CA and PA can model this - what you describe is either no difference to the current system, or just plain broken from the start.

Then the game must ensure that it is possible to have an Ian Wright-like career.

If anything, my system allows it more than the current one. If you find a purple patch like Luca Toni, your development will go through the roof and your "CA" will subsequently skyrocket too - late-bloomer, anyone? The current system puts a limit on PA - so for a real player, during a research phase, a researcher may set his PA to be too low to prevent the purple patch. My system, however, leaves open this possibility, but notes that it is probably not going to happen because the player is getting rather old.

I'm sorry, nothing you posted there is not already modelrd by a fixed BUT HIDDEN pa. The journey you described could equally be enacted by a player with a pa of 130. Just because his pa is 130 it foes not mean he will reach 130 ca, his time at the lower league clubs will limit this.

The problem with your reasoning is that you assume the limiting factor in a players development is the quality of the coaching and facilities. This is clearly wrong. South American players achieve extremely high technical ability not because they get wonderful coaching, but because they are truly talented.

I had a couple of players with half a star potential become established first team players for me in the Championship on my FM10 save. Their star potential rose as well according to the Assistant Manager's report.

I'm sorry, nothing you posted there is not already modelrd by a fixed BUT HIDDEN pa. The journey you described could equally be enacted by a player with a pa of 130. Just because his pa is 130 it foes not mean he will reach 130 ca, his time at the lower league clubs will limit this.

But by setting his PA at 130, you immediately rule out him getting a peak CA of over 130. What I am suggesting is that you cannot generate a value now based on current circumstances, because circumstances change. Or, if you prefer, you cannot predict the future. If I give you a different journey, you can suggest a new value for his PA. And if I give you another one, you can suggest another one. With PA, at some point you are going to have to pick a value that is sensible. With my system, the limit doesn't matter - just play out the journey, and his peak CA suddenly becomes a possibly boring statistic, a mark in the sand showing off how good he was at his peak. His maximum peak should be the case where he moves to the best club in the world, plays first-team football in every match (90 minutes), gets no injuries and gets a rating of 10.00 - if you like, this is the maximum computationally-possible CA. It is probably going to be really high. But then ask yourself - does this value actually matter? It is nearly-impossible to get this scenario - so is it seriously worth considering? If you pick a value for PA below this scenario, you immediately rule out the possibility of doing better - which is incorrect.

Instead of trying to limit a player's maximum ability, SI should try to model a scenario such that the range of scenarios is correct when factoring in probability. For example, there is a 5% chance he finishes in-or-above the Championship, an 85% chance he ends up in League One or Two and 10% chance below League Two. The actual peak values become a probability distribution that should look reasonable and balanced.

Even more succinctly: A player's peak ability is roughly determined by talent, hard-work, attitude, training facilities, coaches, adaptability, few injuries and a whole lot of luck. If you get all these "ingredients", you will get a world-class talent. If you are missing some of these "ingredients", you will not get a perfect recipe but you may get a good one nevertheless. This is a bit like youngsters who have rubbish attitudes or youngsters who never get world-class training facilities. Some "ingredients" are more important than others - adaptability only kicks in if you go abroad, and you could argue talent is the most important.

However, as a player's career progresses, he gets different "ingredients". New club? Different coaches and training facilities. Bad tutor? Attitude goes down the drain, and the recipe is ruined. Every time you change the "ingredients", how good a player will become must also change. You can estimate the best-possible recipe by getting the best "ingredients" but do you really need to bother? You are unlikely to get all the best "ingredients" to begin with - and if you do, congratulations! But you gain nothing by hard-coding in how good your recipe will be when you start cooking!

We can extend the limiting further. A player with passing 1 will probably never have passing 20 at his peak. So why don't we have a "Potential Passing" attribute that limits his passing to, say, 12? I mean, seriously, he's not going to go beyond passing 12. This is a stupid thing to do! You shouldn't need to put a limit on something like this. So why should you put a limit on CA in the form of PA?

This player will have a peak passing attribute eventually - say it turns out to be 10. Does the fact that he had a limit of 12 actually matter? In fact, does it really matter what the limit is at all?

The problem with your reasoning is that you assume the limiting factor in a players development is the quality of the coaching and facilities. This is clearly wrong. South American players achieve extremely high technical ability not because they get wonderful coaching, but because they are truly talented.

Talent is just one thing - look at all the South Americans who fail to adapt to the team ethos and physicality in Europe. My system doesn't have to assume a player's development is fully dependent on coaching - how about tossing in a new "ingredient" called "natural development"?

Everything you talk about X42 is "Maximum CA" and by definition this can never exceed PA as a player can never exceed his potential.

As have been said in many previous threads what needs improving is the way CA develops with less players ever reaching their PA limit.

But by setting his PA at 130, you immediately rule out him getting a peak CA of over 130.

Whoo, long post again. But sorry, it's all unnecessary as your entire argument falls appart in the first sentence. There is utterly nothing wrong with a player not being able to advance beyond ca130.

Everyone has limits, some people reach those limits, some don't. My pa in footballing terms is about 3. There is nothing wrong with that being set in stone because no matter how long I trained and how good the coaches were I would still be able to trip over my own feet when I ran fast.

Having a fixed pa does not limit the game in any way, we the player do not know what a players pa is, we know he's rated as a 5 star prospect but that could mean almost anything, pa of 200? 150? Or is the scout wrong and he's not that good. Then, how do you know he's going to reach those heights?

You seem to think that fixing the pa of players is somehow unrealistic and that I simply don't understand. If all players reached their pa then I would agree with you, but they don't, in fact most never do. Pa is te upper limit of a players talent, something which is very real, the difference between a premiership player and a conference player is not training or coaching but simply that those players were not as talented, nothing would change that.

Whoo, long post again. But sorry, it's all unnecessary as your entire argument falls appart in the first sentence. There is utterly nothing wrong with a player not being able to advance beyond ca130.

Everyone has limits, some people reach those limits, some don't. My pa in footballing terms is about 3. There is nothing wrong with that being set in stone because no matter how long I trained and how good the coaches were I would still be able to trip over my own feet when I ran fast.

Having a fixed pa does not limit the game in any way, we the player do not know what a players pa is, we know he's rated as a 5 star prospect but that could mean almost anything, pa of 200? 150? Or is the scout wrong and he's not that good. Then, how do you know he's going to reach those heights?

You seem to think that fixing the pa of players is somehow unrealistic and that I simply don't understand. If all players reached their pa then I would agree with you, but they don't, in fact most never do. Pa is te upper limit of a players talent, something which is very real, the difference between a premiership player and a conference player is not training or coaching but simply that those players were not as talented, nothing would change that.

I've stressed several times that I accept a player has a limit. I simply don't think it makes sense to generate one and hard-code it in by current circumstances.

If players in my system didn't have a limit, they would hit CA infinity - something I've said will not happen because the game must be balanced. (*)

This part was the most illuminating:

Having a fixed pa does not limit the game in any way, we the player do not know what a players pa is, we know he's rated as a 5 star prospect but that could mean almost anything, pa of 200? 150? Or is the scout wrong and he's not that good. Then, how do you know he's going to reach those heights?

You don't know! Just as in real-life!

You can train the player well and give him lots of first-team football. If he avoids injuries and has a bit of luck, then he will turn out to be absolutely brilliant. If you get Balotelli to tutor him and let him rot in the reserves, he will fall away and will dominate newspaper articles entitled, "Talents that failed."

All you know - and this is in real-life - is how talented a player is and how good their attitude is. It is up to you to continue their development.

His PA is immaterial! If his PA is 200 and he won't reach it, does his PA of 200 really matter? It only matters if they reach it and you wonder whether the limit was high enough to begin with.

Imagine PA doesn't exist. This doesn't mean there is no limit (see (*)). Then all you see is a 5-star talent with a high CA. How do you turn him into a wonderkid? You treat him well and make sure his development continues.

If Players A and B are in the database and have PAs 140 and 170 respectively, and both have reasonable attitudes, then it is difficult for Player A to be better than Player B at his peak. But since they are in the database, you never know what could happen in real-life to these players. Maybe B turns out to have a horrendous injury record and his PA will fall. Maybe A turns out to be a rough diamond whose talent was hidden all along. The PA system ensures, however, that this situation can never occur. My system, on the other hand, lets it happen because you may come across a scenario where A blossoms into an outstanding player to the extent that his peak CA vastly exceeds 140 - and this was never predicted in the original database. This is akin to a youngster like Pedro suddenly exploding onto the world scene as a truly world-class player from absolutely nowhere. Pedro in FM07 could never cater for this scenario because his PA was far too low. But if my system was adopted in FM07, you could give Pedro lots of first-team chances and design a formation so you could get the best out of his attributes - and there is a slight chance you could emulate his abilities today (or 4 years on).

But by setting his PA at 130, you immediately rule out him getting a peak CA of over 130. What I am suggesting is that you cannot generate a value now based on current circumstances, because circumstances change. Or, if you prefer, you cannot predict the future. If I give you a different journey, you can suggest a new value for his PA. And if I give you another one, you can suggest another one. With PA, at some point you are going to have to pick a value that is sensible. With my system, the limit doesn't matter - just play out the journey, and his peak CA suddenly becomes a possibly boring statistic, a mark in the sand showing off how good he was at his peak. His maximum peak should be the case where he moves to the best club in the world, plays first-team football in every match (90 minutes), gets no injuries and gets a rating of 10.00 - if you like, this is the maximum computationally-possible CA. It is probably going to be really high. But then ask yourself - does this value actually matter? It is nearly-impossible to get this scenario - so is it seriously worth considering? If you pick a value for PA below this scenario, you immediately rule out the possibility of doing better - which is incorrect.

Why is setting a static PA incorrect? The researchers or the regen system here are basically playing god. When you and I are born, we're probably capped at some point whether it's football or education. You and I weren't going to be Messi or Eistien no matter how hard we tried because of factors like genetics. This is reflected in the game through PA.

Instead of trying to limit a player's maximum ability, SI should try to model a scenario such that the range of scenarios is correct when factoring in probability. For example, there is a 5% chance he finishes in-or-above the Championship, an 85% chance he ends up in League One or Two and 10% chance below League Two. The actual peak values become a probability distribution that should look reasonable and balanced.

That's what the minus PA values are suppose to mimic for researchers. In terms of the regen system, they also produce PAs according to some probability chart.

Even more succinctly: A player's peak ability is roughly determined by talent, hard-work, attitude, training facilities, coaches, adaptability, few injuries and a whole lot of luck. If you get all these "ingredients", you will get a world-class talent. If you are missing some of these "ingredients", you will not get a perfect recipe but you may get a good one nevertheless. This is a bit like youngsters who have rubbish attitudes or youngsters who never get world-class training facilities. Some "ingredients" are more important than others - adaptability only kicks in if you go abroad, and you could argue talent is the most important.

Your talent here is basically PA. If you get a good PA and have the proper attitude and training, you'll end up a great player. If you don't have a good PA, no amount of hard work will help. Now, that's not to say that hard work isn't necessary. You can still vary your CA, just not above PA.

However, as a player's career progresses, he gets different "ingredients". New club? Different coaches and training facilities. Bad tutor? Attitude goes down the drain, and the recipe is ruined. Every time you change the "ingredients", how good a player will become must also change. You can estimate the best-possible recipe by getting the best "ingredients" but do you really need to bother? You are unlikely to get all the best "ingredients" to begin with - and if you do, congratulations! But you gain nothing by hard-coding in how good your recipe will be when you start cooking!

This should all affect development of CA. And this is reflected in FM through determination, ambition, playing time, tutors, training, etc...

We can extend the limiting further. A player with passing 1 will probably never have passing 20 at his peak. So why don't we have a "Potential Passing" attribute that limits his passing to, say, 12? I mean, seriously, he's not going to go beyond passing 12. This is a stupid thing to do! You shouldn't need to put a limit on something like this. So why should you put a limit on CA in the form of PA?

This player will have a peak passing attribute eventually - say it turns out to be 10. Does the fact that he had a limit of 12 actually matter? In fact, does it really matter what the limit is at all?

Not sure I understand you here. The passing attribute is weighted into the CA calculation. Ideally, it would make sense to have a potential value for every attribute just like not everyone will be able to jump like Michael Jordan or run like Usain Bolt. But in terms of putting that into a game, it would be too time consuming for the researchers.

It's fine having a PA number as they currently do, but I think they need to give more people higher potentials. But then to ensure there is still the same overall mix, make it harder to reach that PA. So give people lower determinations or whatever it is.

Because I think there should be more people that CAN get to Messis level but I think they should be more often limited by their own lack of determination or stopped by injuries rather than being stopped just by the preset number.

If Players A and B are in the database and have PAs 140 and 170 respectively, and both have reasonable attitudes, then it is difficult for Player A to be better than Player B at his peak. But since they are in the database, you never know what could happen in real-life to these players. Maybe B turns out to have a horrendous injury record and his PA will fall. Maybe A turns out to be a rough diamond whose talent was hidden all along. The PA system ensures, however, that this situation can never occur.

Again, this is where your argument falls appart, I don't know that one player has a pa of 140 and the other 170. All I know is that my scout tells me one player is 3 stars and the other 2 stars. The scout could be wrong, the 170pa player might have a really low determination and not reach his potential.

You have this bizzare idea that everything would be better if the 'restrictions' of pa were removed, but frankly how would that make any difference? I won't know that the player I bought has gotten that little but better than he would have otherwise been because I can't see how good he would havebeen in the first place, in the same way I can't see a players pa now.

The only argument that you produce in your favour is that 4 years ago si under rated a player and because they put in a fixed pa he was stuck with it. Under your system it would be no different, si would have under rated his 'talent' and thus he would never have gone past the ability he would have had with a fixed pa.

A fixed or moveable pa make utterly no difference UNLESS you use a scouting program and check it.

Even if a player doesn't have great PA, with good training and using him well, you can mould his attributes to make him very effective. PA doesn't take into account the distribution of the attribute points.

Interesting thread...

I was one of those that said "potential is relative" because in my experience, it is... much of what you've described x42 is already there in the game.

Without looking in the database (why would any real FM player do that?) we don't know what our players PA is... even when our scouts, coaches and assistants tell us that they're good, bad or indifferent, we still don't know anything.

We can look at their attributes and listen to what our staff tell us and make a judgement but we can still be wrong... our staff can still be wrong! If we 100% trust our staff to 'get it right' then we are likely to release players that have a higher PA than we think and they can suddenly appear in the opposition line-up as a star player! I've had it happen to me, several times, where I've released/sold a player whose potential fell short of my very high standards and a few seasons later they've turned up at a rival club as one of their star players!

The staff on FM aren't infallible, they might be very good, and they might have 20/20 for JA/JP but they can still miss the diamond occasionally. As mentioned in the previous thread, in RL we have the Ryan Shawcross factor - released by Man U as not good enough and now being touted for a move back, a big money move back (or to a rival). We have that in the game already! In my opinion it's not easy enough to miss the Ryan Shawcross's! Everything is geared to us being able to, for the most part, tell how good a player is and will be. We rarely make a mistake because the game is designed so that we can't, but in RL mistakes are made by even the very best managers and coaches.

If anything needs to change, it's our ability to discover the talents accurately (without using FMRTE or some other editor/scouting tool) that needs to change. It needs to be harder for us to spot and find the talents, even those within our own youth system. Obviously there still needs to be the Messi's, the Fabregas', the CR's and such, those players whose talent is outstanding at a young age and impossible to miss, but there also needs to be a lot more of the Shawcross's whose talents mature with age and experience - the slow growers.

I agree completely with x42bn6 on this. In reality, the concept of the glass ceiling just doesn't wash. Find me a player anywhere who thinks that they have no capacity for improvement whatsoever, or a coach that thinks that about them. That's not to say I think FM should do away with the concept of PA for the game, as it works for the game. At the end of the day, however realistic SI want the game to, it'll always be a virtual reality, and the system with never mimic reality perfectly. I think SI should maybe look at one day doing away with PA if they're confident the game can work properly without it, or if that doesn't work, then perhaps a dynamic PA could work, whereby over time if a particular player really starts to flourish then they're potential could slowly increase, and similarly if a player gets struck with recurring injuries, as their CA takes a nose dive, then perhaps their PA would also drop a bit too.

EDIT - Also, why does everyone on this forum think Ryan Shawcross is good. He's an average Premier League defender, nothing more, nothing less.

I didn't say he was good, I used him as the example as he'd been mentioned in another thread as being touted for a return to Utd... I don't care one way or the other how good or bad he is.. if he's being touted as a potential Man U signing then he has come full-circle (wouldn't you agree?)

Well, I will say yes, your rubbish newgen could develop into a star:

If he has a PA of 120, let's say (not rubbish, but not great), he could, theoretically, develop into a player with 120 CA. On the other hand, a player with 80 CA and 200 PA could hit 110 and never go beyond that- it all has to do with how they're trained. On FM10, I had players with great PAs develop into bench players, while I had players with lower PAs who came through at around the same time develop into first-team stars for my team.

Why is setting a static PA incorrect? The researchers or the regen system here are basically playing god. When you and I are born, we're probably capped at some point whether it's football or education. You and I weren't going to be Messi or Eistien no matter how hard we tried because of factors like genetics. This is reflected in the game through PA.

Joe Average isn't going to become the next Messi. But why do we need to throw a static PA in to ensure it?

Like I said, Gran Turismo probably doesn't have maximum speeds encoded in to its cars - but a Golf GTI will never be quicker than a Formula One car on the game.

That's what the minus PA values are suppose to mimic for researchers. In terms of the regen system, they also produce PAs according to some probability chart.

It still has the problem that it is a limit.

Assigning a PA allows the game to discriminate who will be better and who will be worse, for no reason why. My system will allow players to be become better or worse, but with a reason - how they develop.

PA predicts the future - removing PA allows the future to play out without the need for a prediction.

PA makes things easy for programmers - but it is less-realistic.

Your talent here is basically PA. If you get a good PA and have the proper attitude and training, you'll end up a great player. If you don't have a good PA, no amount of hard work will help. Now, that's not to say that hard work isn't necessary. You can still vary your CA, just not above PA.

Not really - talent is like a yardstick in the ground, while PA is a wall. Talent is just a rough figure for how talented a player is, while PA is a definite limit. The system I propose takes into account talent as well as various other things to modify how quick a player develops - how good they truly become is purely down to themselves and how the managers treat that player - including you.

For you as a manager, no matter what you do, the player will never exceed his PA, even if he is outperforming everyone in the league and is getting excellent coaching as well as having a fantastic attitude. He stops learning. With my system, he does keep learning and developing - although perhaps he will only do so rather slowly (maybe he's not that talented but makes up for it with his attitude, or he is injury-prone). Talent is just another one of those factors.

Like I said, we don't hard-code individual attributes' limits, so there should be no need to hard-code PA.

This should all affect development of CA. And this is reflected in FM through determination, ambition, playing time, tutors, training, etc...

But how you develop today will influence how good you will be in the future. Hence my "upgraded engine" idea - he's had a boost, and he will never look back.

It does develop your CA. But CA and PA are correlated and the CA-PA gap is meant to be kept realistic - if a player keeps developing a lot, then his future potential is greater. He's opened more doors.

If you like, a player is stigmatised by his generated circumstances in the sense that his PA will be determined by his starting club. And he will keep this PA from a possibly-rubbish club irregardless of how he develops in his career. This is like modelling the world in a video game and hard-coding the future income of a citizen of a third-world country to be very low. But what if this citizen gets into a good University on a scholarship and wins the Nobel Prize? His income will be stuck at the hard-coded level from birth.

Not sure I understand you here. The passing attribute is weighted into the CA calculation. Ideally, it would make sense to have a potential value for every attribute just like not everyone will be able to jump like Michael Jordan or run like Usain Bolt. But in terms of putting that into a game, it would be too time consuming for the researchers.

The reason we don't is because we trust the game not to develop players in an unrealistic way. It is difficult for a player to go from passing 1 to passing 20 - the game simply doesn't let it. In theory, you could place him on full attacking training for the rest of his life and as long as he gets first-team football and has a good attitude it may reach 20 by the time he peaks - but this will be at the expense of various other things, like more talented players.

Anything that is a limit can be removed by modelling why that limit is almost-never achieved. I can either introduce a speed limit for a Golf GTI in Gran Turismo, only to discover that I have a silly speed limit when plummeting down a hill because it's capped - or I can model the physics of the car and terrain so well that it is even more realistic. I could design the next Football Manager to generate the scorelines of a game before it's even begun. Nobody would accept this, because we can change the result of a game by tactical substitutions and changes. The same applies to PA! We can pre-generate PAs, or we can simply influence the current situation (i.e. CA, c.f. the current score and set of tactics in the 20th minute) to change the ultimate outcome (i.e. PA, c.f. the final score).

I may as well bold the last point because I know it's not acceptable, but the same reasoning applies:

- If we determine the scoreline of a game before it is even played, would you enjoy this game? The answer is no - because we can change the CURRENT circumstances (i.e. change tactics/put in a substitution) to influence the FINAL outcome (i.e. the final score).

This is very similar to:

- If we determine the peak CA of a player when he starts, would you enjoy this game? The answer is ??? - because we can change the CURRENT circumstances (i.e. give him more first-team football or loan him out) to influence the FINAL outcome (i.e. his peak CA).

- If we determine the scoreline of a game before it is even played, would you enjoy this game? The answer is no - because we can change the CURRENT circumstances (i.e. change tactics/put in a substitution) to influence the FINAL outcome (i.e. the final score).

This is very similar to:

- If we determine the peak CA of a player when he starts, would you enjoy this game? The answer is ??? - because we can change the CURRENT circumstances (i.e. give him more first-team football or loan him out) to influence the FINAL outcome (i.e. his peak CA).

Firstly, the game does determine the scoreline of a game before its started - we just dont know what that scoreline is. Everytime we make a tactical change it then recalculates that final score.

You make a horrible assumption with your second statement there. The game does NOT determine the peak CA of a player when the game is started, it determines the MAXIMUM CA of a player. This is totally realistic. Every player has a maximum potential. They dont know what that is, we dont know what that is but its there. The game then processes each player based on thousands of permutations, injuries, abilities, determination, games played, attitudes, coaching etc etc and this determines how a player develops.

If a player ever reaches their PA (ive never seen it when using genie scout on previous versions of the game) then they stop developing. So what? Unless you cheat to see it how does this affect you? Maybe that player wasnt ever going to get any better. Some good players have gone to big teams and never improved, the game models for this. Some bad players go to good teams and become great, the game models for this. Having a fixed PA has no bareing on the game at all unless you are somehow able to discover that fixed PA.

You seem intent that every single player should be able to get better, but this blatantly is not the case. - EG = Wes Brown is a good defender - hes had the best training facilities and best coaching - nothing is going to make him any better - hes reached his PA - there is utterly nothing wrong with that. Opposite example - Andy Carrol, young player, still developing, looks great, doing really well - is this his peak? Can he get better? Well in the game, who knows? I dont know what his PA is in the game - maybe he wont. Maybe he will be a one year wonder like Andy Johnson (oh he should have stayed at palace!). We just dont know unless we go looking behind the scenes.

SI do build in a wonderful variable PA system as it is - coded into the database are -1 to -10, therefore each game you play some players can end up great and in the next game they wont be and the KEY thing to this is that you, the player, have absolutely no idea - as far as you know, each player has limitless potential until you assess them, and just like real life, some players are better than others, and sometimes scouts get it wrong.

Reality arguments aside, ultimately nobody has given a convincing enough reason on why having a non-fixed PA in the game is so much better gameplay-wise that it'll justify a re-write of the whole newgen module.

It'll certainly help your case if you can expand on that point rather than concentrating on the reality argument, which IMO has pros and cons on each side.

Edit: Heh looks like Riz has made the same point in a more credible manner (him being a dev and all...)

Reality arguments aside, ultimately nobody has given a convincing enough reason on why having a non-fixed PA in the game is so much better gameplay-wise that it'll justify a re-write of the whole newgen module.

It'll certainly help your case if you can expand on that point rather than concentrating on the reality argument, which IMO has pros and cons on each side.

Thats been pretty much my point all along - but you put it more succinctly.

From where I am sitting I cant see how having a PA that is fixed is any different from having a 'talent' level and a starting CA. Sure with the second option that player could end up a few points better than the player with a fixed PA - but seeing as I didnt know what the players fixed PA was, how from a players point of view is that actually any better? Especially, as Riz pointed out, it would most likely need more processessing power to managed all those open ended abilities which would be detrimental to game play!

I had a couple of players with half a star potential become established first team players for me in the Championship on my FM10 save. Their star potential rose as well according to the Assistant Manager's report.

This is how the game leads us to believe that PA actually is variable. Provided we don't actually look at the numbers and stick to the proper in-game methods for scouting then we will always get players who do better or worse than we were expecting. The only problems with the model comes when you know the true values rather than the ones the game gives you.

If every aspect of players' skills would be translatable into stars or numbers IMO FM would miss a lot of atmosphere.

Those values should give a general evaluation but they have also to leave something to discover, both in bad or good.

If a player ever reaches their PA (ive never seen it when using genie scout on previous versions of the game) then they stop developing. So what?

Good post Maidel, but actually this part isn't quite right.Even with the PA reached, with good training to redistribute the attribute points, the player can still improve.

But why? Unless you use a cheat program you will never know what a players pa is and even having a high pa is no guarantee the player will reach that level, what would changing it add?

I agree without a editor cheat, one would not know the PA. it will solely depend on ours managerial skill to select and hone talent. The current system is good.

Good post Maidel, but actually this part isn't quite right.Even with the PA reached, with good training to redistribute the attribute points, the player can still improve.

Happy to be corrected on that, as Ive said, ive never actually seen a player reach their PA, so I have never been able to prove that one way or the other!

PA system needs an overhaul I think. It needs some sort of dynamicness to it.

I dont understand that statement. The PA figure is the maximum he can achieve and how high he reaches is the dynamic element within the game. Keep a youth with a very high PA away from first team footy and he'll never reach his potential. I hope this helps explain it for you

Mostly agree with Riz. Mind you, I have a brave thought that thanks to Moore's Law, CA and PA will be removed from the picture within 5 years. In 5 years we will be fiddling around with 8 (maybe even 12 or 16!) cores and 5 GHz processors as standard, after all.

Abolishing CA will allow things like managers and scouts to rate a player differently - so "CA" is dependent on who rates the player. For example, Wenger would rate a defender with passing 20 and jumping 12 a lot more than Big Sam, because Big Sam prefers big, strong players and passing is an afterthought - so Wenger might give him a "120" while Big Sam will give him an "80", for example.

PA system needs an overhaul I think. It needs some sort of dynamicness to it.

Couldn't agree more!

Thought I was the only one: http://community.sigames.com/showthread.php/244956-Regens-with-Unusual-natural-positions-for-their-ability...

I dont understand that statement. The PA figure is the maximum he can achieve and how high he reaches is the dynamic element within the game. Keep a youth with a very high PA away from first team footy and he'll never reach his potential. I hope this helps explain it for you

You suggest that we are all born into this world with a fixed innate football skill that we can either reach or not, but that can never be changed?