FM 25 DEVELOPMENT UPDATE

[afailed10]

3 horas atrás, XaW disse:

I agree, it was better in FM24 than before, but there is still a way to go before I'm happy with it. Let's hope for more info soon, as this, the UI, and the match engine are the things I'm most looking forward to.

Before what? FM23? FM22? Possibly. But a clear downgrade from earlier versions such as FM18 and before. 
 

They’ve had 6-7 years to fix those issues, the ai incapacity to build reasonable and competitive squads, to make proper sense of recruitment, to develop youngsters instead of letting them rot in reserves.. but no! Tweak here and there afraid to make bad worse and slap a bunch of features we don’t really need..

In its core they’ve been asking 50€/year for cosmetic and database updates. Let’s hope we have a proper new game this time around.

[trevjim]

50 minutes ago, afailed10 said:

Before what? FM23? FM22? Possibly. But a clear downgrade from earlier versions such as FM18 and before. 
 

They’ve had 6-7 years to fix those issues, the ai incapacity to build reasonable and competitive squads, to make proper sense of recruitment, to develop youngsters instead of letting them rot in reserves.. but no! Tweak here and there afraid to make bad worse and slap a bunch of features we don’t really need..

In its core they’ve been asking 50€/year for cosmetic and database updates. Let’s hope we have a proper new game this time around.

Agreed.

 

I actually can't think of one "feature" that I actually use in the last 10 years other than perhaps  the ability to add/remove leagues between seasons 

The move to 64bit was good though and helped processing speed.

The QOL improvements are good. But they are not feature's.

[Petra90]

On 17/09/2024 at 13:30, XaW said:

I think this is due to how hyped "AI" has been in media and by the AI-industry. The way it's been talked about is almost like it will replace the jobs of everyone in the next few years and can do wonders on its own. The problem is, you need to be VERY specific in how to use it. I've seen people say "just use things like chatGPT to do press conferences". Which I'm sure SI could use, however the press conference answers actually impact how players or other actors in the game view you. And with just open ended text answers getting out the context and adding that into the wide ecosystem of FM sounds like it would be a nightmare! Imagine the testing required, both to make sure it would be impossible to get the game to not say anything controversial (see just about every racist/misogynistic/xenophobic comment made by just about any language simulator out there), but also how to make sure the game picks out reasonable reactions to what is being written by users. And for how much benefit overall? Would a majority of players do press conference if they had to actually type out all the answers? I doubt it. The return on investment would be so miniscule it's impossible to justify.

Now, language simulators might come a long way in the next few years, and if it become another common thing, then sure, why not, but I doubt FM would be the forerunner in this.

I think for static, repetitive stuff that doesn't affect other things AI could provide some interesting variety. Things like news reports text, fan texts on social feed, match reports texts and so on.

Another area where AI could be useful is Newgen realistic facial generation; the FM Newgan facepack is a great example.

[grade]

So technically we entering the last week of September.

When do you think we will get any news (good (the actual first images and video of FM25) or bad (another delay)), this week? Or are you in the camp it will be on September 30th?

[matisptfan]

10 minutes ago, grade said:

So technically we entering the last week of September.

When do you think we will get any news (good (the actual first images and video of FM25) or bad (another delay)), this week? Or are you in the camp it will be on September 30th?

My bet is Monday, September 30th.

[kiwityke1983]

10 hours ago, grade said:

So technically we entering the last week of September.

When do you think we will get any news (good (the actual first images and video of FM25) or bad (another delay)), this week? Or are you in the camp it will be on September 30th?

I'm hopeful for something either Wednesday or Thursday.

If we haven't had anything by then it will be Monday.

Purely because the last two announcements were Wednesday and Thursday.

Edited Monday at 01:01 by kiwityke1983

[XaW]

19 hours ago, afailed10 said:

Before what? FM23? FM22? Possibly. But a clear downgrade from earlier versions such as FM18 and before. 
 

They’ve had 6-7 years to fix those issues, the ai incapacity to build reasonable and competitive squads, to make proper sense of recruitment, to develop youngsters instead of letting them rot in reserves.. but no! Tweak here and there afraid to make bad worse and slap a bunch of features we don’t really need..

In its core they’ve been asking 50€/year for cosmetic and database updates. Let’s hope we have a proper new game this time around.

I think it's better now than it has been in a long time. Back in the versions you mentioned the squads usually got down to too few players as the teams almost didn't recruit player. Then in the last few versions before this, they did, but players they didn't need or use. And especially youth players were left in the dust. After all original players retired it got so much better, but the phase with both types were quite bad. @Daveincid and I did a lot of tests with this (which we reported to SI), and I think he found a way to tweak it in his files in the end (with reputation changes, if I'm not mistaken) to make it a lot better. In FM24 however, it's better than before, but still far from perfect.

So yes, I do hope SI will improve this for FM25 and the future, but I don't think it's as bad as it was before.

16 hours ago, Petra90 said:

I think for static, repetitive stuff that doesn't affect other things AI could provide some interesting variety. Things like news reports text, fan texts on social feed, match reports texts and so on.

Another area where AI could be useful is Newgen realistic facial generation; the FM Newgan facepack is a great example.

Oh, yeah, for sure, in those things it could be a nice addition for immersion, but thinking about AI and use it in things like press conferences or player interactions isn't really feasible at the moment in my view. Of course, the progress is going quickly, so who know some years down the line, but at present I think it's still not usable with these things.

[Daveincid]

vor einer Stunde schrieb XaW:

I think it's better now than it has been in a long time. Back in the versions you mentioned the squads usually got down to too few players as the teams almost didn't recruit player. Then in the last few versions before this, they did, but players they didn't need or use. And especially youth players were left in the dust. After all original players retired it got so much better, but the phase with both types were quite bad. @Daveincid and I did a lot of tests with this (which we reported to SI), and I think he found a way to tweak it in his files in the end (with reputation changes, if I'm not mistaken) to make it a lot better. In FM24 however, it's better than before, but still far from perfect.

So yes, I do hope SI will improve this for FM25 and the future, but I don't think it's as bad as it was before.

It definetly got better in FM24 compared to previous versions. In my opinion, there are very little issues within the code (SI-devs responsibility). The majority is in the database (those are two different things).

I'll give an example (there are other factors playing a role but this is a major one) about AI-squad management:

If you have 200 players with a CA140+ in 2023 but 500 in 2040, then there are several issues at once.

The most obvious point is, that the player strength is much higher (too high). This issue alone may doesn't look that bad. You could argue that footballers in the future will be even better as today.

But it's terrible for squad building! A lot of those players rot in U21 teams in the premier league because no other club can afford them (Financially & reputation wise). On the other hand, a lot of nations develop way too slow. So the issue goes into both directions. 

So what should the devs do? Slow down development speed in the code to lower the amount of too many worldclass players and make the issue in the majority of all other nations even worse? Should they increase development speed so that we will end with 100 CA200 players to solve the issue with the lack of development in smaller leagues?

Answer is simple: 

It's not their task to balance this issue in the first place. It's mainly a research/database task.

There is actually only one thing I wish that the devs should improve in FM25 in terms of squad AI: Teams should use more subs. I assume that the switch to 5 subs per game during the pandemic isn't properly implemented yet.

My two cents

Edited Monday at 07:30 by Daveincid

[grade]

17 hours ago, matisptfan said:

My bet is Monday, September 30th.

May bet is also on the 30th. Sooner would be excellent, but I have my doubts.

[Kevinho7]

I could care less about the whole graphical side for a game like fm. Always played 2d even… 

Also could care less if the ui is built in such a way that it takes me a couple more clicks to get to something. 

As long as the actual simulation part of the game is on point like match engine, squad building, player reactions/morale etc etc, I’m as happy as I can get. The power of fm, in my opinion, has never been about graphics, but about simulating the game world as close to irl as possible so we can have our universes unfold in their own unique way. 

Still love this game to bits, and i’m hoping it can improve to a level of pure satisfaction again. 

[scythian12]

3 hours ago, Daveincid said:

It definetly got better in FM24 compared to previous versions. In my opinion, there are very little issues within the code (SI-devs responsibility). The majority is in the database (those are two different things).

I'll give an example (there are other factors playing a role but this is a major one) about AI-squad management:

If you have 200 players with a CA140+ in 2023 but 500 in 2040, then there are several issues at once.

The most obvious point is, that the player strength is much higher (too high). This issue alone may doesn't look that bad. You could argue that footballers in the future will be even better as today.

But it's terrible for squad building! A lot of those players rot in U21 teams in the premier league because no other club can afford them (Financially & reputation wise). On the other hand, a lot of nations develop way too slow. So the issue goes into both directions. 

So what should the devs do? Slow down development speed in the code to lower the amount of too many worldclass players and make the issue in the majority of all other nations even worse? Should they increase development speed so that we will end with 100 CA200 players to solve the issue with the lack of development in smaller leagues?

Answer is simple: 

It's not their task to balance this issue in the first place. It's mainly a research/database task.

There is actually only one thing I wish that the devs should improve in FM25 in terms of squad AI: Teams should use more subs. I assume that the switch to 5 subs per game during the pandemic isn't properly implemented yet.

My two cents

One thing that is also very prominent in FM24 squad building is CA homogeniety. Teams will recruit to the extent of their Rep and financial possibilities, but not only in the "tip" but also the breadth of the squad. This effect is helped by AI player decision-making prioritising being a less important part of a stronger squad than being a more prominent member of a "smaller" team. One ends up with AI squads which have 16-20 players in the same CA range within 10 points or so. DB (and arguably, real life) squads are more varied, with a CA range of about 20-30 for first team squad. On one hand, the current approach ensures a competent keeping of standards by the AI, but also leads to a bit of a stale gameworld - there will be little movement on the whole for statures of different clubs and divisions. I would argue it also exacerbates the suffocating grip of the PL on all things Rep and player quality. PL teams can easily attract the very best players in the World (think 140-50-60 CA for impact subs and fringe players), making any PL team the pantomime baddie in European competitions. 5/8 PL teams in the top 8 of the CL, them also easily making the quarter final of the CL yesr after year, England getting close to 30 points in the 5-year-ranking, a 40-point lead overall, I hate it. 🤮

[Daveincid]

vor 8 Minuten schrieb scythian12:

One thing that is also very prominent in FM24 squad building is CA homogeniety. Teams will recruit to the extent of their Rep and financial possibilities, but not only in the "tip" but also the breadth of the squad. This effect is helped by AI player decision-making prioritising being a less important part of a stronger squad than being a more prominent member of a "smaller" team. One ends up with AI squads which have 16-20 players in the same CA range within 10 points or so. DB (and arguably, real life) squads are more varied, with a CA range of about 20-30 for first team squad. On one hand, the current approach ensures a competent keeping of standards by the AI, but also leads to a bit of a stale gameworld - there will be little movement on the whole for statures of different clubs and divisions. I would argue it also exacerbates the suffocating grip of the PL on all things Rep and player quality. PL teams can easily attract the very best players in the World (think 140-50-60 CA for impact subs and fringe players), making any PL team the pantomime baddie in European competitions. 5/8 PL teams in the top 8 of the CL, them also easily making the quarter final of the CL yesr after year, England getting close to 30 points in the 5-year-ranking, a 40-point lead overall, I hate it. 🤮

agree but there are again tools within the db to keep it more varied -> player personallties. If players remain happy without getting lot's of playing time, they do not force a transfer. (Eg. too high professionalism, less controversity and other characteristics). There is a bit of a change needed within the code to make newgen player personalities a bit worse or lower the impact of mentoring to get a better/more various representation of player personalities and their consequences in longterm-saves. 

That's just one onther factor of dozens to get the game to a more realistic level.

[Proz]

6 minuti fa, Daveincid ha scritto:

agree but there are again tools within the db to keep it more varied -> player personallties. If players remain happy without getting lot's of playing time, they do not force a transfer. (Eg. too high professionalism, less controversity and other characteristics). There is a bit of a change needed within the code to make newgen player personalities a bit worse or lower the impact of mentoring to get a better/more various representation of player personalities and their consequences in longterm-saves. 

That's just one onther factor of dozens to get the game to a more realistic level.

Having players continuously complaining because they are not playing enough, even if they supposed to be play few matches, is not at all realistic.

Furthermore, their complaints involve the necessity of a lot pf repetitive micro-management which is not at all fun.

 

[Daveincid]

vor 1 Minute schrieb Proz:

Having players continuously complaining because they are not playing enough, even if they supposed to be play few matches, is not at all realistic.

Furthermore, their complaints involve the necessity of a lot pf repetitive micro-management which is not at all fun.

 

If you want a better AI-squad management then this is part of your job as a manager. I won't argue about that anymore. There will always people with other views.

[srvngrc]

I hope the new engine is optimised for Mac Silicon. FM 24 was running very hot on M1. FM 25 would be cooler I hope. 

Edited Monday at 12:36 by srvngrc

[Bahoi]

5 hours ago, Daveincid said:

It definetly got better in FM24 compared to previous versions. In my opinion, there are very little issues within the code (SI-devs responsibility). The majority is in the database (those are two different things).

There is actually only one thing I wish that the devs should improve in FM25 in terms of squad AI: Teams should use more subs. I assume that the switch to 5 subs per game during the pandemic isn't properly implemented yet.

 

 

But how do we address the issue of AI teams not making enough transfers or letting players go and so on. For example, in a save I have in 2027, top 4 teams of the top 5 leagues in Europe (Bayern, Real, PSG, City etc etc, but also almost any EPL team) are using  almost the exact same starting 11 they used since the summer of 2023?

Edited Monday at 12:52 by Bahoi

[Daveincid]

vor 2 Minuten schrieb Bahoi:

 

But how do we address the issue of AI teams not making enough transfers or letting players go and so on. For example, in a save I have in 2027, top 4 teams of the top 5 leagues in Europe (Bayern, Real, PSG, City etc etc, but also almost any EPL team) are using  almost the exact same starting 11 they used since the summer of 2023?

I already adressed all the points I raised here. I'm optimistic that some of the points will be taken into account for FM25

[trevjim]

1 hour ago, Daveincid said:

If you want a better AI-squad management then this is part of your job as a manager. I won't argue about that anymore. There will always people with other views.

The two things are not mutually exclusive.

Broken player expectations and AI squad management 

[spursfan]

6 hours ago, Daveincid said:

But it's terrible for squad building! A lot of those players rot in U21 teams in the premier league because no other club can afford them (Financially & reputation wise). On the other hand, a lot of nations develop way too slow. So the issue goes into both directions.

Thank you for sharing, it does provide perspective :-)

I'm wondering if reputation can be made more variable for players (larger decrease if unemployed / not playing), and for smaller nations/leagues (larger increase if they make 'marquee' signings) -- so there will be more likelihood of a 'reputation fit'?

Also perhaps an internal cap on players in particular positions (not sure if it would be a nightmare coding-wise). 

 

 

[Daveincid]

vor 8 Minuten schrieb spursfan:

Thank you for sharing, it does provide perspective :-)

I'm wondering if reputation can be made more variable for players (larger decrease if unemployed / not playing), and for smaller nations/leagues (larger increase if they make 'marquee' signings) -- so there will be more likelihood of a 'reputation fit'?

Also perhaps an internal cap on players in particular positions (not sure if it would be a nightmare coding-wise). 

 

 

All we can anyway is wait until FM25 is released. It won't take long to check if the issues are the same as before. But as I already mentioned, I'm optimistic that some points will be improved for sure

[talhak]

3 hours ago, srvngrc said:

I hope the new engine is optimised for Mac Silicon. FM 24 was running very hot on M1. FM 25 would be cooler I hope. 

To be honest, I do not have any issues at all with my M2.

[srvngrc]

1 saat önce, talhak said:

To be honest, I do not have any issues at all with my M2.

My M1 is going over 95 celsius with recommended graphic settings. 

[Dotsworthy]

11 hours ago, Daveincid said:

If you have 200 players with a CA140+ in 2023 but 500 in 2040, then there are several issues at once.

I think this might be the fundamental issue with youth development in the game. Even if they 'fixed' youth development a player will see an oversaturated market of +25 year olds in their scout reports and not be able to see any difference.

11 hours ago, Daveincid said:

So what should the devs do? Slow down development speed in the code to lower the amount of too many worldclass players and make the issue in the majority of all other nations even worse? Should they increase development speed so that we will end with 100 CA200 players to solve the issue with the lack of development in smaller leagues?

The difference between going five years into a single save or going through a single year in each iteration of FM is that the devs/researchers adjust CA/attributes to take into account serious injuries, players unable to maintain their form or keep up with tactical changes within the football world.

In a five year (or longer) save most players get better/plateau year on year until 26/27 and don't get worse unless they are over 30 or unattached for a long period. This was fine in FM21 when players over 30 dropped off a cliff, but since then you get big waves of over 25s that you have to wait to retire before the next gen can get first team football.

A solution could be a player regression mechanic where players lose CA depending on factors that could be used to keep the game balanced, but it would open up a can of worms and I'm not sure the forum could handle of volume of people complaining their players are getting worse 

Edited Monday at 19:22 by Dotsworthy

[talhak]

1 hour ago, srvngrc said:

My M1 is going over 95 celsius with recommended graphic settings. 

Do you see a prompt like this on the top right corner of the screen when you start FM?

[sanx]

i saw a interesting comment on reddit that said SI has put all their development time into the ME, hence why SI has removed features. Im aware it may not be true but im tired of the current ME now so im kinda desperate for a change lol. The ME is decent, SI has done a good job but its get tiresome after some months. 

[Dagenham_Dave]

1 hour ago, sanx said:

i saw a interesting comment on reddit that said SI has put all their development time into the ME

One of those classic 'in the know'* people on reddit, eh. 

 

 

*knows sod all. 

[janrzm]

Transfer Market Mechanics is something I desperately hope to see improved in FM25.

I fired up FM24 this last week for a refresher after several months and there are a two major issues (both of which I have raised way back).

The AI clubs sleeping in the period between the seasons end and next transfer window opening allows the human player to easily gather up all their preferred targets. Similarly you can do this during the season in readiness for the following season. Essentially if you have funds you are operating one window ahead of the AI. Player controlled clubs need to be saying we aren't sanctioning that transfer until we've let some people go more often. Players and agents need to be reluctant to sign too early, agents need to drum up interest and clubs need to be dismissive of offers coming in too early. Additionally., player reactions need to reflect that offers not being entertained mid season is fine in many cases and their morale should not tank accordingly. 

The second area of concern is activity in the January (mid season) transfer windows is crazy. In the save I've just played this last week my team was top of the league in January and clubs are in for my top 7-8 players who are all desperate to leave. You just do not see this IRL. If we've over achieved then by all means we should be under pressure in the summer but not half way through the season. The January window is typically made up of clubs looking to do some "housework" squad tidying, covering injuries, shifting unsettled and surplus players. Add into this a handful of bigger surprising deals, maybe but not a given. At the same time clubs should be refusing the human player approaches more, agents should be saying my player is keeping is options open more, or he's focused on the season (especially if his team is in with a chance of winning something). It needs toning down massively. 

There's no nuance to player decisions around transfers that appeal to them, there needs to be much more. A player desperate to leave my silverware winning team competing in Europe for one that's relegation fodder in a league ranked one or two places above mine shouldn't be a given.....

Transfer offers by AI clubs.......I think the balance is delicate here because the game could be ruined if AI teams over pay too easily. But, multiple offers at less than what you paid for players with great ratings, performance and potential are fatiguing. And give me the ability to say no means no.....and if I've turned down 3 million what makes you think I'll accept 2 million the day after......(sheds tear). 

Edited Wednesday at 10:12 by janrzm

[Dagenham_Dave]

How the AI acts in the transfer market is directly linked to how you set your game up in the beginning. SI should do a much better job of explaining how a balanced game world looks, maybe even offering the player some default options explaining the pros and cons of each. 

 

37 minutes ago, janrzm said:

The AI clubs sleeping in the period between the seasons end and next transfer window opening allows the human player to easily gather up all their preferred targets

I usually wait until the season ticks over before signing players, other than pre-contract ones I've agreed on months before. This prevents the above scenario. 

[spursfan]

3 hours ago, janrzm said:

Player controlled clubs need to be saying we aren't sanctioning that transfer until we've let some people go more often. Players and agents need to be reluctant to sign too early, agents need to drum up interest and clubs need to be dismissive of offers coming in too early.

Perhaps this can be done already by letting the DoF reject all offers (delegating responsibility is optional)?

3 hours ago, janrzm said:

The January window is typically made up of clubs looking to do some "housework" squad tidying, covering injuries, shifting unsettled and surplus players. Add into this a handful of bigger surprising deals, maybe but not a given.

Not sure if the AI managers currently work according to this, or mostly a reputation-based upgrade: just that somehow ending up with 6 Right backs while being threadbare thin elsewhere doesn't seem sensible 

 

[wazzaflow10]

13 hours ago, scythian12 said:

One thing that is also very prominent in FM24 squad building is CA homogeniety. Teams will recruit to the extent of their Rep and financial possibilities, but not only in the "tip" but also the breadth of the squad. This effect is helped by AI player decision-making prioritising being a less important part of a stronger squad than being a more prominent member of a "smaller" team. One ends up with AI squads which have 16-20 players in the same CA range within 10 points or so. DB (and arguably, real life) squads are more varied, with a CA range of about 20-30 for first team squad. On one hand, the current approach ensures a competent keeping of standards by the AI, but also leads to a bit of a stale gameworld - there will be little movement on the whole for statures of different clubs and divisions. I would argue it also exacerbates the suffocating grip of the PL on all things Rep and player quality. PL teams can easily attract the very best players in the World (think 140-50-60 CA for impact subs and fringe players), making any PL team the pantomime baddie in European competitions. 5/8 PL teams in the top 8 of the CL, them also easily making the quarter final of the CL yesr after year, England getting close to 30 points in the 5-year-ranking, a 40-point lead overall, I hate it. 🤮

This would have to do with how the game views CA and attributes though right? If CA (and by extension attributes) was treated as an unknown like PA that would give teams both human and the CPU a chance to get it wrong. That would introduce some variety in the squads without forcing some weird game mechanic on CPU managers. 

That said, the game could/should introduce some differing player interaction models that prevents teams from hording a lot of players. Players should be aware of the general squad ability and request playing time equal to them and match pathway promises to their ambition/professionalism. And if you have too many regular starters/important players/star players in a player's position they should be less interested in your team unless you're giving them a big wage packet or promising them more playing time than their positional rival (which should in theory make your current player(s) unsettled too). 

Its a lot of context to code and have it trigger correctly and appropriately.

[CrowManager]

53 minutes ago, wazzaflow10 said:

This would have to do with how the game views CA and attributes though right? If CA (and by extension attributes) was treated as an unknown like PA that would give teams both human and the CPU a chance to get it wrong. That would introduce some variety in the squads without forcing some weird game mechanic on CPU managers. 

That said, the game could/should introduce some differing player interaction models that prevents teams from hording a lot of players. Players should be aware of the general squad ability and request playing time equal to them and match pathway promises to their ambition/professionalism. And if you have too many regular starters/important players/star players in a player's position they should be less interested in your team unless you're giving them a big wage packet or promising them more playing time than their positional rival (which should in theory make your current player(s) unsettled too). 

Its a lot of context to code and have it trigger correctly and appropriately.

The issue is that would make signings arguably worse for the AI more than the player. 

A lot of the "issues with AI squadbuilding" in my view can be fixed pretty easily by taking away the player's ability to know exact attributes. Ergo, making them ranges with Mustermann or Starattributes. Players struggle to make bad transfers and then want to complain about the game being too easy when they can spend 100m on 2 elite players who will be great while the AI can't. Or wonderkids as people above mentioned.

That's not to say there aren't legitimate issues with AI squadbuilding of course.

Edited Tuesday at 02:16 by CrowManager

[scythian12]

4 hours ago, wazzaflow10 said:

This would have to do with how the game views CA and attributes though right? If CA (and by extension attributes) was treated as an unknown like PA that would give teams both human and the CPU a chance to get it wrong. That would introduce some variety in the squads without forcing some weird game mechanic on CPU managers. 

That said, the game could/should introduce some differing player interaction models that prevents teams from ho/a/rding a lot of players. Players should be aware of the general squad ability and request playing time equal to them and match pathway promises to their ambition/professionalism. And if you have too many regular starters/important players/star players in a player's position they should be less interested in your team unless you're giving them a big wage packet or promising them more playing time than their positional rival (which should in theory make your current player(s) unsettled too). 

Its a lot of context to code and have it trigger correctly and appropriately.

It is a very complicated bit of the game, with many feedback loops, so you can easily end up with horrific butterfly wing flapping-related accidents. One thing that jumps out at me is that the players will only look at the club's squads that are interested in signing them and more often than not choose the stronger squad, even if their role is somewhat minor to what it would be at other interested parties. I do not believe (but then again, I could be very wrong, see first sentence) that a small tweak in how players prioritize playing time over quality teammates would ruin the entire system, but lead to bit more varied squads. It is also evident that with the current system, England starts attracting the very best players very easily. On the 'tip' end of the scale, this also leads to a phenomenon which I feel this is also a bit of a limp as England might have the most competitive league, but in the highest echelons of quality, it never is / was as prolific in attracting the best players like the Spanish or earlier the Italian top clubs were. Just throw a glance at the Ballon d'Or wiki page for the top 3 nominees in the last 40 years. On the player development front, maybe playing 38 league games against very tough physical opponents is simply not the best approach to reach those elusive 170+ CA levels. But of course, the majority of the buying public is English so it will always skew towards their desires, and has been a long-standing issue with the gameworld as whole. Move out of London, SI!

But playing with an approach where I switch clubs every 5 years in the same nation I see that the AI will sell the very best players I attracted to my previous clubs (or even let them go... ), but it will cultivate a squad that has 16 players in a CA range within 10 points of the 3rd-4th best players when I left the club, very occasionally punching higher, and this now over more than a 20-year span. They even have 1-2 club grown players who are in this first squad range, ones they signed as talents with 18 and developed them, so I see less of an issue with the development side of things.

 

Edited Tuesday at 05:45 by scythian12

[srvngrc]

10 saat önce, talhak said:

Do you see a prompt like this on the top right corner of the screen when you start FM?

No I don't.

[CrowManager]

19 minutes ago, scythian12 said:

It is a very complicated bit of the game, with many feedback loops, so you can easily end up with horrific butterfly wing flapping-related accidents. One thing that jumps out at me is that the players will only look at the club's squads that are interested in signing them and more often than not choose the stronger squad, even if their role is somewhat minor to what it would be at other interested parties. I do not believe (but then again, I could be very wrong, see first sentence) that a small tweak in how players prioritize playing time over quality teammates would ruin the entire system, but lead to bit more varied squads. It is also evident that with the current system, England starts attracting the very best players very easily. On the 'tip' end of the scale, this also leads to a phenomenon which I feel this is also a bit of a limp as England might have the most competitive league, but in the highest echelons of quality, it never is / was as prolific in attracting the best players like the Spanish or earlier the Italian top clubs were. Just throw a glance at the Ballon d'Or wiki page for the top 3 nominees in the last 40 years. On the player development front, maybe playing 38 league games against very tough physical opponents is simply not the best approach to reach those elusive 170+ CA levels. But of course, the majority of the buying public is English so it will always skew towards their desires, and has been a long-standing issue with the gameworld as whole. Move out of London, SI!

But playing with an approach where I switch clubs every 5 years in the same nation I see that the AI will sell the very best players I attracted to my previous clubs (or even let them go... ), but it will cultivate a squad that has 16 players in a CA range within 10 points of the 3rd-4th best players when I left the club, very occasionally punching higher, and this now over more than a 20-year span. They even have 1-2 club grown players who are in this first squad range, ones they signed as talents with 18 and developed them, so I see less of an issue with the development side of things.

 

That a way of saying that if one were to edit the top 5 CA of players at a club, you believe they'll try to be more aggressive towards higher level talent in the future?

[spursfan]

17 hours ago, Bahoi said:

But how do we address the issue of AI teams not making enough transfers or letting players go and so on. For example, in a save I have in 2027, top 4 teams of the top 5 leagues in Europe (Bayern, Real, PSG, City etc etc, but also almost any EPL team) are using  almost the exact same starting 11 they used since the summer of 2023?

Are these 5 leagues all playable? Number of playable leagues might be a factor to explain 'lack of transfer activity'. 

I'd done a minor nation in previous FMs (mistakenly it was the only league playable), the frustration is  common I suspect -- not enough reputation to attract better players.

 

57 minutes ago, scythian12 said:

But playing with an approach where I switch clubs every 5 years in the same nation I see that the AI will sell the very best players I attracted to my previous clubs (or even let them go... ), but it will cultivate a squad that has 16 players in a CA range within 10 points of the 3rd-4th best players when I left the club, very occasionally punching higher, and this now over more than a 20-year span. They even have 1-2 club grown players who are in this first squad range, ones they signed as talents with 18 and developed them, so I see less of an issue with the development side of things.

So once you've left, the AI would try to sell for a profit, and use the proceeds to continue letting things tick along a certain level, and not really actively seek to boost it (if I'm reading this accurately)? 

 

[Bahoi]

9 minutes ago, spursfan said:

Are these 5 leagues all playable? Number of playable leagues might be a factor to explain 'lack of transfer activity'. 

All playable, all matches played in full detail ! The database is set up with 200.000+ players with various leagues in Europe, South/North America or Asia being playable. This is why I am in 2027 only, I advance at most one matchday at 2-3 days interval, when time and work allows. I want realism, I get Douglas Luiz, McGinn, Kamara for Villa three-man midfield three years on. 

[spursfan]

36 minutes ago, Bahoi said:

All playable, all matches played in full detail ! The database is set up with 200.000+ players with various leagues in Europe, South/North America or Asia being playable. This is why I am in 2027 only, I advance at most one matchday at 2-3 days interval, when time and work allows. I want realism, I get Douglas Luiz, McGinn, Kamara for Villa three-man midfield three years on. 

Nice, would be a decent rig you've got there ;-)

Also, Z released a video featuring some database tool that summarises player attributes into a single number.
Wonder if FM25 would be headed in this direction (fewer detailed attributes, more summarised banded ranges)? 

Edited Tuesday at 07:36 by spursfan
Additional point forgotten to add on (rather than making a new post)

[Costav]

15 minutes ago, spursfan said:

Also, Z released a video featuring some database tool that summarises player attributes into a single number.
Wonder if FM25 would be headed in this direction (fewer detailed attributes, more summarised banded ranges)? 

It exists, and it's called FIFA (or eFootball,or whatever is called now-a-day).

One of the greatest things of FM is that you don't have an overall, just because each role and each situation requires different attributes.
The evaluation stars are already something that is misleading, please let's not add more fluff to the game.

[spursfan]

2 minutes ago, Costav said:

it's called FIFA (or eFootball,or whatever

Yeah I get that -- might move towards that I suspect, but hopefully not as much as what Z's video shows.

3 minutes ago, Costav said:

The evaluation stars are already something that is misleading

I actually like the stars, gives a little bit more variability than a hard number (which is too exact compared to real life for me).
And not as 'fuzzy' as going 'attributeless' (I would probably do so only when the data generated actually makes proper sense) 

[Costav]

5 minutes ago, spursfan said:

I actually like the stars, gives a little bit more variability than a hard number

I agree on that, althought I defined them misleading because of the weight that the attributes have on the CA (as an example, physical attributes are weighted so much more that some mental attributes). This leads to evaluation that most of the time are biased by physical characteristics of the player.

 

7 minutes ago, spursfan said:

And not as 'fuzzy' as going 'attributeless'

I play without attributes, basing my judgment only on the diamond of attributes and on reports....it is fantastic!
I finally enjoy playing FM because it became too easy and now I can even make bad signings 

[jmlima]

1 hour ago, spursfan said:

Nice, would be a decent rig you've got there ;-)

Also, Z released a video featuring some database tool that summarises player attributes into a single number.
Wonder if FM25 would be headed in this direction (fewer detailed attributes, more summarised banded ranges)? 

That would be sort of like the 20-80 system used in baseball scouting.

[spursfan]

4 hours ago, Costav said:

I agree on that, althought I defined them misleading because of the weight that the attributes have on the CA (as an example, physical attributes are weighted so much more that some mental attributes). This leads to evaluation that most of the time are biased by physical characteristics of the player.

I see, I will look at the stars, but I'd place more emphasis on the other attributes (not only physical) -- I'm not that confident it looking at the diamond charts, and I've been known to sign awfully even based on attributes lol

[Lord Rowell]

I also would be in favour of keeping the stars-system, or something similar, that includes an element of uncertainty and inaccuracy.

At the end of the day that's part of sport, no-one can be absolutely certain about how good a player will be.

It also leaves scope for the individual player to make their own judgement based on attributes and gives the possibility of finding bargains that can be developed, if you have the right idea of what to look for.

Edited Tuesday at 13:29 by Lord Rowell

[ForTheLoveOfTheGame]

Still no news today…

Another delay to the delayed announcement almost a certainty then. That journalist knows his stuff after all.

[XaW]

Just now, ForTheLoveOfTheGame said:

Another delay to the delayed announcement almost a certainty then. That journalist knows his stuff after all.

Considering SI have never give an actual date, I find it hard to point to a lack of announcement currently is proof of anything...

[kiwityke1983]

26 minutes ago, ForTheLoveOfTheGame said:

Still no news today…

Another delay to the delayed announcement almost a certainty then. That journalist knows his stuff after all.

Some speculation that they're introducing in possession and out of possession roles.

If true pretty exciting.

[spursfan]

1 hour ago, ForTheLoveOfTheGame said:

Still no news today…

Another delay to the delayed announcement almost a certainty then. That journalist knows his stuff after all.

Could be late night Pacific time to be fair 

[DavutOzkan]

1 hour ago, ForTheLoveOfTheGame said:

Still no news today…

Another delay to the delayed announcement almost a certainty then. That journalist knows his stuff after all.

The leak said last week September 30 (it was originally supposed to be today or yesterday, I believe). 

[grade]

1 hour ago, DavutOzkan said:

The leak said last week September 30 (it was originally supposed to be today or yesterday, I believe). 

The last blog post said the Late September. So... it is this week. We have for this week 3 days and a few hours to go, plus 2 days over the weekend and a Monday. People relax...

 

[decapitated]

 

Sort It Out SI have apparently heard this rumour. I think if true, this would be an interesting innovation that if it works as intended would be a tactical upgrade on FM24. It's just a rumour at this point but I'm hopeful.