[FM21] [Experiment] Youth intake quality and quantity in Spain with different database and varied amount of playable leagues loaded

[Orion_]

1. Introduction

Hello everyone. Today we're going to take a look at how different database setup in terms of amount of loaded divisions and usage of custom database within one country influences number of youngster generated and we will measure their quality based solely on their potential ability attribute (PA). The idea for that experiment came from my fond for playing mostly with lower reputation leagues and clubs with my teams build around youngsters development. For some time I wanted to try this play style with bigger club so obvious candidate became Athletic Bilbao with their Basque-only policy. I asked myself a question how can I maximize my chances for getting promising youngsters in terms of game setup.

2. Database setup cases and simulation conditions

First of all we will look at LaLiga only to have a rather hermetic environment without interference of other big leagues at youth intakes. Just to be clear we're using current at date of writing 21.2.2 version of FM21 with 21.1.1 default database and as database extension we will use database update for Spain from Sangue Blu available in steam workshop as linked. We will use for this test 4 cases of database setup:

• Default FM21 database for Spain with LaLiga (top division) loaded only
• Custom database for Spain with LaLiga loaded only
• Default database for Spain with all available divisions loaded (4 divisions)
• Custom database for Spain with all available divisions loaded (8 divisions)

Then we determined simulation conditions:

• Upon starting the game 'Add players to playable teams' was ticked ON
• 'Detail level' was set to 'None' for all competitions either national or international - it was done to speed up simulation. We have no interest in full match simulations. They don't add anything to our experiment
• Starting date was set as soon as possible for both databases
• Jobless manager was added and immediately sent to holiday until 31/12/2025

Things that can distort data:

• It's possible that some clubs may have different facilities rating between two databases
• Clubs outside Spain are still producing newgens but we're hoping that their amount will be negligible in comparison to Spanish youth intakes

How we determine result:

• When the game reaches 31/12/2025 the game is loaded into FM Genie Scout 21 v1.0 21.2.2 beta 10 build 1110
• We set filter for players age 0-20 and 'Is Newgens' parameter
• We count the players with certain amount of potential ability to show the distribution

The age limit is expected to filter out players that were generated at the start of the game to fill all playable teams, including a lot of U-19 teams, and leave us with only youth intake players. The number was set by checking that youngest newgen at the start of a game was born in 2004 so we're expecting that majority of youth intake players will be younger that that.

3. Expectations

I expected the distribution to be shaped like Gaussian distribution varying only in it's 'height' and changing the position of median closer to 0 if we load more leagues and closer to 200 if we load LaLiga only. My expectations came with assumption that in general good teams with better facilities should produce more higher quality players and lower league clubs should produce more low PA players.

It's worth mentioning to explain why number of playable leagues is so important factor in this case. From my understanding leagues set as 'Playable' upon game start setup will produce full youth intake consisting of 16 youngster every single season no matter their quality. For the leagues set as 'view only' the game will produce limited amount of youth players from intake limiting them to most promising ones. This also applies to leagues that are not loaded at all in the game. With this knowledge we would expect that number of most promising youngsters should be similar across all database setups. The question is can the game produce more high quality youngsters just because it's going to produce much higher amount of all youngsters. But does the quality follows the quantity?

4. Results and analysis

4.1 Newgens quantity

At first we will take a look at number of newgens generated in the different cases.

As we can see default and extended database with only LaLiga loaded produces similar amount of players.

As expected the bigger the number of playable clubs the more players are generated with astonishing number of 17551 youngsters generated in custom database with all divisions loaded.

4.2 Newgens quality

And now the most important graph. What is the distribution of PA among all filtered newgens. There will be show 'raw' graph with resolution of full 200 PA points. There are also prepared graphs with higher intervals of 5's, 10's and 50's to 'smoother' the graph since few points of difference in PA shouldn't be significant. For example players with 151-160 PA are equality good and it doesn't matter if player has 151 PA or 152 PA.

As we can see the first interesting thing that is visible on the graph is enormous amount of players with very low PA in extended database with all divisions loaded. It can be explained with very big number of teams with very poor facilities/reputation/coaches which affects the youth intake. 80% of players generated in this setup had less or equal than 100PA which is far too low to be considered LaLiga senior squad player.

Another interesting thing is that both setups with only top division loaded produced very similar results so database with potentially more clubs won't produce more teams with the same amount of playable leagues loaded. To be honest I didn't check if default database has the same amount of teams but they are only not playable or custom database adds more teams into the game. It's worth noticing that both setups had similar peaks around 70 PA and then around 125 PA. I'm not sure where this gap between 80 and 105 PA came from. If you have any ideas I'll gladly look through them.

When it comes to default database with all leagues loaded it produced almost normal distribution of PA which looks as expected.

All setups produced similar number of world class players (150+ PA) so if we're interested in obtaining only most promising youngsters we can even play with only top division loaded.

5. Conclusion and free thoughts

If we're planning on playing with very top clubs we won't get much from loading more than 2-3 top divisions for country where we play since amount of world class players and players that will be able to fit into senior squads for top division clubs will stay the same.

Another point it that with less playable divisions comes less playable clubs. The difference between 'view only' and 'Playable' league is that if you not choose otherwise in 'Detail level' on default all matches within playable leagues will be fully simulated while for 'view only' the matches will resolve matches based more on clubs and players reputation and overall skill.

We still can set detail level for playable lower divisions to 'None' so they will be simulated in quick match engine but it's still more power consuming than 'view only'.

Last but not least it's worth noticing that if we're planning to play with hidden attributes we could be flooded with low potential players if we search players by ourselves through scouting centre resulting in higher difficulty or finding quality players or it can take more effort and time in compare to smaller database where game will just not produce players with very low potential or their number will be significantly limited.

At the end I'd like to share my point of view and a little suggestion for players that like me try develop youngsters. I personally would check beforehand in what divisions is team of my choose B and C team. Then load the game with playable divisions low enough for B and C team to be in them. If you play with major club your reserve teams should get quick back to back promotions in compare to other regular clubs in their respective divisions and when your reserves reaches highest division possible for them (it's regulated by league rules) I'd just remove lower leagues. This way you can provide more quality playtime for your youngsters. Also remember to set all lower than your first team divisions 'Detail level' to 'None' to improve game speed.

 

If you reached this moment I want to thank you for reading. The spreadsheet with data collected is available here. If you have any questions feel free to ask. I'll try to keep an eye on topic and replay as soon as I can.

Also very minor thing related to data collection. Does anyone of you have Genie Scout with premium access and can check if data export to CSV allow to export CA/PA attribute? Because at my free version you can export current players list to csv but CA/PA column is not there and I think it's made on purpose. So if premium version had access to CA/PA export I could collect more data faster next time for such a test.

 

Cheers everyone and have a nice day!

Edited February 8, 2021 by Orion_