Training format + Dynamic player potential.

hello for all.

1. I didnt find any "suggestions forum" or something like that..so im sorry if its not related to this forum.

2. I know that maybe its too late to talk about it or maybe some other people suggest it before me, but there is 2 things i want to see in fm 2011 and maybe if it will be discusses in here, then it will be in fm 11\fm 12.

about the training:for some people [like me] the training part is really boring and take a lot of time(hiring staff,make special trainig for each player etc.) ,but on the other hand i still want my players to fulfill their potential...so what im suggest is to add an option like "automatic training system"(lets call it ATS)and all you need to do is to invest money in it,for example: if you will invest a lot of money in the ATS, then the club will hire the best coaches that will want to join your club and will set the best trainings for each player and vice versa if you will invest 1$.

about the dynamic player potential,im sure that its already discussed in here,but in the real world the players potential can change a lot throughout the years,if a young player had a great season his potential should be grow a little bit and vice versa.

p.s. the dynamic league is the savior for players who likes to play in leagues like poland,norway,israel,austria etc..

thanx for reading.

For future reference there is a search button on the forum. Just above where the threads start on ther right hand side. A search for 'FM11 suggestions' brings up pages of threads.

I agree about dynamic potential to an extent, although it would need to be implemented properly.

It, for instance, seems very unusual to me that players can peak at 18/19, or even be generated at near to their PA, and that then no amount of training or experience can help them improve at all. Players always talk of constantly improving, so it seems weird to me that in the game this simply cannot happen. While a player maybe is an early bloomer and may see that they have little to improve, in my eyes you can never be perfect at something and training can only ever help you become even better.

Training for most players in the game seems to become a case of attribute maintaince rather than attribute development, which simply seems flawed, at least for players who aren't 'older' anyway.

Is the world's best striker training his finishing so he doesn't forget how to, or so he becomes even better?

StevoRobbo , thanx but i still cant find it..can you give me a link?

about the training:for some people [like me] the training part is really boring and take a lot of time(hiring staff,make special trainig for each player etc.) ,but on the other hand i still want my players to fulfill their potential...so what im suggest is to add an option like "automatic training system"(lets call it ATS)and all you need to do is to invest money in it,for example: if you will invest a lot of money in the ATS, then the club will hire the best coaches that will want to join your club and will set the best trainings for each player and vice versa if you will invest 1$.

Training takes very little time currently.

As for an ATS system, just leave it on the default and it will do a good job.

about the dynamic player potential,im sure that its already discussed in here,but in the real world the players potential can change a lot throughout the years,if a young player had a great season his potential should be grow a little bit and vice versa.

A player's potential never changes in RL only your perception of it.

StevoRobbo , thanx but i still cant find it..can you give me a link?

Cant find what? The search button?

I agree with Cougar. A player's potential does not change IRL.

I see where Hershie is coming from, and IMO few players should be generated within 5 points of their CA, but I think fixed PA is a good thing because it reflects that all players have a peak, it stops people who aren't talented footballers becoming them, and so on.

I agree with Cougar. A player's potential does not change IRL.

I see where Hershie is coming from, and IMO few players should be generated within 5 points of their CA, but I think fixed PA is a good thing because it reflects that all players have a peak, it stops people who aren't talented footballers becoming them, and so on.

Fixed PA is (usually) the solution I agree, but how players get there maybe needs changing. It would of course be ridiculous to allow for all players to reach an infinite level, and some games that allow that are just that - ridiculous.

Maybe the solution lies in how players reach their potential, making it harder to max out their PA or saving a few points for development later in their career, particularly towards mental development without other attributes falling. A player playing at the very top from a young age (say, appearing to peak at 21) is likely to continue to grow mentally throughout their career as they gain experience. The current system doesn't really allow for this.

Fixed PA is (usually) the solution I agree, but how players get there maybe needs changing. It would of course be ridiculous to allow for all players to reach an infinite level, and some games that allow that are just that - ridiculous.

Maybe the solution lies in how players reach their potential, making it harder to max out their PA or saving a few points for development later in their career, particularly towards mental development without other attributes falling. A player playing at the very top from a young age (say, appearing to peak at 21) is likely to continue to grow mentally throughout their career as they gain experience. The current system doesn't really allow for this.

Yes I agree here, and the only thing I can see hapenning with any idea other than a fixed PA.

What most people are going on about when they talk about players' potentials changing, is either they thought a player was better than they turned out, because he was an early bloomer or he's a late bloomer that only came good at 26 or 27 (well relatively anyway).

The game doesn't really model this well. We all know that a player will only improve marginally at the age of 23, and with the right regimes/tutorships and game time will improve steadily until then. There needs to be more randomness in the game, like sudden growth spurts followed by years of plateau or a slow growth which explodes at the age of 27 when a player is bought by the right club, initially as back up, but getting the belief and chances to move up the ladder.

I would like to see club philosophy and playing style also influence the type of player development more. For my team I expect my full backs to be very attacking (albeit good defenders too) and with the mentality to go forward and take chances. I would like to see over a number of years that my youths adopt that philosophy (ok this can be done, but with attention to training) naturally, and if I buy in a player from a club with a philosophy of out and out defence from the back four, I would expect him to be a bit confused at first (unless he's a journeyman or very adaptable), liable to be not as good training wise (i.e. don't pick up stats) and either slowly grow into the club philosophy or get stunted and maybe even ending up getting worse and having his career going into a tailspin.

Ooh, another PA thread...

My view is fairly simple: PA is generated from a set of circumstances at the start of his career, and if these circumstances change dramatically in the future then his PA is not appropriate for his new set of circumstances.

It is better to "plant a flagpole" in the ground defining a player's talent, and letting his development vary in tandem with his circumstances but fixing his talent at the start - the flagpole.

Put another way: If two youngsters players in the database have PA 140 and 160 respectively, and reasonable and comparable development attributes (ambition, professionalism, training facilities, injury proneness, etc.), then the latter will almost always be better than the former. This is despite the fact that the development of a player is dependent on first-team football. In other words, nothing you do from the start of his career will ever change his maximum. In every game you start you will be more likely to choose the PA 160 player if you can afford him, because he's potentially better. However, there may be cases where if you nuture a less-talented player much better than the talented one, you can expect to see the less-talented player turn out better than expected.

Of course there is always a maximum, peak CA, but it doesn't mean we need to decide what it is at the start of his generation (or in the database).

An argument I always hear against movable CA is that everyone will become a Lionel Messi - no! This won't happen. It's possible but can be designed to be near-impossible - rather akin to Lionel Messi suffering cruciate ligament damage again and again to the extent he is later found playing at Blue Square level at his "peak" (around 28 years of age). It's not likely to happen, and is arguably not worth considering.

Training takes very little time currently.

As for an ATS system, just leave it on the default and it will do a good job.

A player's potential never changes in RL only your perception of it.

Thats the problem with the currently default training-like you said-it will do a good job,but i want an excellent job and i dont care to pay a lot of money for it from the team budget. the training part is like obstacle in a middle of a run..because i like to run the game very fast..

Thats the problem with the currently default training-like you said-it will do a good job,but i want an excellent job and i dont care to pay a lot of money for it from the team budget. the training part is like obstacle in a middle of a run..because i like to run the game very fast..

Just download some schedules. SFraser's and Tugs' are probably the best. Best thing is it won't cost you a thing.

Ooh, another PA thread...

My view is fairly simple: PA is generated from a set of circumstances at the start of his career, and if these circumstances change dramatically in the future then his PA is not appropriate for his new set of circumstances.

No it isn't. PA is what a player can acheive with the best training available, with no injuries and plenty of playing time etc, it is not based on his circumstances. As circumstances change so does the likelyhood of a player reaching his maximum potential.

For example, IRL any young player will have a limit that they will reach and never get any better than that. We don't know what that limit is, but it's there. That is what PA is simulating.

Just download some schedules. SFraser's and Tugs' are probably the best. Best thing is it won't cost you a thing.

and then you need to devide them to groups and then you need top hire staff(with hopes for 4 stars)and then you need to teach them new moves and even then you are not sure if you doing it good or not..i jus dont like the training part..it reduse the fun of the game..its like something that you do only because you have to..

and then you need to devide them to groups and then you need top hire staff(with hopes for 4 stars)and then you need to teach them new moves and even then you are not sure if you doing it good or not..i jus dont like the training part..it reduse the fun of the game..its like something that you do only because you have to..

The more you put into the game the more you get from it.

If you just want to race through seasons and win trophies with little effort then perhaps you should consider if FM is the game you are looking for.

and then you need to devide them to groups and then you need top hire staff(with hopes for 4 stars)and then you need to teach them new moves and even then you are not sure if you doing it good or not..i jus dont like the training part..it reduse the fun of the game..its like something that you do only because you have to..

Oh no, how long does that take, like 5 minutes? And then its sorted. Why bother playing the game if you don't want to put that sort of effort in.

The more you put into the game the more you get from it.

If you just want to race through seasons and win trophies with little effort then perhaps you should consider if FM is the game you are looking for.

i didnt said that,training its not the only thing in that game..there is scouting and tactics and finding the next star and to do a lot more things but i dont like training and i am expect for another option[not to replace] for people like me..thats my wish..the game supposed to be fun.

i didnt said that,training its not the only thing in that game..there is scouting and tactics and finding the next star and to do a lot more things but i dont like training and i am expect for another option[not to replace] for people like me..thats my wish..the game supposed to be fun.

and that option is already there, its called default training.

Yes it can be improved upon but that is reserved for those people that want to put extra effort into that area. You don't which is fine but why do you feel you should get that little extra benefit without the work?

Just want to expand on a couple of your earlier points as well:

Top staff: Not needed if you use default training, do you know what your coaches & training really do?

New moves (PPMs?): Takes a split second to ask a player to learn a move and I can't really see how this part could be automated. You also get recommendations via your backroom advice.

the game supposed to be fun.

Actually, it's supposed to be a simulation. Amazingly, real managers actually have to have some input into training.

New moves (PPMs?): Takes a split second to ask a player to learn a move and I can't really see how this part could be automated.

Actually that's given me an interesting thought. The best, most professional and intelligent players are those that recognise their weaknesses and try to do something about it, be it through improvement or adapting their game.

Could that be a route to examine?

Maybe the solution lies in how players reach their potential, making it harder to max out their PA or saving a few points for development later in their career, particularly towards mental development without other attributes falling. A player playing at the very top from a young age (say, appearing to peak at 21) is likely to continue to grow mentally throughout their career as they gain experience. The current system doesn't really allow for this.

Players pretty rarely meet their PA in game from my experience. Most still retain the capacity to improve.

I don't think the current system would take well to players learning their own PPMs (see some of the ridiculous suggestions from coaches) but it could be good in the future IMO.

I'll take a look at my current team in FMRTE.

GK: Age 21 - CA 183 PA 188

DL: Age 19 - CA 179 PA 180

DR: Age 21 - CA 187 PA 188

DC: Age 20 - CA 174 PA 175

DC: Age 20 - CA 178 PA 179

MC: Age 23 - CA 175 PA 176

MC: Age 23 - CA 171 PA 172

AML: Age 26 - CA 177 PA 178

AMR: Age 23 - CA 181 PA 182

ST: Age 23 - CA 185 PA 186

ST: Age 21 - CA 187 PA 189

Even with the best training and plenty of games that doesn't seem quite right.

I've always wanted some form of dynamic potential. I certainly feel that the current CA PA format has potential for improvement, I just can't think of how.

I've always wanted some form of dynamic potential. I certainly feel that the current CA PA format has potential for improvement, I just can't think of how.

My suggestions would be:

1) Fixed PA stays.

2) The higher a player's PA the harder it should be to reach it with far less players ever maxing out with a PA of 150+

3) A change in the way CA develops so players reach their peak at different ages.

4) Maybe separate CA/PA for each area (Technical, Mental & Physical).

4) might be getting on to something.

A top player is often a technically excellent one from a young age. What makes them great is their ability to learn mentally over their career, while physicals could develop more naturally and gradually, and some (strength) can be increased near continually and regardless of other ability - although potentially they may bulk up more rapidly at a younger age.

On the other hand there could be those players who are mentally excellent and have the mindset to train themselves into a technically capable player, or physically excellent ones who look to learn a footballing brain to put their physical presence or ability into effective use.

My suggestions would be:

1) Fixed PA stays.

2) The higher a player's PA the harder it should be to reach it with far less players ever maxing out with a PA of 150+

3) A change in the way CA develops so players reach their peak at different ages.

4) Maybe separate CA/PA for each area (Technical, Mental & Physical).

Number two is essential I think. You can very easily make a mathematical model for it too, as all it is is an exponential curve, e.g. X=1*W*(200/200-CA) X is the stat to be increased, W is the weighting for the stat given the players position (already in game) and CA is CA. Now the equation can be improved, and made a lot tighter, but computers are really good at these kind of calculations.

I like the idea of 4 too, never considered it before, but it is very interesting.

the issue is how potential is rated in fm. Is it rated as what the player could reach given the best possible circumstances? Or is given based on what a reasearcher thinks the player would reach given their current performances/circumstances.

I think the former would be hard to rate so its done on the principle of the latter example, which isn't entirely accurate. How many times have we seen a player make a step up that no one thought they would make? And as a result these players can never achieve that level in fm until the next version or update, so in that sense i think a dynamic pa rating is a good idea.

fixed PA+10 would be great, if player reaches his potential very early, then PA+10

you guys always say fixed PA is OK, just say me: what was Forlan's PA in FM2005 and before ?...

fixed PA+10 would be great, if player reaches his potential very early, then PA+10

you guys always say fixed PA is OK, just say me: what was Forlan's PA in FM2005 and before ?...

In RL his PA has never altered. What has changed is a researchers perception of what it should be.

My suggestions would be:

1) Fixed PA stays.

2) The higher a player's PA the harder it should be to reach it with far less players ever maxing out with a PA of 150+

3) A change in the way CA develops so players reach their peak at different ages.

4) Maybe separate CA/PA for each area (Technical, Mental & Physical).

This is what I've been asking for for years, especially 2) and 4)

I think it's disingenuous to say that players rarely max out young in real life though - particularly mediocre players. You frequently see academy youngsters at big teams simply stop improving and end up in non-league.

Number two is essential I think. You can very easily make a mathematical model for it too, as all it is is an exponential curve, e.g. X=1*W*(200/200-CA) X is the stat to be increased, W is the weighting for the stat given the players position (already in game) and CA is CA. Now the equation can be improved, and made a lot tighter, but computers are really good at these kind of calculations.

I like the idea of 4 too, never considered it before, but it is very interesting.

What will also be required is modelling of how much each stat can realistically improve, which stats can be most easily improved, and which ages each stat usually undergoes the most development at. For example, you will rarely see players improve markedly on their dribbling in their mid 20s, but you would see their decisions, teamwork, anticipation or even first touch improve more.

A top player is often a technically excellent one from a young age.

I would add to this that they are likely to be technically excellent in a few certain areas at a younger age and become more technically well-rounded as they get older. I can't think of a top-class player who hasn't significantly rounded their technical game moving into their early and mid 20s.

What will also be required is modelling of how much each stat can realistically improve, which stats can be most easily improved, and which ages each stat usually undergoes the most development at. For example, you will rarely see players improve markedly on their dribbling in their mid 20s, but you would see their decisions, teamwork, anticipation or even first touch improve more.

Yeah that too, as I said (and frankly should have expanded on) is that my calculation can be improved. My expansion should have been, this is only a simple calculation to give an example of what can work, and it leaves out many variables that can be added in as well.

Any way let me try and model another equation with your improvement: X=1*W*(200/200-CA)*(Y+Z) {IF x is in group 1 then y=1.05 If x is in group 2 then y=1.02 If x is in group 3 then y=1} {If age is >28 then Z=1.2 if age is <28 but >23 then Z=1.1 and if age is <23 then Z=.9}. Groups 1,2 & 3 are the current groupings (and don't consider hidden ones).

As my warnings above state this is an incomplete model, only useful for illustrative purposes. Much revision would be needed to get this close to reality.

No it isn't. PA is what a player can acheive with the best training available, with no injuries and plenty of playing time etc, it is not based on his circumstances. As circumstances change so does the likelyhood of a player reaching his maximum potential.

For example, IRL any young player will have a limit that they will reach and never get any better than that. We don't know what that limit is, but it's there. That is what PA is simulating.

I know what it means. I just disagree with it.

For a regen, his PA is determined at the start of his career, based on certain circumstances (i.e. the reputation of the club, its training facilities, coaching staff, and so on). In real life, how good a player becomes (ignore the notion of a limit, or PA, for now) is dependent on things like talent, luck, training facilities and first-team football. The circumstances modeled in-game do not include first-team football - the maximum is not based on first-team football.

I see talent as being largely constant but the remaining circumstances as being highly variable - therefore it does not make sense to have the notion of a constant limit.

The metaphor I used previously was a car - a car has a theoretical maximum speed on a flat road, and it is certainly possible to estimate it. However, we do not put a limiter on the car to justify its maximum speed - we simply let the car go as fast as possible and the maximum speed will appear eventually (in theory). Here, you replace the car with the game, the maximum speed as the peak CA and the limiter as PA.

The model I talk about instead doesn't try to model the limit, but puts a rough ballpark figure for PA somewhere, almost like an average. For example, you could say that the majority of players coming out of Arsenal's academy will have on average a career somewhere in the upper levels of the Championship, and you can plant various flags in the sand. This is opposed to the current system, where we say that players will never be better than a certain level - something difficult to model because PA moves between data releases. To model this, my model accepts that this ballpark figure isn't an upper limit but a rough idea of where a player will turn out. Any uncertainty is modeled via the game - injuries, training facilities, and first-team football.

Yes, there will always be some limit, because players don't continue developing forever. I just don't think there is a need to figure out what it should be at the start. Just like a car, why not let the car (development engine) show off how a car will perform in different scenarios - rubbish surfaces (i.e. poor training facilities), driven by a clumsy driver (i.e. injuries) and as a driver becomes more experienced (i.e. first-team football)? The model then justifies the system.

However, I do think the main reason why PA exists is because it makes things easy in the black box - it's easier to use PA to predict how good a player will become, rather than predict how good a player will be based on many attributes (ambition, professionalism, first-team football, injuries, training facilities...). Scouting is therefore easier. However, since PA is a hard-coding of some sort, and CA a weighted average, I predict that within 5 years both CA and PA will be gone, due to Moore's Law - there is no need to have a weighted average and a hard-coding - instead, the circumstances behind this weighted average and hard-coding should be modeled.

The circumstances modeled in-game do not include first-team football - the maximum is not based on first-team football.

In-game, first-team football is the most effective way of raising CA.

The model I talk about instead doesn't try to model the limit, but puts a rough ballpark figure for PA somewhere, almost like an average. For example, you could say that the majority of players coming out of Arsenal's academy will have on average a career somewhere in the upper levels of the Championship, and you can plant various flags in the sand. This is opposed to the current system, where we say that players will never be better than a certain level - something difficult to model because PA moves between data releases. To model this, my model accepts that this ballpark figure isn't an upper limit but a rough idea of where a player will turn out. Any uncertainty is modeled via the game - injuries, training facilities, and first-team football.

But this is simply wrong. The majority of players at the Arsenal academy never develop enough to have an eventual career any higher than conference or so. Neither do they all have a roughly equal chance, even if they all have identical injuries, training facilities, and first-team football. Most of them just simply hit a wall where they cannot improve any further. A few continue to develop. Fixed PA models this perfectly, it's just that the game makes it too easy to reach this for players with any kind of decent PA - injuries, lack of first team football, bad coaches etc don't cause enough resistance to development.

there is no need to have a weighted average and a hard-coding - instead, the circumstances behind this weighted average and hard-coding should be modeled.

I disagree. Some players are simply more talented than others. Circimstances alone cannot be made to account for this single fact.

In-game, first-team football is the most effective way of raising CA.

I am aware of that, but the best youngsters tend to have lots of first-team football during their tender ages. While their development of course is also quicker, why can't their peak be?

But this is simply wrong. The majority of players at the Arsenal academy never develop enough to have an eventual career any higher than conference or so. Neither do they all have a roughly equal chance, even if they all have identical injuries, training facilities, and first-team football. Most of them just simply hit a wall where they cannot improve any further. A few continue to develop. Fixed PA models this perfectly, it's just that the game makes it too easy to reach this for players with any kind of decent PA - injuries, lack of first team football, bad coaches etc don't cause enough resistance to development.

Hm... You don't really give Arsenal's academy the credit it deserves. Those who go on to get scholarships go on to perform very respectably. The worst I can think of right now are Hoyte and Aliadière, who have still had decent careers.

Movable PA also can model players hitting a brick wall. They could simply not perform well enough to sustain their development (maybe they play out of position, or they simply weren't ready).

I disagree. Some players are simply more talented than others. Circimstances alone cannot be made to account for this single fact.

Talent isn't the only discriminant that determines how good a player will be. Talent sometimes falls by the wayside, after all. While it is doubtless a major factor, I don't believe it should prevent less-talented players from becoming very good players, sometimes better than more talented ones, and not necessarily because the more talented ones fall from grace dramatically.

Yes, some players are more talented than others - we can model this with a movable PA (i.e. higher talent to begin with => higher probability of them becoming a world class star).

For a regen, his PA is determined at the start of his career, based on certain circumstances (i.e. the reputation of the club, its training facilities, coaching staff, and so on). In real life, how good a player becomes (ignore the notion of a limit, or PA, for now) is dependent on things like talent, luck, training facilities and first-team football. The circumstances modeled in-game do not include first-team football - the maximum is not based on first-team football.

I would agree with the highlighted part to a certain degree but this is not what PA is.

PA is simply the best a player could ever become and is theoretical whereas the above is actual.

Yet more people not understanding what the word "potential" means.

You can't exceed your potential, it is set from birth through nature(your genes) and you either reach it or you fall short, it cannot increase. Current ability increases as you move towards your potential through nurture(training, playing games etc). You then reach a peak CA which is any value upto and including your PA and then you decline.

I am aware of that, but the best youngsters tend to have lots of first-team football during their tender ages. While their development of course is also quicker, why can't their peak be?

This could be more easily and realistically achieved by having a variety of different development curves that could then be affected by factors such as first-team football. I'm not sure quite what you're trying to say about their peak, but under your suggested model it seems that for good players, their peak would just keep rising away from their current level till they get too old to develop. I don't believe that's realistic at all.

Hm... You don't really give Arsenal's academy the credit it deserves. Those who go on to get scholarships go on to perform very respectably. The worst I can think of right now are Hoyte and Aliadière, who have still had decent careers.

If you go back to the squads that won FA Youth Cups in 1999-2000 and 2000-1, there's around 26 players - the cream of the academy at the time.

3 of those never ever played a professional game, not even at conference level.

4 played professionally for a while, but haven't found a professional club for years.

1 was playing professionally until fairly recently, but is now semi-pro

5 are playing at various levels abroad

1 is playing in League 2

5 are playing in league 1

1 is unattached but a Championship level player

2 is playing in the championship

3 are backup players in the premiership

1 is sadly deceased

Lets be charitable and say that the players playing abroad are playing at League 1 standard on average (some are playing higher, some far lower). Let's also be charitable and say that there were only 5 players each year who weren't good enough to make the squad for the final. Presumably few of them have played professionally.

So, we see there are quite a core of scholars, maybe 30-40% who never make the professional grade, while around the same number or slightly more end up at roughly league 1 level. Maybe I was being a tiny bit harsh with my numbers earlier, but I think my point still stands.

Presumably, those 8 players who are no longer professional had very good CA for their age since they won the FA youth cup. I only know of 1 of those whose development was badly affected by injury. I think your model would allow for too many players of this background and standard to continue developing when compared to the reality we see here.

Movable PA also can model players hitting a brick wall. They could simply not perform well enough to sustain their development (maybe they play out of position, or they simply weren't ready).

Talent isn't the only discriminant that determines how good a player will be. Talent sometimes falls by the wayside, after all. While it is doubtless a major factor, I don't believe it should prevent less-talented players from becoming very good players, sometimes better than more talented ones, and not necessarily because the more talented ones fall from grace dramatically.

Yes, some players are more talented than others - we can model this with a movable PA (i.e. higher talent to begin with => higher probability of them becoming a world class star).

But, if the CA development worked properly, the PA system would do pretty much everything you say here. There are players in FM who have high PA but never go on to reach them due to injury, bad training, the factors you mention above - with the result that players with lower PA overtake them in terms of CA. I feel this should happen more than it does at the moment, but there's nothing wrong which a more carefully constructed CA development model couldn't solve.

I would agree with the highlighted part to a certain degree but this is not what PA is.

PA is simply the best a player could ever become and is theoretical whereas the above is actual.

This is true, but I'm saying you shouldn't need to have PA (or it shouldn't be called "PA"). I'm saying that a limit indeed exists, but we don't have to state it from day one, and that a model of how a player could turn out may be better measured by some model where there is no physical limit.

Yet more people not understanding what the word "potential" means.

You can't exceed your potential, it is set from birth through nature(your genes) and you either reach it or you fall short, it cannot increase. Current ability increases as you move towards your potential through nurture(training, playing games etc). You then reach a peak CA which is any value upto and including your PA and then you decline.

I think I understand what potential means, thanks.

I'm saying that a better idea isn't to define PA, but some yardstick figure based on the notion that a player on average will turn out to have peak CA X (say, 160), but allowing variation in both directions - as opposed to saying that his maximum CA will never exceed some figure Y (say, 180). In one game his peak CA might turn out to be 160; in another 150 and in another 170. But on average it will turn out to be 160. The reason this has an advantage over the PA model is because it allows variation in both directions - there is a slim chance the player may have a peak CA of 181 (>180) - but a very slim chance nevertheless.

Think about it - the probability that a player (CA 60) having a peak CA one greater than his PA (i.e. PA 140; CA 141) is zero - less than the probability that this player will have peak CA 60, by virtue of the fact the limit is hard - despite the fact it is extremely unlikely that the player will fall so vastly from grace.

----

Current model

CA

PA (*)

Maximum CA = PA

Average maximum CA = slightly less than PA

My model

CA

"Talent level" (yardstick figure)

Maximum CA = 200 (if extraordinarily unlikely)

Average maximum CA = PA as defined in (*)

----

The main reason I think this is better is because it allows you to have a much better say in how a player develops - simply because first-team football is so important. For two real players with different PAs and acceptable mental/training attributes, the one with the lower PA will be on average worse than the other. But if there is a movable peak level of some sort we can be allowed to force the lesser one past the other through better treatment via first-team football.

I know there's a limit and this is how PA is defined, but I'm saying that it's better to "abolish" PA in favour of a movable target.

I think it's a lot easier to guess how good a player will be on average rather than guess the maximum limit of a player's ability, simply down to the many things that can happen in a player's career. You can say a player will only be on average a League One level player yet if this player finds a good set of circumstances he could vastly exceed this average. One may say he should have had a higher PA to begin with - I say with my model, you don't really have to worry about it.

Indeed, more-talented players will clearly be easier to work with - I would rather sign a 16-year-old Lionel Messi rather than a 16-year-old Shaun Wright-Phillips - the chances of Messi becoming world-class are clearly greater than that of Wright-Phillips. But one day I would like the game to show me that because of this talent, without a limit, Messi will be better 99% of the time. The 1% is the slim chance that Wright-Phillips doesn't stagnate and Messi suffers lots of injuries. You may say that this isn't the point because this is unlikely - I say the game should prove this!

With youngsters, there is a large uncertainty to as how good they will turn out to be - you can get a rough idea, but you don't have a crystal ball. As they get older, the uncertainty becomes less - you become much more certain about how they will turn out. At their peak the uncertainty is zero. To me, this means that you cannot put a limit on at 16 because the upper end of the tail shrinks (uncertainty decreases) as you get older.

A criticism about my model is that it allows any rubbish player to become Lionel Messi. This is true - but it can be designed such that this scenario is incredibly unlikely to happen. It is always going to be easier to work with more-talented players - would you rather bust a gut and force a rubbish player into lots of first-team football to get this far, or pick out the most talented wonderkids and ease them in?

This is true, but I'm saying you shouldn't need to have PA (or it shouldn't be called "PA"). I'm saying that a limit indeed exists, but we don't have to state it from day one, and that a model of how a player could turn out may be better measured by some model where there is no physical limit.

But in RL it is set from day one from the moment you are born and there is a limit to how good you can be.

EDIT

and TBH if you never use 3rd party applications you'll never know what that limit is.

This could be more easily and realistically achieved by having a variety of different development curves that could then be affected by factors such as first-team football. I'm not sure quite what you're trying to say about their peak, but under your suggested model it seems that for good players, their peak would just keep rising away from their current level till they get too old to develop. I don't believe that's realistic at all.

I'm not saying that. Development doesn't always have to be in one direction. It's like swimming in thick liquid - at some point it becomes harder to move (i.e. it's harder to develop far away from your talent level) and sometimes things may go pear-shaped and backwards.

Pathways to a limit is an interesting idea but it doesn't solve the problem that it makes little sense to decide how good a player is based on current circumstances at 15-16.

If you go back to the squads that won FA Youth Cups in 1999-2000 and 2000-1, there's around 26 players - the cream of the academy at the time.

3 of those never ever played a professional game, not even at conference level.

4 played professionally for a while, but haven't found a professional club for years.

1 was playing professionally until fairly recently, but is now semi-pro

5 are playing at various levels abroad

1 is playing in League 2

5 are playing in league 1

1 is unattached but a Championship level player

2 is playing in the championship

3 are backup players in the premiership

1 is sadly deceased

Lets be charitable and say that the players playing abroad are playing at League 1 standard on average (some are playing higher, some far lower). Let's also be charitable and say that there were only 5 players each year who weren't good enough to make the squad for the final. Presumably few of them have played professionally.

So, we see there are quite a core of scholars, maybe 30-40% who never make the professional grade, while around the same number or slightly more end up at roughly league 1 level. Maybe I was being a tiny bit harsh with my numbers earlier, but I think my point still stands.

Presumably, those 8 players who are no longer professional had very good CA for their age since they won the FA youth cup. I only know of 1 of those whose development was badly affected by injury. I think your model would allow for too many players of this background and standard to continue developing when compared to the reality we see here.

You've done your research and I concede. I don't think this is the case for modern-day Arsenal though - the current crop of youngsters will on average go on to have solid careers. Sure you may get the odd player dropping into semi-professional level but this is balanced out by the odd world-class player like Fàbregas.

But, if the CA development worked properly, the PA system would do pretty much everything you say here. There are players in FM who have high PA but never go on to reach them due to injury, bad training, the factors you mention above - with the result that players with lower PA overtake them in terms of CA. I feel this should happen more than it does at the moment, but there's nothing wrong which a more carefully constructed CA development model couldn't solve.

Yes, but what about players with lesser PAs that go on to exceed it? As I mentioned, PA factors in things like talent, training facilities and first-team football. First-team football can change a player's standing rather dramatically. Likewise a lack of first-team football can change it rather dramatically too - in the opposite direction.

We know a player's talent, their training facilities and coaching staff, and a rough idea of their ambition, professionalism and injury-proneness. We don't know how much first-team football they will get - this is variable. Likewise, if a player moves around, the training facilities and coaching staff are also variable. Through tutoring, ambition and professionalism are also variable. An Arsenal youngster is judged largely based on his current circumstances - a youngster at one of the best academies in England (if not the world), with top-notch coaches and a manager who has a penchant for fielding youngsters in the first-team. His limit will be higher than if he suddenly moves to a club that has rubbish facilities and coaches. What stays constant is his talent. But due to the fact PA factors in lots of other things that are variable, I am unconvinced that this means his PA should be constant too. A youngster that makes a move to a vastly inferior coaching and training circumstance will have a poorer future than if he stays at Arsenal. This is reflected by the fact that PAs go up and down all the time between data releases.

I don't see why you can't have the two in tandem - a better CA model and no limit to CA (but a talent level to make it hard to go too far).

If a player has good training facilities and coaches, has a good attitude and continues to play very well with lots of first-team football, has a good attitude and is lucky with injuries, he should continue to develop well irregardless of his talent (but a more talented player may be able to make the most of it and develop quicker - such is the nature of talent). If a player struggles with injuries or form he could hit a stumbling block; if he later rediscovers his form he may become a late-bloomer, and if he is talented (or is that "was"?), he could well turn out to be brilliant (this may be the difference between, say, Luca Toni and Stephen Carr - two relative late-bloomers but one was always more talented).

You'll notice in the above paragraph that I have not mentioned at all anything about his limit or peak. It never comes into play. A player's development is only partially-influenced by his talent (perhaps majorly, but still only partially) - where it peaks is therefore only partially-influenced by his talent. You don't need to know the peak at the start - you just need the facts and figures, and the game will simulate it all for you. Yes there's a peak but does it really honestly matter? It's a bit like saying a player's passing will never exceed 14 - does it really matter? A stupid training schedule may allow it to hit 15, but then again that's an outlier circumstance and in reality this isn't going to happen on average - the fact the schedule is stupid is enough to suggest that it's not worth considering. Like the so-called indestructible buildings that are destroyed by a nuclear missile - not an average circumstance.

But in RL it is set from day one from the moment you are born and there is a limit to how good you can be.

EDIT

and TBH if you never use 3rd party applications you'll never know what that limit is.

Yes, I know there's going to be a limit, but does it make sense to set this at the start of your life? Can you even determine it without a crystal ball?

Also see my arguments about current circumstances factoring in the determination of PA, and what happens if those circumstances change.

I do not buy the argument that "if you shouldn't see it, there isn't a problem."

Yes, I know there's going to be a limit, but does it make sense to set this at the start of your life?

Yes, because thats what happens in real life.

Can you even determine it without a crystal ball?

No but we don't have to as the game assigns it randomly and we never see it.

Also see my arguments about current circumstances factoring in the determination of PA, and what happens if those circumstances change.

Your arguments all refer to a maximum CA figure NOT PA.

I do not buy the argument that "if you shouldn't see it, there isn't a problem."

Thats your choice and I would agree there is an issue but its not with PA, its with CA and how it develops over time within the game.

Yes, but what about players with lesser PAs that go on to exceed it?

That never happens as by definition you can never exceed your potential.

As I mentioned, PA factors in things like talent, training facilities and first-team football. First-team football can change a player's standing rather dramatically. Likewise a lack of first-team football can change it rather dramatically too - in the opposite direction.

Again this is CA and how it develops NOT PA.

I don't see why you can't have the two in tandem - a better CA model and no limit to CA (but a talent level to make it hard to go too far).

Because CA has a limit, its called PA both in RL and in the game.

Luca Toni and Stephen Carr - two relative late-bloomers but one was always more talented).

Late bloomers, yes but neither exceeded their PA in RL. They simple had a lower CA for much of their career.

No but we don't have to as the game assigns it randomly and we never see it.

We see the symptoms of it, though, and it doesn't stop it from existing. And it doesn't work for real players.

Again this is CA and how it develops NOT PA.

The average Arsenal youngster has on average a higher PA than the average Championship youngster. The idea of how a player will turn out will differ from average Arsenal to average Championship too. There is a correlation (and an obvious causation: Arsenal are a better team). Therefore it follows that what we see as a youngster at Arsenal will usually have a higher PA than that of a youngster in the Championship, for similar levels of talent.

PA factors in the best-possible scenario - best training facilities reasonably expected from the player's career, for example. Yet there will be an inherent bias because a youngster at Arsenal is largely expected to not quit in the first year of his professional career, and will therefore get a solid amount of good training facilities and coaches at Arsenal. In other words, the notion of a player's peak is partially determined by his circumstances at present.

This includes first-team football.

Because CA has a limit, its called PA both in RL and in the game.

Again, I'm not bothered by the fact a maximum exists - just that there may be no reason to stick it in-game.

In real-life a Golf GTI is going to be slower than an F1 car. However, in a racing simulation, I don't need to stick in a maximum speed for both cars to prove it. I will simply design my game such that the mechanics of the car are very different and that the F1 car simply has better components, leading to a faster maximum speed.

Likewise, if a car has a known maximum speed, I should aim to model my game and the car well-enough that the maximum speed in-game and in real-life match - without having to stick a limit in. I can therefore reuse this code for other cars. I can then use this code for fictional cars (read: regens) with an unknown limit - and they will perform with an unknown maximum speed - albeit one I can find out through testing the game (read: a rough idea of the peak CA of a player at his peak, or at least how good he is).

Late bloomers, yes but neither exceeded their PA in RL. They simple had a lower CA for much of their career.

I agree they never exceeded their (true) limit, but their PAs would have gone down and then gone back up through research phases - something a variable CA development would take care of, as long as there is no limit (or at least it doesn't move around so much).

I think I understand what potential means, thanks.

You clearly don't, based on the rest of that post.

You clearly don't, based on the rest of that post.

It depends how researchers rate PA, yes we all know its the absolute limit a player can reach but researchers can't know that limit. So instead they rate a players PA based on how they are performing and what they think they are likely to reach based on their current club and coaches. But thats not the PA we know as their limit.

Essentially it means there are going to be plenty of players whose PA in game isn't accurate to their real life PA and so the system isn't complete either way.

I agree they never exceeded their (true) limit, but their PAs would have gone down and then gone back up through research phases - something a variable CA development would take care of, as long as there is no limit (or at least it doesn't move around so much).

One thing that's bugged me for years with the black and white PA system we have now is how some players "potential" changes so dramatically. If potential really is just how good a player can possibly get, then why do players who players who are considered to have great potential yet maybe don't develop too well end up with lower potential in future games. Has the player "lost" his potential? Is it researchers admitting they've made mistakes, or that players were indeed just "good for their age"? It's a rather grey area.

Take someone like Jermaine Pennant. He was once a highly rated youngster with high potential, but he has never really fullfilled it. If he had this potential at all though, why is it not in the game? Should he not have the potential, ie the high PA, to become the player that it was believed he could become, even if it's very likely he'll never reach it? Look at someone like, as mentioned previously, Luca Toni. Was he considered to have great potential at a young age, or even in games just a few years ago? What made him fulfill his potential at last, and perhaps why was it not noticed previously?

Players won't neccessarily exceed their potential, but a lot may have potential that isn't obvious at a young age, and there needs to be some way to allow for players who maybe aren't the cream of the crop at 16, but have the right abilities and mindset to become a better player later in life, and similarly allow those who have the clear potential at a young age yet don't appear to reach it to have the opportunity to eventually. Not phrased that very well at all, but I'm tired.

It's all a tad hypothetical, but to show how it's not as simple as it all seems.

to the fact that we seem to have a proper, sensible discussion going on in GD. wp.

edit: also, I like and agree with the fact that presently personality is a major factor in how well a player develops and I really think this is an area to look further into in this respect.

You clearly don't, based on the rest of that post.

Please elaborate.

I'm aware that PA is a limit and is defined as the best-possible CA attainable by a player.

I'm saying that PA is a function of various factors based on current circumstances. However, our perception of how good a player will be (not PA - real life) will change if these circumstances change. A bit like opening more doors - or closing some. If a player moves from non-League to a Premier League side his PA will inevitably be relooked at at the next research phase, because his circumstances have changed.

This is why I don't think PA is a good idea in-game. I know what it is but I think we can design a better model.

Talent remains constant throughout your lifetime - how good you will ever be will depend on the entire set of circumstances throughout your lifetime - not just the circumstances you began your career with.

@Hershie: I agree. Take your Pennant example. He would have had a decent PA from Notts County presumably, although a lot of his ratings would have come from the Arsenal researcher at the time. Pennant would probably be rated as a mid-table Premier League player initially, due to his talent - but with a troublesome personality (low professionalism, adaptability, and so on). Under my model, a player who played the current version of Football Manager (or Championship Manager I suppose) at the time would have been able to perhaps develop Pennant into a solid player for Arsenal as long as he could get around his poor personality - which would probably 9 times out of 10 stop him from getting there. Yet his talent would only provide a very rough guide to as how good Pennant would become - how good he would really become depends on how the user treats him. In this way you get to better influence Pennant without hard-coding a limit that changes with every research round.

Youngsters have especially uncertain futures - which is why negative PAs exist - to allow some form of randomness. To me, you could ask the question whether there is a need to actually generated a random PA upon game start - or simply see the negative PA as something like "the player will be at most best in the range X to Y" - the actual value being indeterminate, and the better you treat the player the more likely his actual limit will be higher.