The best leagues to load in Football Manager for Newgens [Data Based]

JohnConnor

Member
Joined
Nov 16, 2020
Messages
20
Reaction score
27
Points
13
See the full article here

I have sampled each and every playable club in the game for their real median PA values. This is the best indicator of newgen quality. I explain why in the article, but basically the PA of the best player in the intake is much more random than the median PA, which is quite stable and is a good indicator of expected quality.

I have then weighted the quality of each division with the processing time for the division, to create a ranking of what leagues should be loaded and which aren't worthwhile. You can read my earlier article which explores processing time here.

A few examples of my findings:

- South Korea is one of the best leagues to load
- Colombia has 17 clubs with 100+ median PA (for comparison, Greece only has 2)
- Argentina Div 2 isn't worth it due to the high number of mediocre clubs in a burgeoning division (worse than South Africa when weighted)
- Chile Div 1 is even worse than Argentina Div 2
- France Div 2 is better than many other options, including the first divisions of Belgium, Portugal, Denmark and Austria.
- Romania has a 130 PA club (FC Viitorul), but the rest of the division was so mediocre that it makes Romania Div 1 not worth selecting overall

In the future I will be releasing the full data (i.e. exact median PA values I recorded for each individual club) if there is enough interest

All feedback, questions and suggestions welcome
 
Great analysis. I observed in FM19 that top teams buy many South Korean newgens as the game progesses. An the SK national team benefit directly from those newgens as they dominate Asia and constantly stay in top 10 World Ranking.
 
Excellent article. Changed the way I set my games up.
 
Cheers lads, stay tuned for something even better soon. Don't want to pump it up too much because I'm still working on it, but basically I've reversed engineered the actual newgen formula, and it includes one factor you might not be expecting. Right now I'm getting results with ~95% accuracy (predicts the median PA). I'm trying to work out if the difference is just random variation, or if I've not identified a factor, or if I can nonetheless tune my numbers to get higher accuracy.

I observed in FM19 that top teams buy many South Korean newgens as the game progesses

Another reason I like South Korea (don't know if it's same in future FM versions) is that its one of the countries that gives bonus mechanizing revenue too. A curious fact is also that one team in South Korea Div 1 consists purely of loaned players - it's a team that players play for when they do mandatory military service.
 
@fc.cadoni

Thanks

Unfortunately the formula I came up with turned out not to be as effective as it initially seemed to be. It works for getting a rough idea, but it's only about as accurate as a single sample of the median PA data.

My latest article is Proving 'Youth Facilities' has no effect on newgen PA in Football Manager.
In addition to what the title says, in it you'll find some of the raw data I've gathered (for your own assessment), as well as what I've found does or does not contribute to newgen PA.
 
Last edited:
@fc.cadoni

Thanks

Unfortunately the formula I came up with turned out not to be as effective as it initially seemed to be. It works for getting a rough idea, but it's only about as accurate as a single sample of the median PA data.

My latest article is Proving 'Youth Facilities' has no effect on newgen PA in Football Manager.
In addition to what the title says, in it you'll find some of the raw data I've gathered (for your own assessment), as well as what I've found does or does not contribute to newgen PA.
Just a heads up, seems the link you posted doesn't actually re-direct to any URL. I was able to navigate to the article via the links you posted in the OP, though.
 
What's next John?

Just sent you a message, but in case someone else wants to know, I plan on making posts on these two things soon:

Training
Set of steps to creating a near-perfect custom DB (hours long process, but gives better result than default game)

I can put out some more stuff on PA (i.e. more specific data for one thing), and a collection of little interesting tidbits.. for instance, have you ever wondered where the permanant 'scouting knowledge' of a scout is located in the editor? It's unintuitively in 'career plans' under 'person data' (would be a major tip, if it weren't for the fact that certain 'career plans' aren't searchable in the editor).. but I don't really feel like posting more on that stuff right now
 
Just sent you a message, but in case someone else wants to know, I plan on making posts on these two things soon:

Training
Set of steps to creating a near-perfect custom DB (hours long process, but gives better result than default game)

I can put out some more stuff on PA (i.e. more specific data for one thing), and a collection of little interesting tidbits.. for instance, have you ever wondered where the permanant 'scouting knowledge' of a scout is located in the editor? It's unintuitively in 'career plans' under 'person data' (would be a major tip, if it weren't for the fact that certain 'career plans' aren't searchable in the editor).. but I don't really feel like posting more on that stuff right now
Ever done any kind of testing on which attributes tend to be most important for players to have? It's difficult to isolate the variable, especially when attributes presumably matter differently across positions, and even more so across roles. But for example, I tend to value passing very highly on all my midfielders, and dribbling + pace on strikers. I think people overvalue finishing and long shots, too -- sure, you wouldn't want a striker with a terrible finishing rating, but finishing seems to suffer from diminishing returns beyond a certain point, as it becomes more beneficial for the striker to create more/better chances than to better finish the chances he is already capable of creating.
 
Not commenting on your research or anything

But Why load leagues when you can force load players? Loading leagues slows down sim time and is only useful if you actually want to manage in the league. Or if you want more than the top 2 teams in that league to be ultra competitive.

Plus you are missing out on all the good Africans and Asian-Middle East players that don't have any leagues that could be loaded.

Again great research but you should be applying the research to loading players by league, nation, region and continent. Then seeing what combination of players loaded is the best.

I know in my saves I force load 100% every continent except Europe. I do all tier 1 and select tier 2 leagues of the big nations. Also force load 100% starting nation. This gets me about 200k to 250k players. But it could certainly be trimmed.

While loading leagues kills sim time on potato computers and most FM users have potato laptops.
 
Ever done any kind of testing on which attributes tend to be most important for players to have? It's difficult to isolate the variable, especially when attributes presumably matter differently across positions, and even more so across roles.

This is something that interests me, that as you say, is difficult to isolate. My theory on training will identify key attributes that are beneficial across all players, and the degree (roughly) to which they are beneficial (for instance, 'professionalism' is more important than 'penalties'). For the key attributes for a role, I just take for granted what SI recommends for the role, but also in combination with more vague explanations of attributes given in the past by the developers. Still compiling the training post, but will post it soon.

Trying to work out what attributes do precisely is something completely in the future for me at the moment. I reckon the best way of approaching it would be to look at actual results, like goals scored, chances created, passes completed, and so on. Came across this recently, which I found interesting:

EkVmywTXcAAbGwy.png

Source: twitter.com/StatSav (he's got a few more charts on there)

Took a look at some of the best vs. some of the worst players, and it seemed like 'strength' was one attribute that mattered for strikers (was just glancing though). I think this is probably a decent way of deducing what attributes actually matter.

But Why load leagues when you can force load players? Loading leagues slows down sim time and is only useful if you actually want to manage in the league.

It's a fair point that loading leagues for newgens may be unnecessary. I haven't examined it, but I'm pretty sure that not loading leagues does at least reduce the quantity of newgens significantly. From non-loaded teams I've looked at, the median PA (and presumably the spread) does stay the same though. I might do a quick test to see if manually loading all players also increases the number of newgens produced.

Personally I like to load all the key leagues I reasonably can, for immersion. I do actually have a custom DB made where I've stripped out all the unnecessary players and some of the staff, and I plan on posting the steps to creating this sometime after a post on training. My idea initially was that it would lead to a big performance improvement, but it ended up only having a minor performance improvement over default DB of the same number of players. However, it does distribute players much better (78k is what I could reasonably get it down to) and you get ~90k more staff at zero performance cost.

Note that loading more players still has a significant performance impact, even if leagues are not loaded. It also bloats the save size.

In regards to performance, unfortunately even with a cutting edge CPU it's hard to recommend loading beyond the usual suspects (i.e. spain, germany, brazil, etc.).. so in the end it might not even matter for a lot of people, because whether it's Serbia or Czech Republic that's better, either is probably a bit too much. Personally Serbia Div 1 is about the limit to what would I load down to for an enjoyable save.
 
Loading players doesn't really start to slow anything down until you are like 15+ years out in save. The average FM user doesn't take it that far. But loading full playable leagues on certain potato computers can slow it down from day one.

The player database works like this. Whatever your number is lets say its 100k. The game will always try to have 100k players any more then that your get the fake "Grey" players and anything less then that you get more newgens until you hit your 100k number.

Let's use Africa/Asia as examples Africa had 1 playable nation, Asia has a couple but in terms of NewGens the only good one is South Korea. Both continents have many good newgens from leagues that dont exist in FM. If you don't force load those 2 entire continents they you will get 90%+ greyed out players even on their national teams.
While if you force load 100% of them you wont get any grey players anywhere. Africa has a handful+ really good newgen nations and Asia has several that are in the middle east but Asia continent.

Some notable nations are Egypt, Jordan, Nigeria, Japan they all rank top 15 youth in FM since like FM18 or something. But the African national teams are pretty well stacked as an example take the Cameroon team. The players are all really up there in age but they are quality so you will get all of their newgen replacements. When you don't load players you are missing out on all of that. And if you think South American and Korean wonderkids are cheap wait until you see Africa/Asian prices.

Now making all of the fm leagues viewable you still get league tables, history and they are easy to browse but it doesn't effect sim speed at all. Biggest drawback is only a couple teams will be ultra competitive in the league. Example would be France. If it's only viewable PSG will be the top team for 100 years. As a player you are still getting the best from France in champions competition anyway so it doesn't matter if it's PSG or another team really. Now if the league is fully playable another team is more likely to eventually take PSG spot and even become the big dog of the nation. Doesn't always happen but it's possible.

The speed factor of playable leagues vs all viewable and large player databases is night and day. Especially when playing 10+ years out. The viewable option even with 200k+ players will always be faster at sim speed.

Overall your research is great. But you should be doing it based upon how to load players. It's honestly the biggest misunderstood thing in FM. You have all these potato laptops setting South America to playable 2 or 3 tiers deep. When they are never leaving Europe and it kills their sim time. It is probably also why people don't play saves deep eventually that daily sim speed takes it's toll on people.

Z even did a video I believe last year about players, playable and view only leagues but it probably needs another one. Honestly the majority of FM users don't even know there is option to load just players. I see it on reddit everyday. People's minds are blown when I'm like just load players not leagues.
 
I felt like view only leagues didnt produce much newgens. After a while the real players disappear and they arent being replaced with newgens, but with grey players.
 
I felt like view only leagues didnt produce much newgens. After a while the real players disappear and they arent being replaced with newgens, but with grey players.

Correct. If you do not force load players by setting up the database then you get grey players. Force loading players is the key. It can only be done before you start your save when setting up the database. That is why its so important. It is also why it is so overlooked.

So many fm users miss this. Honestly SI does an awful job hiding the custom database options. It should be in a drop down just like playable/viewable should be an option in there viewable + players. It is one of the few areas where the user interface is a complete miss.

If you want say the Ivory Coast to have newgens the only way is to force load players at the nation/region/continent level. And they don't even have a league in FM.
Let's use Belgium as an example now. IF Belgium is view only then you will get grey players even though it is a big European nation. However is you force load Belgium in the player database they will have 100% of their players and full newgens the entire save even on view only or even if you dont load the league they still get all their players forever.
 
Loading players doesn't really start to slow anything down until you are like 15+ years out in save. The average FM user doesn't take it that far. But loading full playable leagues on certain potato computers can slow it down from day one.

So I did a test of this:

England Premier League + 'Large' Database
Time to load 1 year: 5:41
Total newgens after 1 year: 851
Rennes newgens: 2
Kano Pillars newgens: 0

England Premier League + 'Large' Database + All Players
Time to load 1 year: 31:38
Total newgens after 1 year: 44,929
Rennes newgens: 5
Kano Pillars newgens: 5

16 Leagues + 'Large Database' (data taken from when I did my article)
Time to load 1 year: 25:44
Total newgens after 1 year: 11,292
Rennes newgens: 8
Kano Pillars newgens: 0

For my article I also found that loading all players of English nationality (+16.2% players) caused a 4.26% slowdown.

Personally I would favor a custom edited DB + 16 leagues loaded than 1 league with all players loaded. If you're playing in England, you won't be able to sign a great deal of those 45K newgens because they're non-EU and outside top 50 nation ranking. It's also a waste to have a bunch of 30 PA newgens for New Caledonia. With a custom edited DB you can leave a country such as New Caledonia with ~30 players which should last it for a number of years before the players start getting too old (a few newgen replacements should come in anyway).

I guess though it might be better to 'load players from top clubs' or similar for good newgen producing divisions that don't have much relevance or transfer activity, such as Serbia or Ukraine
 
Good stuff. I don't think anyone should load 100% all players everywhere though.

England has insane number of players cause it has so many tiers. But the vast majority aren't important to anyone.

We need the info to show us which players to load where so we can be selective about it.

I've been cutting it up and getting it to around 225k total players. But I'm sure it could be really sliced up to give both best newgens and best speed.

I mostly just focused on slicing up europe individually. But north america and asia probably could be really sliced up to get the best outcomes.
 
So many fm users miss this. Honestly SI does an awful job hiding the custom database options. It should be in a drop down just like playable/viewable should be an option in there viewable + players. It is one of the few areas where the user interface is a complete miss.

Yup - that's me. Will definitely change the way I set up games in the future. For a game that's already started, presumably the only way to increase newgens is to add playable leagues then...grrrrrrr
 
So many fm users miss this. Honestly SI does an awful job hiding the custom database options. It should be in a drop down just like playable/viewable should be an option in there viewable + players. It is one of the few areas where the user interface is a complete miss.

Yup - that's me. Will definitely change the way I set up games in the future. For a game that's already started, presumably the only way to increase newgens is to add playable leagues then...grrrrrrr
 
Top