Jump to content
Sports Interactive Community
Erith22

Increase the randomness of player development!

Recommended Posts

As someone who plays long term games, I think the #1 biggest failing long term is that the cycle of 'buy young, develop, sell for huge profit' is pretty much automatic.

What I mean is that, assuming you've set up a good training program, you cycle in young players into your squad, and avoid crippling injuries (which are very, very rare), at the top level you can routinely sell a few players a year at a 30-50M euro range. It's totally worth it to buy 3+ star prospects for 10-20M euros because you KNOW that you'll get double that in a couple years. Since the AI doesn't do this, it gives the human player a huge advantage. The AI simply won't pay 20M for a 16 year old prospect. They will pay huge money for an 'established' player. In real life this makes sense because you know pretty much what you are getting with a 24 year old. But the problem is, in FM, you really know pretty much exactly what you are getting with a 16 year old too, so the whole mechanism of buying and selling breaks down.

In real life, there's a huge chance that young prospects will fail. That's why you don't see 16 year olds going for that kind of money on a regular basis. Not because they lack potential, but rather because so much can go wrong along the way (major injuries, personal problems, drug problems, lack of motivation, poor effort in training, etc, etc). In FM, development is all but automatic if you know what you are doing.

In Out of the Park baseball, they have implemented a function where a player's PA can undergo random ups and downs over the course of their development. Ideally, far more players should start with high PA, and there should be more downs than ups, though it should be balanced so the overall PA level of the world stays about the same.

I feel like FM desperately needs this. There should be a big risk involved in paying huge money for an undeveloped 16 year old.

A sampling of my transfer spend over last several seasons in FM 2014

In: Out:

26M 84M

22M 97M

24M 65M

11M 67M

159M 264M

24M 219M

44M 171M

61M 185M

All this while winning La Liga every year by a huge margin and going deep in Champions League. With Tenerife. And I only use in-game scouting.

Share this post


Link to post
Share on other sites

OOTP definitely handles potential and career development much better, along with having so much more usability in it's GUI. That said, FM is much more long-term playable than OOTP due to roster management and general AI. Given OOTP used to be part of SI, I'm surprised they haven't adopted some of the better elements in FM

Share this post


Link to post
Share on other sites

OOTP and FM have just made very different game design decisions.

In OOTP the user can choose the level of scouting accuracy. You can even set the game up so that you always have 100% accurate knowledge at all times. Knowledge is also fixed in the sense that a 3 star player will always be a 3 star player no matter what team he is on. In FM scouting is imperfect and relative. A 3 star player for one team might be a 2 star player for another.

OOTP adds a bit of randomness to simulate prospects who level out in the minors or take off and exceed previous projections. If I'm being fair, I'd say changing the PA in a baseball sim reflects that at different points in time scouts project players differently. Take a prospect like Bubba Starling. He was thought worthy of the fifth pick in the draft and a $6 million signing bonus as an 18 year old. Scouts might have said that he had 5 star potential at the time. Three years later, Starling is struggling to stay above the Mendoza line in High A ball. Now if you have heard scouts talk about him, then you know that don't rate him as much of anything. But, it's not that his potential has really changed. He never really had the potential. Projecting amateur baseball players is exceedingly difficult and inaccurate. When Starling was drafted, the Royals knew they were taking a big risk. He was a boom or bust prospect. What the team knew was that he might have Hall of Fame potential, but that he might also have the potential to never make the majors. They didn't know his real potential and took the chance that it was high. OOTP would simulate this by giving Starling the highest in game potential and then randomly lowering it dramatically later. But, that's not really what happened. The OOTP solution is really a gamey solution that's somewhat necessary because a large portion of the game's user base do not want scouting knowledge in the game to be nearly as inaccurate as it usually is in real life.

FM doesn't need that because the randomness is built in. Your scouts aren't always right. They probably are far too accurate when you have scouts with great attributes. Even when they miss, they don't miss by very much or in a way that really costs the user. I suppose that could be tweaked. But, there are things you can do to make it more difficult if you want. I believe that there is a skin that hides the star ratings or you can just not purchase every high potential regen in the game.

Share this post


Link to post
Share on other sites

It's a massive part of the game, which i hope they re-work for FM15... scouting is far too accurate/too easy to nab the best scouts, and player dev too linear.

I would suggest doing away with scout attributes - their ability should be hidden, and instead replaced with a list of their recommendations that have made it.

Share this post


Link to post
Share on other sites
It's a massive part of the game, which i hope they re-work for FM15... scouting is far too accurate/too easy to nab the best scouts, and player dev too linear.

I would suggest doing away with scout attributes - their ability should be hidden, and instead replaced with a list of their recommendations that have made it.

Player development isn't too linear and staff are often wrong, more often than many people seem to think.

There needs to be more variety in staff opinion though rather than the 1/2* difference we currently have.

Share this post


Link to post
Share on other sites

My biggest peeve is it takes scout one report (in a day) to get the sort of information it would take a real life scout a couple of trips to obtain. Too much is available too soon IMO.

You can set yourself your own rules such as having to scout a player in three matches, or watch him yourself etc, but it's still pretty easy to grab the best talent just off the initial reports.

Share this post


Link to post
Share on other sites

The problem is that scouts know how good a 16-year-old who has never played competitive football is at that exact moment. Even if knowledge of PA was completely removed, that evaluation makes it possible to find high-CA youngsters and thus with good precision predict a possible potential. In real life, if a player has not played competitive football nobody would know how good he is.

If a 16-year-old becomes the top goalscorer in the Greek league, you know that he's a great talent. If the same player has not had his debut yet, you wouldn't even know about him. The need for the "not for sale" functionality would disappear because we would realistically not be able to identify wonderkids before they actually had an impact on their domestic league.

Share this post


Link to post
Share on other sites

The PA solution would likely be the hardest, given how touchy the FM engine is. It's the most reliable, since it couldn't be gamed. Increasing scouting randomness helps, though eventually a smart human player can just aggregate many different reports and make a pretty solid judgement.

I feel like the PA solution is necessary because prospects SHOULD fail sometimes, and they really never do. The only other way to solve this would be to make the AI just as aggressive in bidding for 16 year olds as the human player. That would make it more competitive, though it would turn the game on its head as far as a simulation goes.

Share this post


Link to post
Share on other sites
The PA solution would likely be the hardest, given how touchy the FM engine is. It's the most reliable, since it couldn't be gamed. Increasing scouting randomness helps, though eventually a smart human player can just aggregate many different reports and make a pretty solid judgement.

I feel like the PA solution is necessary because prospects SHOULD fail sometimes, and they really never do. The only other way to solve this would be to make the AI just as aggressive in bidding for 16 year olds as the human player. That would make it more competitive, though it would turn the game on its head as far as a simulation goes.

Why do you think they don't fail?

Because they reach their perceived PA based on the stars? If so your logic is flawed.

I'm pretty sure I read a post this year or last where SI stated something like 70% of newgens don't reach their actual PA ever based on their test saves.

Share this post


Link to post
Share on other sites
If the same player has not had his debut yet, you wouldn't even know about him.

Well that's just nonsense. Real life scouting is a little more in depth than just reading Sky Sports match reports.

Share this post


Link to post
Share on other sites

I'm pretty sure I read a post this year or last where SI stated something like 70% of newgens don't reach their actual PA ever based on their test saves.

That number drops significantly when a human is in charge of the club. As long as you get the player early enough, you can pretty much guarantee they'll reach their PA, unfortunately.

Share this post


Link to post
Share on other sites
Well that's just nonsense. Real life scouting is a little more in depth than just reading Sky Sports match reports.

Well what are they doing then? Watching youth team football where he competes with teens who are still tagging along for the social bit?

Share this post


Link to post
Share on other sites
That number drops significantly when a human is in charge of the club. As long as you get the player early enough, you can pretty much guarantee they'll reach their PA, unfortunately.

So that begs the question why are the players improving significantly better with a human in charge?

More forward thinking? more preparing for the future and giving them more game time than AI managers?

You can block out PA stars and make staff worse at judging but if it all comes down to playing time then it won't help. The answer has to be doing something with the rate of improvement so even human users can't max players out as easily.

Share this post


Link to post
Share on other sites

Yes.

To use an example, Wayne Rooney was spotted by Everton while playing in the Walton and Kirkdale junior league at U-11s level.

Clubs regularly watch local junior leagues in the hope of spotting someone special.

Share this post


Link to post
Share on other sites
Well what are they doing then? Watching youth team football where he competes with teens who are still tagging along for the social bit?

Pretty much.

Around here there are registered independent FA clubs taking kids from around 4yos. The league & non-league clubs in the area build a relationship with these clubs and local schools with with the aim of spotting the talented kids.

Didn't Messi get spotted by Barcelona when he was 10yo or something?

Share this post


Link to post
Share on other sites
So that begs the question why are the players improving significantly better with a human in charge?

More forward thinking? more preparing for the future and giving them more game time than AI managers?

You can block out PA stars and make staff worse at judging but if it all comes down to playing time then it won't help. The answer has to be doing something with the rate of improvement so even human users can't max players out as easily.

They get more games when they're young, something that doesn't really happen with the AI often. Most players under AI control will reach a middling level of their PA. A human manager will see a huge prospect and actually nurture him. The AI will mostly neglect them and, as the best prospects from the lower leagues get snapped up by the larger teams, they get completely neglected and waste away. Usually, by the time the AI starts playing the kids, it's too late.

Share this post


Link to post
Share on other sites

Human managers actually use the tools available to them: tutoring, individual training, PPMs, loans, first-team game time in a way that tends towards maximally efficient. The AI is terrible at all of the above.

In a long-term save, I have to build all of my players from a young age, because anyone I buy "premade" over the age of 23 will have a terrible personality and wasted potential in stats they don't need. In general, the hidden stats of AI-controllled players fall off a cliff as real players are replaced by regens in a long-term save.

Share this post


Link to post
Share on other sites

The real answer is a complete overhaul of the CA/PA/Reputation systems that will narrow the gap between human valuation and training of youngsters and AI valuation and training of same. But that will probably never happen.

Share this post


Link to post
Share on other sites

I think broadly three things are needed

1) it should be less easy to pick out the talented youngsters

2) AI should develop their players better and hang onto the perceived future talents better

3) fewer youngsters should make their potential even if perfectly handled

Lots of people on here have been saying these things for several years but it remains as easy as ever to outwit the AI over the long term with developing youngsters a key part of it.

Share this post


Link to post
Share on other sites

When you see a 16 yr old with 4 star PA, he's almost guaranteed to be world class. there should be a bit more randomness i agree.

right now it's too rigid and not realistic at all.

Share this post


Link to post
Share on other sites

That's just plain wrong. PA stars at 16 are wildly inaccurate.

Share this post


Link to post
Share on other sites

My solution? Play FMC where you have no control over PPMs, individual training, or mentoring. They play and do well, they get better. If they don't, they don't.

Share this post


Link to post
Share on other sites
My solution? Play FMC where you have no control over PPMs, individual training, or mentoring. They play and do well, they get better. If they don't, they don't.

someone is really putting an effort to his playing career :lol:

Share this post


Link to post
Share on other sites

The way to 'fix' this is just have the AI take care of their kids better. If it's a small club, and you offer them 20m for their 16 year old, of course they will sell him on, this is fine. If Bigger club AI were to look after their kids better, not sell them on or train them much better, then that would be a perfectly fine challenge. You'd know that particular kid is very good, but off limits to you, and you'll have to find a window to get him elsewhere.

So yeah, just having the AI take better care of their kids would be good, although that's easier said than done I believe.

Share this post


Link to post
Share on other sites

FM doesn't need that because the randomness is built in. Your scouts aren't always right. They probably are far too accurate when you have scouts with great attributes. Even when they miss, they don't miss by very much or in a way that really costs the user. I suppose that could be tweaked. But, there are things you can do to make it more difficult if you want. I believe that there is a skin that hides the star ratings or you can just not purchase every high potential regen in the game.

Once again, you shouldn't run wearing two left shoes because your opponents are slower...

The issue here isn't (just) FM's scouting being way too accurate and way too quick even with mediocre scouts. It's about players developing in a rather linear and predictable way that makes most of the scout-buy-develop-sell process too safe.

Case in point: you get raving scouting reports for two strikers "you should sign whatever the price". Both have 4.5* potential. Then you look at their attributes: one is 18 and is a well-rounded forward, CA 130 PA 160. The other is 20 and is technically gifted, but lacks in mental strength and/or is desperately slow, CA 130 PA 180.

AI clubs will likely buy the latter because of his higher PA, no matter how flawed he is, and also failing to acknowledge at 20 he has fewer years left to reach his PA... something that is unlikely to happen anyway.

But back to the well-rounded forward, unless he picks up serious injuries and/or doesn't get enough playing time at the club (which rarely happens for high PA players), he'll fulfill his potential.

Then if he's at the human manager's club, it's pretty much a given he'll become as good as he possibly can within a few seasons.

I know hidden mental traits and personality can affect how fast and how much a player will improve, but in too many instances all they need is a high PA and match experience at a good level of football.

There isn't enough diversity and it's too easy to spot the 5* players who'll make it and those who won't.

FM needs:

1) different growth curves, based on personality and on in-game experiences (that'd allow us to have actual Late Bloomers, Early Bloomers, Flashes in the Pan etc)

2) Dynamic PA within the db-set range (a player can "lose" some PA following a poor season, but should also get some back if he finds his way again. But nothing higher than the original PA of course, to avoid the "everyone can become Messi" risk/argument)

3) Less accurate and much slower scouting process, with the elimination of the "Potential" factor for unproven players... There's no way a scout can say a 17yo who's barely able to make an impact at Reserve level will be a future EPL star".

Also, a rework of the CA+PA+Reputation criteria for AI's signings would be great, but again, that's a different part of the story.

Share this post


Link to post
Share on other sites

or most of you could stop using genie scout and similar and just play the game, unless regen is wonderkid and there are 5-10 in whole game at a time you wont have 20 stars in your squad and most will never reach their potential, if you are putting that 2.5 star out of 4.5 in your squad from age of 20, you are already sacrificing a lot to make sure he grows, people buy regen for 10mln, sell for 50, after 5 years, forgetting how much effort they put and wages they paid and they imagine they made huge profit, do you need those extra 20 mln you made at this time, all you need is 11 world class players to "win" the game, whats the point in buying all regens and selling them 5 years latter, btw most regens develop way faster and better with ai than with human, i have never had my world class regen to become league star by age 18-20 yet I have seen AI regen become just that 24/7 and in the end ill buy that regen, because hes 3 times the player mine is at its age, even if both will end up being similar 5 years down the road.

I had so many 4.5 and even 5 star players become 2.5 and less star players over a year, over 2 years, over 3 months,

4.5 winger ended up as 2 stars. another 4.5 winger, was 3.5 never got to 3.5, and eventually went to become 2.5 because of injuries, same for many other players, signed 5 star gk, that ended up being 3.5 star, scouting might be accurate at times but its sure **** other times

Share this post


Link to post
Share on other sites
spot on Kalle

Two seasons ago i had a CM (retrained striker) who just couldn't stop scoring whole season. He was always confident, never overconfident and as I waited for him to hit the wall and get on scoring drought he never did. He finished the season with 28 goals in 30 matches and was 2nd best goalscorer in the league.

Despite the magnificent season for a fairly mediocre player (he is good all round but not excellent in any of his attributes) his reputation, ability, interest from other clubs or national team manager never came as you would expect in real world.

Such a performance in any top/medium league in europe would surely stir some dust but in game, nothing happened. It is way too hermetical and predictable. In this instance i blame player's reputation and PA/CA system that prevent from happening what would happen in real world in this instance.

Yep, you could have a player score 30+ in every division as you fly up the leagues, and not even got a flicker of serious interest if his PA isn't high enough.

Share this post


Link to post
Share on other sites

I would LOVE to see players progressing or not depending the tactical system , remmember santon at inter :)

So when u buy a hot brazilian and u put him in a wrong system for him he just hits the wall while his PA is there he aint progressing maybe peaking later in his career late 20's and whe saw that in real life too many times :)

Share this post


Link to post
Share on other sites

smikis and cougar - don't think you guys are on the same wavelength here... you probably haven't got a team setup or scouting setup where it just becomes so easy to spot the good 15-17yr olds and milk the system.

There is a issue here - there is absolutely no doubt, that once you see the system for what it is, you can easily spot youngsters that will make it , either leading top div player or world class.. and mostly they make it. Not all of them, but most, especially those with the right mental attributes.

Scouting is too easy, player development isn't properly tied to performance, and players do develop quite easily and in a linear manner if you follow some simple steps (good coaches, loans, first team exposure). Not too many late bloomers - i've not noticed many players who stall due to lack of playing time then being able to recover and still make it to the top if a team gets them late, eg. 23yr old and then plays them, also few early peakers who then get worse - they simply stay at their peak longer, nor many who peak then fall and then peak again. A bit more randomness would make the game more interesting.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...