Potential Ability: The "Jamie Vardy" problem

[leonavich]

That's not the ethos of the game though. All the players down the years that we've gotten wrong and over-rated in terms of potential ability are generally considered to be failings by researchers. There are genuine issues that come along as barriers to progression, such as serious injuries but if there is a player who one of us has rated as should almost certainly make it to the top and then gets no where near in real life subsequently that is collectively on us and so we strive to be better at researching.

It goes against it all then completely to throw in some random elements each season just to provide the game with some lower-league RNG mechanics for development. SI could do that, but based on everything that's happened so far it'd likely be a change in approach towards the striving for ever improving and ever more accurate player data.

I get that for the real life players in the database, but in regards to the comments above around late bloomers and players still growing, it is just a thought to cater for that without making the system too complicated and it may only apply to regen players later down the line.

The system works fine for me as is, but offering up an idea that popped in to my head whilst reading peoples thoughts.

[Matshit]

In FM players can still improve once they've hit their PA value, this has been repeated many times in every thread of this nature.

Well, yes. In terms of the player being able to lower some stats and improving others. In terms of proper improvement not so much.

This statement is totally unacceptable & pretty much undermines anything you have to say on the subject.

Well, I still think it is true. Not because I think the people doing the research aren’t as qualified as they can be, but because PA has to be a big amount of guesswork.

You can only see how good the player is now and if he seems to have the correct attitude (both of wich are already in the game).

[bwilliams23]

The CA/PA ratings work fine. I really fail to see the need to incessantly make the game even more complicated.

Exactly. Not to mention, adding complexity without really changing the output adds more potential for things to go wrong.

Ok, lets try some more pertinent examples than Jamie Vardy and Harry Arter. Apologies in advance for using FM13 rather than FM16, it’s the latest version I have:

Robert Lewandowski:

• FM 2008, potential ability -7 (110-140)

• FM 2013, current ability 167

Sergio Busquets

• FM 2008, potential ability -6 (90-120)

• FM 2013, current ability 169

Kevin de Bruyne

• FM 2008, potential ability -6 (90-120)

• FM 2013, current ability 140

Granted these aren’t late bloomers a la Jamie Vardy, and I’m not having a go at the researchers, who I’m sure set PA at their best estimate with the information available at the time. But these three very famous players highlight my point that PA restricts upper growth, so much so that in FM08, even if all three had achieved 100% of their PA, they could not been anywhere near as good as they start out in FM13.

In which case, my suggestion would be to set all PAs typically high, especially for young players, but to make it very hard to achieve 100% of this, particularly when the player has a low starting CA. I.e. all the stars have to align. So to answer a number of points above

• Would it require work on the game mechanics? Yes, but isn’t that the case with every new release

• Would it add to anything to the game? In terms of unpredictability, I think it would, appreciate others will disagree

• Will it make the game more complicated for the player? I don’t think so

However, having read this:

There are numerous stats off the CA calculation, these are unweighted attributes within the sphere of CA and these can change throughout a players career. The actual development of players and such later in their careers is nothing to do with the CA/PA system.

CA and PA purely serve as loose reference points, with the nuts and bolts of how good a player can truly be, and how good a player truly is being nestled away much deeper.

I acknowledge that I may place too much importance on the PA / CA system. I had understood that PA / CA drove development of all other attributes, (to ensure a Conference player couldn’t have lots of 20s for example) but as you can probably tell from the above, my days of playing FM for months on end are gone and I may be behind the times on how crucial PA / CA are. But I do maintain that the above examples highlight how restrictive PA can be.

[Craigus89]

Well, I still think it is true. Not because I think the people doing the research aren’t as qualified as they can be, but because PA has to be a big amount of guesswork.

You can only see how good the player is now and if he seems to have the correct attitude (both of wich are already in the game).

I agree with what you are saying to an extent. Professional Scouts who work for the best clubs in the world get it wrong sometimes. I don't mean any offense to the researchers who obviously do a fantastic job, but it is impossible to predict something like this accurately, especially a static ceiling for how good someone could be in 10 years time. It is simply impossible. Let alone being able to accurately predict who is going to be a late bloomer and their development stunted for say 7 years. How could you possibly know that!?

This is crossing between the realms of what is accurate to real world players and what is happening in a computer game. Any "late-bloomer" would have to be random and not based on real world knowledge, otherwise they wouldn't be a late bloomer because all the professional football people would know it and be trying to help them improve.

EDIT: Basically what I'm saying is, you cannot predict the future, and that is what PA is attempting to do. The only way to add variance to the current system is to randomise it, not by better scouting of real-world players.

[sammayel666]

What problem?

I'm playing in the Premier League

One of my starting centre backs was signed from League 2 at the age of 26 (He was also a D/AM R retrained to CB). Two seasons averaging ~7.

One of my starting strikers was signed from League 2 at the age of 23 (after being the leading goalscorer). He's got 7 in 18 so far and is not looking out of place.

I realise that I've just picked two specific examples to combat the Vardy example, but it shows that the late bloomers are there if you give them a chance.

Focus on attributes and match ratings rather than the number of stars.

[Guest]

If people stopped looking at the pure CA/PA values in isolation there wouldn't be a problem, unfortunately the AI is less capable of that at the moment but some more tweaking of the role CA & transfer logic plus the way match squads are selected would see significant improvements from the AI.

Even with the current AI weakness in this area on my FM15 save Chelsea had a former player of mine when I played them in the CL as AC Milan manager, the player was a lad I signed from League 2 when I was managing a Championship side.

[s1111]

As pointed out by others in the thread, it's possible for players with low CA to put in Vardy-esque performances well above their supposed level if given the right role in the right system. Especially limited strikers and centre backs who only need select attributes to score goals/defend and play well. That's the real Vardy effect IMO.

[Matshit]

If people stopped looking at the pure CA/PA values in isolation there wouldn't be a problem, unfortunately the AI is less capable of that at the moment but some more tweaking of the role CA & transfer logic plus the way match squads are selected would see significant improvements from the AI.

Sure, but a player CA is a quick way to see how good he is, on average you will expect the quality of players to increase as the CA increase.

My biggest problem isn’t the PA in itself anyway, more the PA your scout and assistant gives you. It is WAY to accurate. The last time I checked, two identical players with different PA got different ratings from the scouts as well. That should never happen.

When I get a new list of youths I want my assistant to sometimes say “This player will most likely never be good enough for the club” only for him to turn into a world beater. And I want a chance for a player with the potential to be as good as Messi to start up at a weak Conference South club with a CA on the same level as his team mates. 99/100 times he never ends up anywhere, but with a bit of luck he might slowly progress to turn into a quality player. And I want a 18 year old to look like a superstar, only for his career to fizzle away. In my view, big clubs should produce players with a higher starting CA, but PA should not be in any way affected by training facilities. That makes no sense.

[Guest]

A player with the potential to be as good as Messi will be very unlikely to start their FM life as low down as the Conference because that will be next to a zero chance of happening irl, there will be examples irl of 8 or 9 year old players starting out at their local club but by the time they reach FM newgen age (14-16) the real life equivalent will have left that small local club for much larger club to get access to the better quality of coaching & facilities. This can be achieved if a real life lower league clubs has a world renowned youth academy & in fact this can be seen in the case of Crewe in FM as they can & do produce youngsters with PA in excess of 150.

All you've done in this thread is display how little you've spent investigating & testing what's going on in FM, take the time to fully review a 20+ season save that's has been run on full detail & you will start see that much of what you claim doesn't happen in FM actually does happen.

As for quality being linear with CA I'm not sure how many times people need to repeat that CA is not the true definition of how good a player actually is, a 125CA with perfect attributes for their position & role will perform better than a 150CA player who has attributes that lack any balance & is being forced into a role they are not comfortable in, add in that the 150CA player might be lacking in hidden attributes while the 125CA players has excellent hidden attributes & that margin in actual performance just becomes greater.

Edit: On the point about newgens with Premier League potential coming through in the lower divisions I've just checked a test save I'm running & a 146PA player came through at Dartford who where in the Conference South, his career has stalled a bit & he is currently at Cardiff in League 1. Barnet (L2) had a 143PA come through, AFC Wimbledon (L2) sold a 150PA & the highest rated English newgen so far (194PA) came through at Birmingham City who were in the Championship.

[Sheer Class]

Speaking from years upon years of researching the game.

The only players who are able to buck the trend are players with exceptional pace.

Vardy, Wright are players with that pace because its something that cannot be taught or easily defended against.

Then they have luck, right manager willing to take a chance on them.

He is having the best possible season if he recreates it then time to have a look but he like Wrighty are in the top 0.1% of players who made something when they looked dead and buried.

[Matshit]

A player with the potential to be as good as Messi will be very unlikely to start their FM life as low down as the Conference because that will be next to a zero chance of happening irl

Not Messi quality, but there are lots of examples. Jordi Alba would have started in Barcelona, but with such a low PA that the club would decide to let him go at the age of 16, and he would be signed by a level three club in Spain.

Domenico Berardi would have started his Football Manager career at Cosenza at the lower part of level three. Giaccherini at a club so weak I cannot even find out what level they are on. Diego Costa at some sort of hobby team in Brazil. Lewandowski at a level three team in Poland.

but by the time they reach FM newgen age (14-16) the real life equivalent will have left that small local club for much larger club to get access to the better quality of coaching & facilities

But that makes no sense. The fact that you are not playing football seriously and lack the discipline does not mean your potential is any lower. Again, the big clubs would easily swoop up the big CA youngsters, but if the player isn’t showing any quality to prove his potential, how would anyone know?

All you've done in this thread is display how little you've spent investigating & testing what's going on in FM, take the time to fully review a 20+ season save that's has been run on full detail & you will start see that much of what you claim doesn't happen in FM actually does happen.

I have. Way above 20 as well. It doesn’t.

As for quality being linear with CA I'm not sure how many times people need to repeat that CA is not the true definition of how good a player actually is, a 125CA with perfect attributes for their position & role will perform better than a 150CA player who has attributes that lack any balance & is being forced into a role they are not comfortable in, add in that the 150CA player might be lacking in hidden attributes while the 125CA players has excellent hidden attributes & that margin in actual performance just becomes greater.

That is true. But again, I said “on average”. In the same way as having three dice would on average give you a higher number than two, even though potentially the one with two dice could end up with a number four times as high as the one with two.

[themadsheep2001]

A player with the potential to be as good as Messi will be very unlikely to start their FM life as low down as the Conference because that will be next to a zero chance of happening irl, there will be examples irl of 8 or 9 year old players starting out at their local club but by the time they reach FM newgen age (14-16) the real life equivalent will have left that small local club for much larger club to get access to the better quality of coaching & facilities. This can be achieved if a real life lower league clubs has a world renowned youth academy & in fact this can be seen in the case of Crewe in FM as they can & do produce youngsters with PA in excess of 150.

All you've done in this thread is display how little you've spent investigating & testing what's going on in FM, take the time to fully review a 20+ season save that's has been run on full detail & you will start see that much of what you claim doesn't happen in FM actually does happen.

As for quality being linear with CA I'm not sure how many times people need to repeat that CA is not the true definition of how good a player actually is, a 125CA with perfect attributes for their position & role will perform better than a 150CA player who has attributes that lack any balance & is being forced into a role they are not comfortable in, add in that the 150CA player might be lacking in hidden attributes while the 125CA players has excellent hidden attributes & that margin in actual performance just becomes greater.

Edit: On the point about newgens with Premier League potential coming through in the lower divisions I've just checked a test save I'm running & a 146PA player came through at Dartford who where in the Conference South, his career has stalled a bit & he is currently at Cardiff in League 1. Barnet (L2) had a 143PA come through, AFC Wimbledon (L2) sold a 150PA & the highest rated English newgen so far (194PA) came through at Birmingham City who were in the Championship.

Any CA/PA debate quickly become circular because, while they post with the best of intentions, there are not enough people who understand the current system well enough.

[Guest]

The 146 PA player who generated at Dartford is rated by the West Ham scouts as having 2 star potential but might be as high as 3.5 stars, if he hit his PA his pure CA rating would make him the 3rd best player in the squad yet the scouts think he'll be no better than a backup & if he's lucky he might be bang average. I reckon most people would ditch or ignore a 1* CA 2* PA player from theirs or another clubs youth intake yet in the case of this player they could yet be proven to be wrong in that decision.

Matshit, don't state a point using Messi quality as the benchmark & then ignore that aspect when I reply to that point, you said Messi therefore you mean Messi, not anyone else who have similar raw ability to those I have mentioned in my reply to you.

If you want to have a discussion have the decency to not twist replies or change your intent midway through.

[scott MUFC]

at the end of the day the Vardy thing is based more on a few areas rather then a broad radius, which the stars system is based on, that's why 3 star players can sometimes be more worthwhile then a 5.

[Matshit]

Matshit, don't state a point using Messi quality as the benchmark & then ignore that aspect when I reply to that point, you said Messi therefore you mean Messi, not anyone else who have similar raw ability to those I have mentioned in my reply to you.

Well, that doesn't make a big difference. If a player of Lewandowskis quality can end up on level three in Poland, then surely a player of Messis quality can as well. And besides, Who is to say Messi had a much bigger potential than the other mentioned? There are tons of reasons Messi is as good as he is. A big part is a natural talent, but there are probably a lot of other players who COULD have been as good as him, it is just a lot of hard work and luck.

[leonavich]

There are a few instances where there are players that exceed expectations, or the scouted potential.

I signed a regen player, Niccolo Mora from Titrium, on a whim. I was a Serie B team at the time, and I think we played them in a friendly, or a scout recommended him as a 3 star player at the time.

Needless to say, he only got 2 Italian Caps, but ended up with over 1000 appearances for the club, and was probably one of the best players for the best part of a decade.

He started out with a 6.92 and 6.84 in Serie B and topped his career with a 7.47 in Serie A 9 years later, with 16 goals and 24 assists as a MC / AMC.

He retired at 40 years of age, averaging 6.81 in the Serie A over 31 games, managing 5 goals and 10 assists.

Not a 5 star rated player, even for a then Serie B team, from a non league side, so there are players in there that can become world beaters.

I never checked his CA or PA, but I loved being able to find a gem among the rough. He has become one of my all time favourites.

Just to add to the discussion,

It may be worth having the average rating discussed in the scout reports as a way to encourage well performing players to be considered, instead of just high attribute stats, or PA levels. I think the current system does what people are looking for, its just not a visible because the AI don't necessarily find these players, and as mentioned above, players may overlook a 3 star CA / PA for a 4.5 rated player, even though given the chance, they could become one of the best on the team.

Perhaps, not just signing of the season, but most improved could be a category monthly and yearly, even as a news item / subscription, to create more visibility to the players who have had the highest jump in there performances month to month or year on year? Even acknowledging a player who moves from say League 1 to the Championship and performs well, creating this acknowledgement and news within the game, will highlight improved players and well performing player and add this element to the gaming world?

[Sunstrikuuu]

Messi by himself is actually kind of an interesting edge case. His PA coming through at Barcelona is 200. His PA had he come through at, I dunno, Cefn Druids would probably have been 0, because Cefn wouldn't have been able to afford the treatments necessary to treat his growth condition.

[Guest]

Well, that doesn't make a big difference. If a player of Lewandowskis quality can end up on level three in Poland, then surely a player of Messis quality can as well. And besides, Who is to say Messi had a much bigger potential than the other mentioned? There are tons of reasons Messi is as good as he is. A big part is a natural talent, but there are probably a lot of other players who COULD have been as good as him, it is just a lot of hard work and luck.

He was on Legia's (Big Polish club) books as a youth player & for whatever reason moved to a smaller club as a 17 year old before reasonably quickly being picked up by Poznan (another big Polish club), in terms of Polish football be probably didn't appear out of the lower leagues without anyone having heard of him & unless you have knowledge of a particular nation using examples from that nation is very unwise as you cannot know the circumstances at hand.

I'm all for encourage debate & discussion but so far all you've done is throw out unsubstantiated statements on how FM works & reference real life data or records but without having any apparent understanding of them.

[Craigus89]

Let's not also overlook the fact that for this to be authentic or realistic, we shouldn't actually be able to easily spot someone like Vardy by scouting until they are actually performing at such a high level, otherwise he would have been signed by Real Madrid in the pre-season... so all this talk of better representation of PA by scouts is redundant.

[Guest]

See post #63, a player with similar potential as Vardy is rated as being a less than inspiring backup for an average Premier League side.

There is also the matter of world depth, in most nations in FM players are generated in the top 2 divisions because that's all that exist in FM, it's very rare for players to be generated at clubs outside of the active leagues because those clubs do not need any players.

[scott MUFC]

on FM in the past I have seen 5 star potential players pop up at no club at all think it was 14 where a scottish player popped up, and my scout found him, at no club at all, and another one in 15 in Cyprus that had great potential.

[Overmars]

Just thinking out loud, trying to fit in to the late bloomer conversation and simpler ways of implementing. Without having to make it too complicated the back end could select random players every year to get a PA or CA increase, whether they are young players or older players.

This will allow for the potential of a late bloomer as well as young stars who have rapid growth spurts in skill. It will also allow for additional variance in each game.

I like this idea. Even if CA/PA are not sole determinants of player ability, it would be interesting to have the possibility of one's PA move about a bit.

[mack4ever]

My idea would be to effectively give every player two PAs.

One is the current system of PA, so a static value that is assigned to the player which dictates their maximum possible CA.

The second is a randomised PA that each player is assigned that can replace their static PA dependent on certain factors, a PPA (Potential Potential Ability) if you will. When this happens is determined by things such as the reputation of the club and the standard of the training facilities that the player has access to.

So for example let's take two players.

Player A has a PA of 120 and a PPA of 150 and is playing for Bristol Rovers.

Player B has a PA of 120 and a PPA of 120 and is also playing for Bristol Rovers.

Player B will never be able to reach a CA higher than 120 no matter which club he plays for.

Player A on the other hand can. If he stays at Bristol Rovers and their training standard stays the same then his PA will stay at 120. However if he moves to somewhere that will give him a higher standard of training then his PA can increase up to the maximum value dictated by his PPA. So if he takes a small step up to Bristol City, receiving a slightly higher standard of training then his PA might increase to 130. If he moved up to Birmingham then it might go up to 135, and if he was picked up by Manchester United then it will go up to nearly 150.

As the players PPA is random it doesn't mean that all players will suddenly have more potential if they're signed by a bigger club, but it would mean that some players will thrive when given the chance, and unlike now where you can see exactly how good a player will be before you've even signed them, this would introduce the element of luck again, signing a young player who is good for his age but isn't predicted to develop that much, he might now be worth picking up because there is a chance that he will really push past the initial expectations.

[Guest]

As ever that only addresses the issue that researchers can be mistaken on a player's potential & why is it that being unable to apply hindsight to the starting data is such an issue?

If a researcher believes that a player could reach a high level should they got a move to a top club with better coaching then they should already be rating them with a PA value to take that into account & in game if the player doesn't get that move & gets stuck in the lower leagues they will not reach that high PA value because they will not get access to the coaches, facilities & level of opposition that will drive them to improve.

Edit: People are just seeking to introduce additional data points & complexity where none is needed, all this will achieve is a greater chance of data errors or coding bugs for no tangible gain in the end product.

[Overmars]

Edit: People are just seeking to introduce additional data points & complexity where none is needed, all this will achieve is a greater chance of data errors or coding bugs for no tangible gain in the end product.

The tangible gain would be a more realistic representation of how a player's ceiling can adjust over time. It never made any sense to me that a fast developing player could have his CA equal his PA before he turns 25. My staff tells me that the player is already playing at his full potential, and I wonder how that can be possible when we can always make him a little bit smarter or technically skilled.

[Guest]

So you've never come across a player in real life who showed promise as a youngster but failed to develop? Everton have produced a fair few over the years.

Could be suggested that Fernando Torres hit his peak in his early 20s, stalled & then regressed before flirting with an Indian summer as his career winds down.

[scott MUFC]

Ravel Morrison surely best example.

[Guest]

He never really peaked, such a great talent but very a troubled young man.

[Mrlee.1986]

So you've never come across a player in real life who showed promise as a youngster but failed to develop?

Step forward Freddy Adu...

Huge potential at 14 years old with DC United, and then it all went sour.

[cseyoum]

The major thing i like about FM has always been the development of players, the current system impedes my fun and affection for the game without editor. Game needs to be more dynamic with regards to the potential carreer trajectories of players.

[Guest]

What do you mean by more dynamic?

What's aspects of the current approach do you find lacking & can you qualify with examples?

What changes would you propose & how would they be a positive change on the current system?

[fmfan74]

I take it many people forget ballotelIi. Sensational skills. World class player someone who you would choose over vardy any day only if he had the same or half of mentally focusing ability. I don't want selfish players in my team I want a strong mental team that play sensibly.

From what I can see a lot of people are not looking at mental stats enough.

[swansongs]

Step forward Freddy Adu...

Huge potential at 14 years old with DC United, and then it all went sour.

Adu wasn't actually 14 at the time, though.

[Mrlee.1986]

Adu wasn't actually 14 at the time, though.

How? He signed his first contract at 14, debuted and scored at 14 so how was he not 14 at the time? ?

[Guest]

I think it is a generally accepted but unspoken truth that his mother lied about his age, iirc it helped their US residency application oby having a dependant who was a minor.

[Mrlee.1986]

Oh ok, to be honest I hadn't heard about that - or even seen anything like that before as I checked 'facts' before posting - although I read somewhere he's a nightclub promoter now lol

[mack4ever]

Oh ok, to be honest I hadn't heard about that - or even seen anything like that before as I checked 'facts' before posting - although I read somewhere he's a nightclub promoter now lol

Well somebody might want to tell the Tampa Bay Rowdies that...

[bwilliams23]

As ever that only addresses the issue that researchers can be mistaken on a player's potential & why is it that being unable to apply hindsight to the starting data is such an issue?

Predictability. Might just be my personal preference, but I find it quite boring to run different saves for 10-15 years and see the same group of core players winning world player of the year, playing for England, being signed for big money etc. on each of those saves.

To put it another way, I also find it quite boring that it big-5 teams will tend to win the Premier League for a number of years before another side breaks through. With the game being left to its own devises (i.e. running a simulation or not being a PL team), how often do we see a non big-5 team winning the PL in 2016?

Edit: People are just seeking to introduce additional data points & complexity where none is needed, all this will achieve is a greater chance of data errors or coding bugs for no tangible gain in the end product.

So I'll refer to what I said before; higher PA, harder to achieve. No additional data, work behind the scenes on the game engine but no additional complexity for the player.

Tangible benefit? If it introduced greater variability or unpredictability then for me there would be, but again appreciate that might just be my personal preference.

[enigmatic]

So you've never come across a player in real life who showed promise as a youngster but failed to develop? Everton have produced a fair few over the years.

Francis Jeffers is the ultimate example. The only level he ever looked any good at was the Premiership, as a teenager.

-

Vardy had a mediocre (still high by conference standards) potential in FM12 because researchers perfectly reasonably underestimated his potential, not because every conference level youngster has a random non-zero probability of emulating him. And since they underestimated his potential because of his low ability relative to his age at the time, removing PA limits wouldn't really have solved the problem of him never turning out any good in the game anyway. But it would have also all but guaranteed an early-developing but somewhat limited 18 year old like Francis Jeffers had the opportunity to develop to Messi-like levels given regular game time, instead of just becoming a useful part of the England setup and regular goalscorer as actually happened in the CMs from that era.

The whole reason PA exists is to prevent the opposite scenario, where you can absolutely guarantee the majority of your Premiership youth academy become international stars simply by maxing out coaching and giving them game time from a young age, players with good starting attributes reliably develop linearly into exceptional finished players and players with low starting abilities can safely be ignored because they'll definitely not grow further and faster than their peers.

And of course real life coaches and scouts can identify that of two teenagers of similar ability, one has a good chance of reaching the first team and the other isn't necessarily guaranteed a professional career. In real life it's a bit more tied to tangible factors such as "this player is the best in his age group but relies solely on being bigger than the other kids and hasn't really improved his game awareness after three years of top coaching" vs "this player is very raw after we found him at a low level, but he already has good technique, learns quickly and we reckon after he's developed match awareness and spent some time in the gym he could be a future first teamer" and far from an exact science, but a game where coaches are usually (and inexplicably) pretty accurate in their predictions is better than one where they don't assess potential at all.

[Torminator]

There is also the "Kluivert" problem, players peaking at 18-21 and then steadily declining and burn-out at 28.

For me its not an issue that this is not implemented in the game. I would be bummed out if my worldclass talent declines at an age that should have been his peek age.

Not being able to code in "Vardy-type" development is ok by me. I look forward to next edition and to check which talents and players are updated.

And as mentioned by the mods, a 150 CA/PA player with the right distribution is a world beater and can easily outperform a 180PA player. Its so much fun to create a team of 150CP/PA players that just click and do the right things and help you win the league!

[YKW]

There are some bits of the game that I think could be much better, and some of these are limited by the way CA/PA works, in my opinion.

For example, if my objective is to sign good youngsters, then I just need to scout, scout and scout some more, perhaps with a manual scout of some good clubs youth teams, and then just look at the ones that have the highest recommendation. Similarly, it is far too easy to disregard large numbers of players based on the same.

Why would I ever sign a 2 star CA 2 star PA 18 year old, if I have anyone available with at least 3 Gold PA stars? Why would I even consider signing a player with less Gold Stars PA than my current first teams' CA? Maybe this is just me, and everyone else regularly signs young players who scouts say cannot improve anymore over other players with higher PA stars.

If you have a player with no PA stars left to fill, but who is young and has good personality traits, on what basis are all of my scouts reporting to me that he has no prospect of improvement? It's because they KNOW (based on having some awareness of PA) that the player can't improve, NOT that factors other than the PA value suggest a limited development from that point.

Consider also youth intake day. It's obvious immediately whether or not you have anyone who is going to be any good. Sure, your 4 Gold 5 Black PA newgen could actually have a PA just above the CA of your first team, but those 1 Gold 2 Black star players are basically just a waste of time. These guys are so young, but so easy to disregard.

Debating whether or not a fixed PA makes sense is only one part of the issue. It's about making things interesting and unpredictable for the user in the context of the in game mechanics, like scouting and youth intake.

Now I have tested the above on this version of the game by using a scouting program to identify young players with great determination, professionalism and ambition, but who are at or close to their PA. Scouting these players is always the same result. None of my scouts think the player can improve.

A similar test was done by duplicating a player (no 0 entries) to end up with 4 copies, each with a different PA. Despite there being no difference in the players apart from their PA value, all my scouts, good and bad, differentiated between them successfully.

I do not believe that the above can be solved AND the current CA/PA system be left exactly as it is. With the current available information to scouts, they HAVE to be able to perceive PA either directly or indirectly in order to have any function within the game at all. I don't think Perceived PA is the answer because it's not possible that this isn't somehow a function of PA. Players who are similar apart from PA will always be too easy to distinguish between.

This is why I think some kind of development curve type thing has to be in the game moving forward. I think it is the single most obvious solution to the weaknesses in game play.

There could be 8 or so development curve templates, such as the old-school PES' "early lasting" and "late peak." These could be generated upon game start based in part due to CA/Age (a 22 year old with CA 170 isn't going to be a player that peaks late) and natural fitness, but also largely random.

The development curve would then govern all 'reports' from within the game. Scouts would now judge potential based on Age, CA, and likely improvement over the next X years of the development curve, where better scouts have a higher value for X.

So for example, and I'm going to use Youth Intake as that is one of the worst bits for me, imagine you have three players come through, all 17. For the sake of the example also assume that they have similar personalities.

Player 1: CA 80 PA 90

Player 2: CA 90 PA 190

Player 3: CA 100 PA 120

In the current system, Player 2 will ALWAYS be reported as the best, which he is. This will likely be the case whether your Assman is 20/20 JPA/JPP, or 1/1 JPA/JPP.

So now you assume the system whereby development curves have been generated at the "births" of the players.

Player 1: CA 80 PA 90 - Curve: Early Lasting (early CA growth, then sustained until an old age)

Player 2: CA 90 PA 190 - Curve: Late Peak (initial slow CA growth, then quick CA growth after around 25, and a fairly decline after 30.)

Player 3: CA 100 PA 120 - Curve: Early Peak (initial fast CA growth, then fairly quick decline after 26 or so)

Let's say you compare the intake while having an Assman 20/20 Judging, compared with one of 1/1 Judging.

Our 1/1 Assman can only make an informed judgement based on his forecast development over the next 1 year. Let's say our 20/20 Assman can make his judgement based on forecast development over the next 6 years. They have to guess (or extrapolate based on their skill level) for development beyond this point.

Assman 1/1 Intake (judging PA after 1 year, age 18)

Player 1: Reports PA 87

Player 2: Reports PA 93

Player 3: Reports PA 110

Assman 20/20 intake (judging players after 6 years, age 23)

Player 1: Reports PA 90

Player 2: Reports PA 130

Player 3: Reports PA 120

So immediately you have Assman 1 who can barely distinguish between Players 1 and 2, and reports 3 as the best prospect, as far as he knows. Assman 2 doesn't differentiate much between Players 2 and 3, but correctly identifies 2 as the best prospect, although he still underestimates him.

You can extend this kind of uncertainty to scouting, with the true PAs of the players ALWAYS totally separate from the in game agents. Being reported to that a player can't improve by no means even implies that he won't improve. Players reported as being likely to improve a lot may very well do so, but may still end up with a lower CA than a late-blooming player who was reported as having no development.

In the examples above the players have almost identical personalities. If you consider the effect of personality on development, and a scouts impression of PA (PPA), then this would further enhance the uncertainty as the game unfolded. For example, a scout would apply their personality multipliers after they have done they development curve analysis. E.g. Scout studies the development curve and concludes that the player could have a CA of around CA+40 in 3 years time, representing a possible CA at that time of 140. If the player has an awful personality, then the scout would downgrade their recommendation. If the player has a great personality then the scout may extrapolate their 140 figure after 3 years, to a proposed 160 beyond this point (after which their scouting ability dictates they have to guess.) This could happen even if the player has an actual PA of 140.

[Guest]

If you have saves showing all this I'd suggest raising them in the bugs forums.

[enigmatic]

To be honest, pre-built development curves that determine from birth that a newgen's CA growth will speed up at a certain point in his twenties and scouts that can predict this on a six year time horizon sound a lot less realistic than the current state of affairs where it's actually mostly affected by when and whether they're playing enough at the right level.

(you could, to be fair, probably pre-build a development curve that determines how quickly they'll physically grow and whether they'll stop at 18 or 22, which might offer scope for scout feedback that leaves you with interesting decisions over whether you still want that centre half your scouts think has plenty of room to improve their game but will remain quite small and weak, or whether that 16 year old centre back with 6 for jumping might actually be OK in the air by the time they break into the first team)

I do quite like your idea that scouts' estimates of potential for very young kids should be biased upwards or downwards based on the player's personality (though it appears to already be already a factor in whether the scouts assess a player as a "good signing" or "some doubts" or "not worthwhile").

As frustratingly redundant as all those 1.5 black star newgens are (it's even worse when you're playing a Youth Challenge and have to seriously consider whether they might actually have the attribute distribution to be useful backup in a higher division in future!) it's probably not an entirely unrealistic situation for coaches that have worked with kids for a couple of years to have a pretty good idea which players aren't likely to make it even though they're good enough for the U18s league even on "trial day"

[mack4ever]

Would the simplest solution to this problem not be to simply make it harder for in-game scouts to assess a players potential?

I mean, right now we can tell pretty much exactly how good a player will be after having a scout look at them for a couple of matches. What if it was a much longer process. So no matter how long your scouts watch a player they're never going to have more than a rough estimation of their potential, the only way you'll know for sure is by signing them and having your coaching staff assess them over an extended period of time.

Without any real changes to the PA system you've now got a game where there is a proper element of luck involved when it comes to signing young players, where it's harder to predict exactly who will turn into a great player and who won't.

[AllyJoseph]

I actually just think you are flat out wrong about no late developers in the game.

The big English clubs always sign kids or bring through a host that get released at 23 and go on to have great careers.

Go on your most recent save, look at Chelsea, pick a regen with good potential but poor CA at 23.

Sign him and give him game time. 10/10 times he will go on to be a Prem player.

[Guest]

Ally, that kind of sweeping statement is no better than those with a contrary position, of course as with my reply to YKW if you have save game evidence that 100% of newgens at a club like Chelsea have the potential to make a career in the Premier League please do raise that as a bug.

mack, I do agree with you that a scout's ability to get a a full picture of a player in no more than a month is going IMO less than ideal, moreso when that player doesn't make any appearances. The problem is that FM has to strike a balance between realism & fun, while the current system isn't perfect I do think it has that realism:fun balance about right.

Those who want more realism can use scouts in such a way to be more realistic, I only send my scouts on competitom assignments & I do not have them constantly watching each & every player they report on.

[enigmatic]

Think Ally's point is more that there are always a few high potential players that develop late, which in the game is usually due to big clubs not giving them enough chances to shine

(more Danny Drinkwaters and Ryan Shawcrosses than Jamie Vardy).

So the game mirrors real life in that respect, though I'm a bit disappointed with Man Utd and Chelsea's treatment of a England U21 PA 186 regen with great personality in my experimental holiday game, who I'd have happily given game time to based just on his starting stats and PPMs (at 27 instead of being the next Stevie G he's now a rotation player at Newcastle, hasn't played more than 10 games a season since he was a teenager and has accepted a callup from Wales). Still, I've noticed he is very injury prone, so possibly that's held his development back more than his clubs' selection policies, which again is fair enough.

[AllyJoseph]

Yeah my point was basically to show there are late developers, hell there are late developers that show up not at the big clubs. Just one common example is the big clubs who obviously get the better youth, the release them.

Your main issue OP is that you are assuming that Jaime Vardy's of this world do not happen within the game. For that to be true you'd need to know ever single regen, at every single club, at every single level and see their development. You can't see that. Its really hard to argue that Johnny Newkid on my save had a similar career to Jaime Vardy because the same kid on your save stayed at Colywn Bay his entire career before retiring at 27.

I would argue the onus is on you to prove that not one single regen on any fm16 save you've had has started low, developed late and ended up in the Premier League. It's a neigh on impossible task.

It is far more realistic that young kids that do well in lower leagues get signed by big clubs and either develop or fall by the wayside for every Rickie Lambert and Jaime Vardy there are 10 Gareth Bale's that were sold in League One to the prem as a youngster and go on to have top careers. Just like there are 10 John Bostocks who dont cut the mustard.

[RocheBag]

Eh? Why would any club in real life sign any young player if any of that was true?

What? I didn't say young players don't improve....

[aderow]

Every time this discussion comes back I say the same thing. I don't believe that there's an issue with the CA/PA system. I do believe that the player development model should be more volatile. That would address some of the issues people have in regards to late bloomers etc.