I simulated to 2039 on FM23 - my observations

[Andrew James]

@noobcakethanks a lot for your analysis and detailed post

2 hours ago, noobcake said:

They are down ~12.5 points of CA at 25 years old compared to players at game start

Could you let us know what league setup you used for this test? This kind of drop-off would usually indicate a decent proportion of the players playing in inactive leagues. Some decrease in inactive nations can be expected, as the higher CA players are prioritised for inclusion at game start, and we don't boost newgens to keep this constant as it would be too easy to pick up high quality youth players. 

Internally, we're seeing CA rates remain very stable at Active clubs - and as has been mentioned above in thread, other factors beyond progression feed into this, such as AI transfers and team selection. 

We're still looking into ways we can balance out the progression rates in inactive leagues, and whilst I'm glad to see more evidence that attribution distribution isn't "broken" anymore, it's still an area we're not fully satisfied with and are looking to improve further. 

[Slippy_]

14 minutes ago, Andrew James said:

@noobcakethanks a lot for your analysis and detailed post

Could you let us know what league setup you used for this test? This kind of drop-off would usually indicate a decent proportion of the players playing in inactive leagues. Some decrease in inactive nations can be expected, as the higher CA players are prioritised for inclusion at game start, and we don't boost newgens to keep this constant as it would be too easy to pick up high quality youth players. 

Internally, we're seeing CA rates remain very stable at Active clubs - and as has been mentioned above in thread, other factors beyond progression feed into this, such as AI transfers and team selection. 

We're still looking into ways we can balance out the progression rates in inactive leagues, and whilst I'm glad to see more evidence that attribution distribution isn't "broken" anymore, it's still an area we're not fully satisfied with and are looking to improve further. 

Hi Andrew,

Not the person who posted the analysis but noobcake did add the save to his orginal post. I had a poke around to check the average age of squads in the English prem and it looked to average out around 27/28. That is perhaps a little high but not unreasonable (especially if we consider more players will probably continue to play at higher level for longer).

I have downloaded and checked the active leagues but probably best for noobcake to confirm:

[SupraFlush]

Here we go, I've finally simulated @LeoFM save until 2039. I haven't looked at the save yet, I just got home and need to eat something lol! But here's the save, 1st of January 2039 with the hotfix from 25st of October. Here's the link for anyone who either want to study how the development has changed or just want to play in the future, enjoy!

 

https://www.mediafire.com/file/y31ewgnjgjf8a7l/FM23_Player_Development_Test_%28Oct_25_Minor_Update%29_1st_January_2039_.fm/file

EDIT: For anyone interested, here's the leagues loaded: 

Edited October 27, 2022 by SupraFlush

[noobcake]

1 hour ago, Slippy_ said:

Hi Andrew,

Not the person who posted the analysis but noobcake did add the save to his orginal post. I had a poke around to check the average age of squads in the English prem and it looked to average out around 27/28. That is perhaps a little high but not unreasonable (especially if we consider more players will probably continue to play at higher level for longer).

I have downloaded and checked the active leagues but probably best for noobcake to confirm:

This is correct, the only things missing are the inclusion of players in continental competitions from Europe and South America via load options.

Out of these I only used players of >=130 PA to weed out players that don't seem to relevant for long-term development

EDIT: Gonna take an hour or two but I'm gonna filter out all players in non-selected leagues and post the finding with only the relevant ones

Edited October 27, 2022 by noobcake

[LeoFM]

@noobcake how did you get access to hidden attributes in game since the editor is not out yet?

[JamieL9898]

42 minutes ago, LeoFM said:

@noobcake how did you get access to hidden attributes in game since the editor is not out yet?

FMRTE 23 is out which can be used

[LeoFM]

Some players in 2039 before vs after the update:

Camavinga

Spoiler

Musiala

Spoiler

Bellingham

Spoiler

Gavi

Spoiler

Endrick

Spoiler

Some pretty clear improvements here! (obviously their CA might not be the same in both pictures but still looks good)

Edited October 27, 2022 by LeoFM

[LeoFM]

Here's my analysis of the updated attribute development:
https://docs.google.com/spreadsheets/d/1jAQg-r33xpNp8GCzl_CLT79_ckysI2Kh-2Z9wPCi5YA/edit?usp=sharing

Kept it very basic since I think @noobcake already covered most stuff.
This is basically just the same thing as I did in the beginning of this thread, I exported all attributes of all top players and then averaged them and compared different in game years.

Summary:
It's a lot better, but I still think there are improvements to be made (glad SI also share this view). 

[Slippy_]

Completly anecdotal but after messing around in the 2 saves supplied but it feels that the regen players are a little less technically gifted than say 'current' players but they seem to have better physicals/mentals. Of course SI have said that they are still looking with the code to improve further.

Although it looks like LeoFm came to the same summary. Glad to see it looks alright, looking forward to the full release!

[GreenTriangle]

A comparison between the database from 2022 (130000 players) and the one from the save posted by noobcake (year 2050, 64000 players) : link

[SerCska10]

3 часа назад, Andrew James сказал:

@noobcakethanks a lot for your analysis and detailed post

Could you let us know what league setup you used for this test? This kind of drop-off would usually indicate a decent proportion of the players playing in inactive leagues. Some decrease in inactive nations can be expected, as the higher CA players are prioritised for inclusion at game start, and we don't boost newgens to keep this constant as it would be too easy to pick up high quality youth players. 

Internally, we're seeing CA rates remain very stable at Active clubs - and as has been mentioned above in thread, other factors beyond progression feed into this, such as AI transfers and team selection. 

We're still looking into ways we can balance out the progression rates in inactive leagues, and whilst I'm glad to see more evidence that attribution distribution isn't "broken" anymore, it's still an area we're not fully satisfied with and are looking to improve further. 

And the top 5 leagues in Europe with their second leagues are enough for the normal development of newgens?

And please tell me which files and how in the game files you can change yourself to bring it to the state after update 23.0.5?

Edited October 27, 2022 by SerCska10

[Mars_Blackmon]

6 hours ago, diLLa88 said:

I responded in a different thread which seems related to this.

CA development is directly related with first team playtime. AI managers and AI squad building is highly favoured towards older players who have a high CA and that does not drop off as much as it used to do in older FM versions. Therefore young talent do not get enough exposure to playtime which significantly hampers their development and also reputation. Young talents are often signed and then simply not being used -> therefore no CA development.

This is also the reason why national teams are full with old players that are just not replaced by young talent as often happens in real life.

In my opinion old players should fall-off quicker whenever certain mental/physical attributes are not up to scratch (think: determination, work rate, natural fitness, etc.). Also AI teams should use young talents more in easier/cup games or as a substitute for the sake of development, this is simply not happening enough.

Young players need more ways to gain reputation. There is barely any international competitions for youth. Also old player shouldn’t drop off a cliff, the problem is that the game has a lack of injuries and therefore the old players are always fit.

[b.sinisa]

3 minutes ago, Mars_Blackmon said:

Young players need more ways to gain reputation. There is barely any international competitions for youth. Also old player shouldn’t drop off a cliff, the problem is that the game has a lack of injuries and therefore the old players are always fit.

A lot of people report a lot of injuries i don't remember a lack of injuries 

[noobcake]

Ok so here's an update as per the suggestion of @Andrew James

I filtered my results for players >=130PA in loaded leagues only (~5k vs ~8k before) and here it is

Loaded league players do develop better, but I'd hesistate to call it well, as they're still on average 6.6 points behind at 25 vs 12.6 with non-loaded. The worst is actually -10.4 at 21, whereas with non-loaded league players it was -13.94 at 26

So I guess if you want to play despite this at least load leagues as opposed to loading players though the custom database setup. It reduces the impact of the problem

Players transferring in from unloaded leagues might still be affecting this but if you're running a large DB this is likely to happen to some extent anyway. Reducing the DB size and loading more leagues instead might make the issue even smaller. Just don't expect as many cool super exotic wonderkids if you don't load the league I guess

Loaded leagues only spreadsheet

Edited October 27, 2022 by noobcake

[Mars_Blackmon]

8 minutes ago, b.sinisa said:

A lot of people report a lot of injuries i don't remember a lack of injuries 

Injuries are only at a 80% rate of the real life rate. Not to mention most users play with high intensity tactics. But of course we aren’t talking about the users here, we are talking about the overall game world where the AI might not be using high intensity tactics and may see minimal injuries especially to veteran players who will not sit out and make room for younger players.

 

Not to mention the lack of intentional play for development and lack of friendlies where some 1st team players might miss a game or two and start younger players.

Edited October 27, 2022 by Mars_Blackmon

[Dotsworthy]

I'm pretty sure SI tweak injuries to about 80% of real life injuries.

[b.sinisa]

2 minutes ago, Mars_Blackmon said:

Injuries are only at a 80% rate of the real life rate. Not to mention most users play with high intensity tactics. But of course we aren’t talking about the users here, we are talking about the overall game world where the AI might not be using high intensity tactics and may see minimal injuries especially to veteran players who will not sit out and make room for younger players.

Yes it is possible that in the player profiles are selected options like play till you can and some other options, there is option for some other stuff

[Arnar]

4 hours ago, LeoFM said:

Some players in 2039 before vs after the update:

Camavinga

  Hide contents

Musiala

  Hide contents

Bellingham

  Reveal hidden contents

Gavi

  Reveal hidden contents

Endrick

  Reveal hidden contents

Some pretty clear improvements here! (obviously their CA might not be the same in both pictures but still looks good)

This looks a little better after the update but am i the only one who thinks they are unreasonbly high in crossing, corners  and penalties vs other technicals?

[Dotsworthy]

13 hours ago, noobcake said:

Loaded league players do develop better, but I'd hesistate to call it well, as they're still on average 6.6 points behind at 25 vs 12.6 with non-loaded. The worst is actually -10.4 at 21, whereas with non-loaded league players it was -13.94 at 26

Are view only leagues identical to playable in this regard? If you were to load a view only league and the players for it, would there be a significant difference in attribute growth compared to loading the playable league instead?

Edited October 28, 2022 by Dotsworthy
More clarity

[fc.cadoni]

9 hours ago, Dotsworthy said:

Are view only leagues identical to playable?

No

[abcdf]

16 ore fa, noobcake ha scritto:

Ok so here's an update as per the suggestion of @Andrew James

I filtered my results for players >=130PA in loaded leagues only (~5k vs ~8k before) and here it is

Loaded league players do develop better, but I'd hesistate to call it well, as they're still on average 6.6 points behind at 25 vs 12.6 with non-loaded. The worst is actually -10.4 at 21, whereas with non-loaded league players it was -13.94 at 26

So I guess if you want to play despite this at least load leagues as opposed to loading players though the custom database setup. It reduces the impact of the problem

Players transferring in from unloaded leagues might still be affecting this but if you're running a large DB this is likely to happen to some extent anyway. Reducing the DB size and loading more leagues instead might make the issue even smaller. Just don't expect as many cool super exotic wonderkids if you don't load the league I guess

Loaded leagues only spreadsheet

As I said, I still see technicals lagging a bit behing mentals and physical in players development. Can we expect to see further balancing on the issue @Andrew James?

[Proffa]

Will there be more effort put in to this matter or is it case closed for now? 

[Andrew James]

On 28/10/2022 at 12:57, abcdf said:

As I said, I still see technicals lagging a bit behing mentals and physical in players development. Can we expect to see further balancing on the issue @Andrew James?

2 minutes ago, Proffa said:

Will there be more effort put in to this matter or is it case closed for now? 

Definitely something we're looking to improve further. As I think has been mentioned earlier in thread, a lot of the remaining issues have been present in previous versions and are unfortunately not simple to resolve, but it's certainly not case closed. 

[quadrifoglio]

2 hours ago, Andrew James said:

Definitely something we're looking to improve further. As I think has been mentioned earlier in thread, a lot of the remaining issues have been present in previous versions and are unfortunately not simple to resolve, but it's certainly not case closed. 

If you can resolve not only this year's problem but remaining issues such as how early a wonderkid steps up to the 1st team , which are the development years for players and the decline of development for over 30 players then we are in a good road

[Stevedox]

On 31/10/2022 at 13:45, Andrew James said:

Definitely something we're looking to improve further. As I think has been mentioned earlier in thread, a lot of the remaining issues have been present in previous versions and are unfortunately not simple to resolve, but it's certainly not case closed. 

I don't frequent the forums often (usually around each release) but have been around a while, and I must say this is a very different attitude of old, where any fan concerns were dismissed and even ridiculed by mods and creators. So it's been nice to see you engaged throughout. 

[LeoFM]

Improved Attribute distribution – NewGen generation + progression (eg. Crossing for fullbacks)

Has anyone tested this yet? Will be very interesting to see the results!