The first underlined number is the median PA of a newgen intake of that club in FM19. Each subsequent number is a sample from FM21. The final number is the average of all the FM21 samples.

This final number is not precise, but to keep it in perspective, the degree of error is less than the random variation in quality of intakes year-to-year. For my thoughts on such caveats, and why I use direct testing of the median PA rather than any other measure or method, scroll to the bottom of this article.

Summary of Findings

The results differ considerably from FM19. Whilst most of the ‘visible’ attributes have remained the same, ‘hidden’ attributes (scroll to bottom for more on this) have on the whole bumped up many clubs from mediocre or middling PA to fairly high PA. The extreme values remain about the same, but England is considerably improved and is now top dog, with almost all Premier League clubs — and even a few Championship clubs — making the top list. France, Brazil & Argentina remain strong. Italy has slipped. Many previous ‘value’ nations such as Serbia and Greece now have comparatively reduced quality, but are nonetheless decent with most of their Div 1 clubs at a solid~90–100 PA median. South Korea remains a standout candidate for inclusion.

The Top 100 clubs

And here are the next 50, but note that while this lot is comprehensive, it is not exhaustive:

The Best Divisions to Load

The numbers inside each column above denote the number of teams that have 120+ median PA. I’ve found that my recommendation for what divisions to load in FM19 was too burdensome. I would now recommend loading England through to Belgium, and South Korea, as well as all divisions in your selected nation. Div 2 of England & France should get the chop as their players are hard to buy. Colombia, Ukraine, Mexico, Brazil Div 2 & Uruguay are non-EU and thus should be dropped. Denmark has to go to slim things down. South Korea should be included due to having good players, merchandizing benefits, and a possible loan club option. For the same reasons, USA should be considered, but South Korea is about half the burden on processing.

Please note that the images will be subject to revision, and I will at some stage add other inactive nations and lesser divisions. I figure it’s good enough to release as is, even though it’s a work-in-progress. I have already sighted an error in that ‘England Div 2’ should be 125, not 130, but I will update this later.

Methodology, reasoning, and my response to SI’s call for ‘researchers’

Ideally a newgen has high starting CA, PA, consistency, professionalism, determination, and so on.. of all these things only one is set by the origin club/division/nation that cannot later be altered, and that is the PA.

Now there are four possible ways to assess the PA data: Average, mean, median and distribution. Average and mean are useless, because the tail ends of the distribution are randomized. ‘Game importance’ of a nation does effect the likelihood of a broad distribution (i.e. 70–195 PA, instead of 70–150 PA), but whether your best newgen is 170 PA or 190 PA one year, is mostly randomness. The median proves to be a reliable indicator of quality, and while it varies by about -/+~15 PA each year, taking multiple samples generally reduces this to ~5–10 PA of the ‘real’ average. Considering one is unlikely to play beyond a few years in the game anyway (5 years = 5 samples), the imprecision simply reflects what you get in reality. Taking into account the distribution would be important, and useful, but it is too difficult for me to do. In any case, it’s common sense that Real Madrid is going to have a broad distribution, whilst Guarani FC in Brazil Div 2 is going to have a bunched up one. The most important thing is to know how divisions *actually* compare to each other, then you can predict the value of particular clubs, which is what I will now briefly explain.

Earlier I mentioned ‘visible’ and ‘hidden’ attributes. By ‘visible’ attributes I refer to ‘youth rating’, ‘game importance’, ‘league reputation’, and club facilities. By ‘hidden’ attributes, I refer to some hidden national or competition factor, or some as yet unknown confluence of factors, that is necessary to explain the differences. I won’t go deep into this here, but consider the following examples:

Saint-Etienne — 15YF/15YR/16TR/15YC — 133.6 PA — 140 youth rating
Schalke — 16YF/15YR/16TR/15YC — 113.25 PA — 155 youth rating

Wolfsburg — 16YF/14YR/15TR/13YC — 107 PA — 155 youth rating
FC Metz — 16YF/12YR/12TR/11YC — 129.75 PA — 140 youth rating

Charlton (Div 3) — 12YF/13YR/11TR/13YC — 118.25 PA — 120 youth rating

As you can see, either the nation or competition these clubs are in, is having a hidden effect. And if you think of all the other possible factors, such as ‘cities’, ‘youth importance’ and so on.. yes, I have tested all of these independently, and have observed no relevant effect. Because of this hidden factor, or set of hidden factors, I have been unable to create a predictive formula. It’s quite close, and I have some ideas, but for now the formula that unlocks everything remains elusive.. so instead I’ve gone for the next best thing, which is to simply draw on the actual results to create a full picture.

One critique you may have, is that if one can’t pin down PA values with 100% accuracy, or indeed load more than a few leagues whenever one plays, then what is the point of knowing all this information? First off, it’s worthwhile knowing what clubs produce the best newgens so we can see who to look at or manage, if you like cultivating newgens like me — the level of accuracy is adequate for this. Secondly, and this is much less a matter of opinion, if people are being kept in the dark and also deluded by ‘youth rating’ and so forth, then custom databases will be inherently flawed, messy and pointless. And in regards to just the player, one should know whether or not ‘youth facilities’ actually train young players (it doesn’t), or whether or not their team from a minnow nation can become world-class.

To give this an ending, I would like to use this opportunity to take a shot at the untermenschen that ‘work’ at SI. I present the following without further comment:

The real effect of ‘youth facilities’ is only a small effect on starting newgen CA. It has no effect on PA, at all. If you try to post this on the SI forums you get held up by mods & SI staff repeatedly saying you need to ‘upload your full data for analysis by the SI team’ (as though it’s data that they’re entitled to and no one else can understand) or ‘post it in the bugs forum’ (to which you receive no response), and then you get banned. And if you try to post again later, you get banned for having an ‘alias account’.

Methology

One way to prove it is to simply change Barcelona’s ‘youth facilities’ from ‘20’ to ‘1’ in the pre-game editor. But several critiques could be made of this method. Perhaps the game resets the value to the default when it starts. Or maybe there are already pre-existing but hidden youth players, and any change to the ‘youth facilities’ value would take a few years to play out. There’s also the possibility that it only comes into play if say youth recruitment between two clubs is even (which is how club reputation works), so setting ‘youth facilities’ to ‘1’ might make no difference for some clubs, but possibly will in others.

Therefore rather than creating and testing an artificial difference, I believe the best way is to simply to compare the default values of clubs with their actual results in the game (youth intake median PAs).

South Korea — Gyeongnam (17 YR/16 YC/12 YF) 103 PA vs. Ulsan (17 YR/16 YC/16 YF) 101 PA
Russia — Lokomotiv Moscow (12 YR/10 YC/11 YF) 104.5 PA vs. Krasnodar (10 YR/12 YC/20 YF) 98 PA
England — Liverpool (16 YR/16 YC/17 YF) 120 PA vs. Tottenham (16 YR/16 YC/20 YF) 113.5 PA
Italy — Empoli (14 YR/13 YC/12 YF) 115 PA vs. Bologna (14 YR/13 YC/16 YF) 119 PA
Northern Ireland — Glenavon (17 YR/14 YC/5 YF) 90 PA vs. Dungannon (16 YR/14 YC/12 YF) 91 PA
Bulgaria — Levski (20 YR/15 YC/14 YF) 118 PA vs. Ludogorets (20 YR/15 YC/17 YF) 99.5 PA
Ireland — Bohemians (6 YR/6 YC/4 YF) 73 PA vs. Sligo Rovers (6 YR/5 YC/12 YF) 66.5 PA
Turkey — Bld. Erzurumspor (2 YR/5 YC/5 YF) 88 PA vs. Kayserispor (2 YR/5 YC/13 YF) 66 PA
Portugal — Vit. Setubal (12 YR/14 YC/6 YF) 109.5 PA vs. Boavista (12 YR/15 YC/12 YF) 97 PA
Denmark — OB (14 YR/13 YC/6 YF) 82.5 PA vs. AaB (14 YR/14 YC/13 YF) 95.5 PA
Poland — Wisla Plock (12 YR/8 YC/6 YF) 92.5 PA vs. Slask (11 YR/9 YC/10 YF) 79 PA
South Africa — Cape Town City (13 YR/11 YC/9 YF) 101 PA vs. Kaizer Chiefs (13 YR/12 YC/15 YF) 99 PA

Taking all these results together we get an average of -1.04 PA per +1 ‘youth facilities’. Don’t worry, youth facilities don’t make your PA go backwards; what is happening here is mostly just random variation, but also to some extent slight differences in youth recruitment and youth coaching. For instance, even with Liverpool and Tottenham having identical youth recruitment, Liverpool ends up higher in the pecking order because it is then sorted by club reputation.

For comparison, here’s what we get with differences in ‘junior coaching’:

Russia — CSKA Moscow (14 YR/11 YC/12 YF) 102 PA vs. Dinamo Moscow (14 YR/14 YC/12 YF) 107 PA
England — Newcastle (13 YR/8 YC/13 YF) 96 PA vs. Crystal Palace (13 YR/13 YC/12 YF) 107 PA
France — LOSC (14 YR/11 YC/16 YF) 107 PA vs. FC Nantes (14 YR/14 YC/15 YF) 106 PA
Uruguay — Cerro (13 YR/12 YC/7 YF) 100 PA vs. Liverpool de Montevideo (13 YR/15 YC/11 YF) 95.5 PA
Belgium — STVV (14 YR/11 YC/12 YF) 87 PA vs. Charleroi (14 YR/14 YC/14 YF) 99.5 PA
Northern Ireland — Linfield (18 YR/6 YC/5 YF) 69 PA vs. Glenavon (17 YR/14 YC/5 YF) 90 PA
Ireland — St Pat’s Athletic (14 YR/5 YC/8 YF) 65 PA vs. Limerick F.C. (14 YR/8 YC/12 YF) 70 PA
Slovakia — Ruzomberok (14 YR/11 YC/9 YF) 78 PA vs. Michalovce (14 YR/14 YC/10 YF) 88 PA
Argentina — Huracan (16 YR/9 YC/14 YF) 96 PA vs. Newell’s (16 YR/13 YC/14 YF) 120.5 PA
Norway — Start (17 YR/11 YC/10 YF) 83 PA vs. Stabaek (17 YR/15 YC/16 YF) 101.5 PA

There are a few blips here and there, but the average is +2.61 PA per +1 ‘junior coaching’. I can say with a fairly high degree of confidence that the ‘real’ figure is between +1.14 PA and +6.25 PA, with my best guess being 3.08 based on various things, although as I’ll now attempt to explain, the figure has significant variation due to other factors (and not just random variation either).

But before that, if you’re still not yet thoroughly convinced that ‘youth facilities’ has no effect on newgen PA, consider that only 1 club in Brazil Div 1 has ‘youth facilities’ higher than ‘13’ (and even that is only ‘15’), yet as I’m sure we all can agree from our own observation, Brazil produces some of the best newgens in the world. ‘Boca’ in Argentina only has ‘13’. And conversely, ‘Burton’ in England League One has ‘20’, yet who has heard of good players coming out of Burton, at least like we do of Crewe or Charlton? Additionally I can say that artificially changing the ‘youth facilities’ to ‘1’ in the editor, I’ve had the same results over and over again — no statistically significant difference in newgen PA.

One day I hope to be able to finally write an article titled ‘The mechanics behind Football Manager revealed’, however as of now the exact formula continues to elude me. What I can say with certainty is that ‘youth recruitment’ and ‘junior coaching’ are key factors for newgen PA. But a 17 YR/14 JC club in England is not the same as a 17 YR/14 JC club in Brazil. You might think that’s for a number of reasons, but in the same way I’ve ruled out youth facilities through deduction based on comparison of real results, the following factors can be ruled out:

City population
City attraction
Nation reputation
Club reputation
Division reputation
Training Facilities
Youth importance
Manager quality
Country population
U18 division reputation
Game importance
State of development (nation)
Continent regional strength
Economic factor
And so on..

I have also tested many of these factors individually and found they did not effect newgen PA.

With ‘youth recruitment’, ‘junior coaching’, ‘nation youth rating’, ‘nation ranking points’ and some numbers thrown in, I can get fairly accurate predictions that are about as useful as having taken 1 sample of each club’s real results. It only works if I weight youth recruitment, junior coaching, and national youth rating (divided by 10) as equal.

In individual testing, I found that the difference between ‘200’ and ‘1’ ‘national youth rating’ for a certain club reduces PA by ~25%. I didn’t find any notable effect for changing nation ranking points, but I only did a small test of that at the time.

I have two theories. One is that there is some unidentified or hidden nation or division factor, where the values generally match up with nation rankings. Another is that ‘youth recruitment’ is more complex than a simple number from 1 to 20, and perhaps ties into ‘nation youth rating’, cities/regions, the number of clubs in a division, the youth recruitment ratios in that division, and so on. What I did notice that low junior coaching with high youth recruitment seems to have highly variable results, even though the average ends up as expected. To illustrate:

Linfield (18 YR/6 JC — Northern Ireland Div 1):

71.5
54.5
67.5
78.5
74
65
83.5
67
86
64.5

Average = 71.2

Glenavon (17YR/14 JC — also Northern Ireland Div 1)

91
88
90.5
90.5
92.5
82
89
89.5
88
87

Average = 88.8

It’s as though ‘youth recruitment’ is setting the peak potential, while ‘junior coaching’ determines the chance of that peak potential being actualized, or conversely, falling short of that potential. This makes sense if ‘junior coaching’ is to represent what it means. Because of this, ‘youth recruitment’ would also matter a bit less as the end PA result is doubly dependent on junior coaching. This seems to match up with individual testing I’ve done showing that ‘youth recruitment’ has a max ~25% effect, whereas junior coaching can have a max ~40% effect. That said, I would take this with a massive grain of salt until more clubs are tested.

I’ll conclude this article with samples taken of four clubs, to help you draw your own conclusions:

Arsenal (18 YR /16 YC/16YF — 14 samples)

126
113
109
123
125
102
97
137
117
123
115
118
128
131

average = 118.857
min = 97 (-18.39%)
max = 137 (+15.26%)
max variance excluding two outliers = -9.17%/+10.21%

Brighton (14 YR/11 YC/18 YF — 10 samples)

117
105
101
106
113
106
118
103
104
127

average = 110.0
min = 101 (-8.18%)
max = 127 (+15.45%)
max variance excluding one outlier = -8.18%/+7.27%

Stuttgart (20 YR/14 YC/16 YF — 10 samples)

121
123
109
115
113.5
108.5
107.5
123
125.5
114.5

average = 116.05
min = 109 (-6.44%)
max = 125.5 (+7.72%)
max variance excluding one outlier = -6.44%/+5.58%

Sao Paulo (20 YR/14 YC/13 YF — 10 samples)

122.5
127.5
116.5
134.5
123.5
112
123.5
107
115
120

average = 120.2
min = 107 (-10.98%)
max = 134.5 (+11.90%)
max variance excluding one outlier = -10.98%/+6.07%

Important note: All testing done with FM19.

In my last article where I tested what effects processing time, I attempted to make a division selection guide by weighting both processing time and the median Newgen PA of the best club in the division.

After some thinking, I’ve come to realize that this is an inadequate method. Rather than using just the best club in the division, all clubs in the division should be taken into account. As you’ll see, this is shown most starkly by comparing Romania Div 1 to Colombia Div 1. In Romania, FC Viitorul produces median PA of ~130, but only 4 clubs produce median PA of 90 or more. In Colombia, their best is Atletico Nacional’s ~125 PA, but 17 clubs produce median PA of 100 or more, making Colombia Div 1 a significantly better choice to load than Romania Div 1.

First, I’ll explain the methodology. I have observed the actual median Newgen PA of every playable club in the game. As only 1 sample of each club has been taken, this will be within ~10% of the ‘real’ average median PA ~95% of the time. For instance, Shamrock Rovers are recorded as 71 PA. In reality, the actual average might be 65 PA or 78 PA.. or even higher/lower if this is one of the rare outliers. So take with a grain of salt a difference within ~10 PA, but you can be confident that Chelsea (142 PA) will very rarely be worse than Aberdeen (81 PA). When you consider that in-game values such as facilities and reputation will fluctuate over time to a degree, being more precise makes little sense anyway.

Why the median PA? And why not just use national youth rating, youth recruitment, etc.— I will explain in more detail in a future article the mechanics of FM, but I’ll give a basic overview now in this one. The most reliable measure of newgen quality is the median PA, which is fairly consistent. The best PA in an intake is on the other hand random, only loosely tied to the median through a deviation curve. For instance, Connah’s Quay (68 PA) will very rarely produce a 180 PA regen, but Paris SG (141 PA) is likely to do it with some regularity. The PA must be measured rather than calculated, because the facilities are only part of the equation, and even adding the nation youth rating isn’t sufficient.

There is a lot of misinformation out there about the game’s mechanics, which is in large part due to SI’s incompetence, not only in their failure to communicate the mechanics of the game (i.e. confusing naming, refusal to explain) but to even understand it themselves!

In reality, ‘youth facilities’ don’t have any effect on newgen PA. The story for ‘game importance’ is a bit more complicated, but certainly doesn’t effect PA ‘equally’ as other factors. Here are a few factors that have no effect at all on newgen PA, which might surprise you:

Youth Importance
Youth Facilities
Training Facilities
Nation Reputation
City Attraction
City Inhabitants Range

The most important factors for PA are youth recruitment, youth coaching (aka junior coaching), nation youth rating, and unique nation and division. No one else seems so far to have appreciated the last two factors, because they are hidden from the user. You can swap absolutely everything between Romania and England, and still England will produce significantly better newgens and Romania worse. It is also the reason why the club in Div 2 with 15 youth recruitment can’t match a Div 1 club with 15 youth recruitment, but can still be better than a few of the bottom Div 1 teams — it’s because each division is treated separately and unequally. Two other factors, game importance and club reputation play lesser roles. All club reputation does for PA is determine who comes first out of two clubs in the same division with equal youth recruitment.

You may be wondering why SI staff do not seem to understand how their own game works. This is because the core of the game that we all love it for, was actually developed over two decades ago by the Collyer brothers, and has remained largely unchanged since. Rather than continue to develop the game, the underpaid soyboys hired over the past ~15 years don’t know and don’t care to know about the mechanics of the game, preferring instead to dazzle us with all their fluff such as regen faces, a 3D match engine, MoCap (now abandoned), yearly reskins, press conferences, and injecting their political agenda such as BLACKED charities, Brexit punishment, homosexuals ‘coming out’, removing a core mechanic of the game because it was pointed out SI’s own assessment of reality was racist (national personality attribute templates were removed the following year, with no replacement to balance), and so on. And in case you were wondering, the Collyer brothers themselves have, understandably, moved on from the core programming of the game to pursue fresh projects.

Getting back to the subject, I’ll explain the scoring system. This is subjective and you can substitute your own values.

8 points for 150+ PA
7 points for 140–149 PA
6 points for 130–139 PA
5 points for 120–129 PA
4 points for 110–119 PA
3 points for 100–109 PA
2 points for 90–99 PA
1 point for 80–89 PA

In this system, a division with one 130–139 PA club is equal to a division with six 80–89 PA clubs. This seems about right to me, as high PA players are what we are looking for, but lesser decent players can serve a role too.. not to mention each 80–89 PA club has a low, but not negligible chance of producing a 150+ PA regen.

After that, the total points are divided by the number of teams in the division. This roughly weights the quality of the division with it’s processing time.

The Results

Brazil Div 1 = 1 x 140, 2 x 130, 5 x 120, 6 x 110, 3 x 100, 1 x 90, 1 x 80 = 80 / 20 = 4.000
France Div 1 = 2 x 140, 2 x 120, 6 x 110, 9 x 100, 1 x 90 = 77 / 20 = 3.850
Argentina Div 1 = 3 x 130, 7 x 120, 3 x 110, 6 x 100, 4 x 90 = 91 / 24 = 3.792
Germany Div 1 = 2 x 120, 9 x 110, 6 x 100, 1 x 90 = 66 / 18 = 3.667
England Div 1 = 1 x 140, 1 x 130, 6 x 120, 2 x 110, 6 x 100, 2 x 90 = 73 / 20 = 3.650
Colombia Div 1 = 2 x 120, 1 x 110, 14 x 100, 3 x 90 = 62 / 20 = 3.100
Brazil Div 2 = 2 x 120, 7 x 110, 4 x 100, 4 x 90, 2 x 80 = 60 / 20 = 3.000
South Korea Div 1 = 2 x 120, 1 x 110, 6 x 100, 2 x 90 = 36 / 12 = 3.000
Switzerland Div 1 = 4 x 110, 3 x 100, 1 x 90, 2 x 80 = 29 / 10 = 2.900
Spain Div 1 = 1 x 130, 4 x 120, 3 x 110, 5 x 100, 2 x 90 = 57 / 20 = 2.850
France Div 2 = 5 x 120, 1 x 110, 5 x 100, 3 x 90, 5 x 80 = 55 / 20 = 2.750
Denmark Div 1 = 1 x 130, 1 x 120, 1 x 110, 6 x 100, 2 x 90, 1 x 80 = 38 / 14 = 2.714
Mexico Div 1 = 1 x 120, 2 x 110, 6 x 100, 6 x 90, 3 x 80 = 46 / 18 = 2.556
Belgium Div 1 = 1 x 120, 1 x 110, 6 x 100, 4 x 90, 4 x 80 = 39 / 16 = 2.437
Uruguay Div 1 = 1 x 120, 1 x 110, 3 x 100, 9 x 90, 2 x 80 = 38 / 16 = 2.375
Ukraine Div 1 = 1 x 120, 1 x 110, 4 x 100, 2 x 90, 2 x 80 = 27 / 12 = 2.250
Austria Div 1 = 1 x 120, 3 x 110, 1 x 100, 5 x 80 = 25 / 12 = 2.083
Sweden Div 1 = 3 x 110, 3 x 100, 4 x 90, 4 x 80 = 33 / 16 = 2.063
Czech Republic Div 1 = 6 x 100, 5 x 90, 5 x 80 = 33 / 16 = 2.063
Serbia Div 1 = 2 x 110, 2 x 100, 7 x 90, 4 x 80 = 32 / 16 = 2.000
Italy Div 2 = 1 x 110, 5 x 100, 4 x 90, 9 x 80 = 36 / 19 = 1.895
Russia Div 1 = 1 x 110, 6 x 100, 2 x 90, 4 x 80 = 30 / 16 = 1.875
South Africa Div 1 = 1 x 110, 3 x 100, 6 x 90, 3 x 80 = 28 / 16 = 1.750
Slovenia Div 1 = 1 x 110, 3 x 100, 1 x 90, 3 x 80 = 17 / 10 = 1.700
Spain Div 2 = 3 x 110, 4 x 100, 4 x 90, 3 x 80 = 35 / 21 = 1.667
Australia Div 1* = 2 x 110, 2 x 100, 5 x 90 = 16 / 10 = 1.600
Scotland Div 1 = 1 x 110, 1 x 100, 4 x 90, 4 x 80 = 19 / 12 = 1.583
Peru Div 1 = 1 x 110, 3 x 100, 5 x 90 = 23 / 16 = 1.438
Poland Div 1 = 2 x 110, 5 x 90, 7 x 80 = 23 / 16 = 1.438
Slovakia Div 1–2 x 100, 3 x 90, 5 x 80 = 17 / 12 = 1.416
Argentina Div 2 = 1 x 110, 7 x 100, 7 x 90, 2 x 80 = 41 / 29 = 1.414
Belgium Div 2 = 1 x 100, 2 x 90, 4 x 80 = 11 / 8 = 1.375
Bulgaria Div 1 = 1 x 110, 5 x 90, 5 x 80 = 19 / 14 = 1.357
Germany Div 2 = 4 x 100, 2 x 90, 8 x 80 = 24 / 18 = 1.334
Norway Div 1 = 1 x 110, 2 x 100, 3 x 90, 5 x 80 = 21 / 16 = 1.312
Netherlands Div 2 = 4 x 100, 4 x 90, 6 x 80 = 26 / 20 = 1.300
Romania Div 1 = 1 x 130, 2 x 100, 1 x 90, 3 x 80 = 17 / 14 = 1.214
Chile Div 1 = 1 x 110, 2 x 100, 2 x 90, 5 x 80 = 19 / 16 = 1.188
Turkey Div 1 = 1 x 120, 1 x 110, 3 x 90, 5 x 80 = 20 / 18 = 1.111
Finland Div 1 = 1 x 100, 2 x 90, 6 x 80 = 13 / 12 = 1.083
Greece Div 1 = 2 x 100, 2 x 90, 7 x 80 = 17 / 16 = 1.063
Switzerland Div 2 = 1 x 110, 1 x 90, 3 x 80 = 9 / 10 = 0.900
Hungary Div 1 = 9 x 80 = 9 / 12 = 0.750
Portugal Div 2 = 2 x 100, 2 x 90, 3 x 80 = 13 / 18 = 0.722
China Div 1 = 1 x 100, 1 x 90, 5 x 80 = 10 / 16 = 0.625
Northern Ireland Div 1 = 2 x 90, 1 x 80 = 5 / 12 = 0.416
Ireland Div 1 = 1 x 90, 2 x 80 = 4 / 10 = 0.400
Croatia Div 1 = N/A
USA Div 1 = N/A

*Australia has additional A-League youth teams at the inactive second level which nonetheless produce additional good regens. While the PAs will still be fairly accurate, the end ‘score’ probably needs to be reassessed.

EU Only

France Div 1 3.850
Germany Div 1 3.667
England Div 1 3.650
Switzerland Div 1 2.900
Spain Div 1 2.850
Netherlands Div 1 2.777
France Div 2 2.750
Denmark Div 1 2.714
Belgium Div 1 2.437
Portugal Div 1 2.111
Austria Div 1 2.083
Sweden Div 1 2.063
Czech Republic Div 1 2.063
Serbia Div 1 2.000
Italy Div 2 1.895
Slovenia Div 1 1.700
Spain Div 2 1.667
Scotland Div 1 1.583
Poland Div 1 1.438
Slovakia Div 1 1.416
Belgium Div 2 1.375
Bulgaria Div 1 1.357
Germany Div 2 1.333
Norway Div 1 1.312
Romania Div 1 1.214
Finland Div 1 1.083
Greece Div 1 1.063
Hungary Div 1 0.750
Portugal Div 2 0.722

Non-EU only

Brazil Div 1 4.000
Argentina Div 1 3.792
Colombia Div 1 3.100
South Korea Div 1 3.000
Brazil Div 2 3.000
Mexico Div 1 2.556
Uruguay Div 1 2.375
Ukraine Div 1 2.250
Russia Div 1 1.875
South Africa Div 1 1.750
Australia Div 1 1.600
Peru Div 1 1.438
Argentina Div 2 1.414
Chile Div 1 1.188
Turkey Div 1 1.111
China Div 1 0.625

Observations and caveats

Croatia produced few regens as the clubs were already stocked up with existing young players, and therefore were not assessed. USA youth intake seems to work differently and requires separate assessment. I have not yet tallied the lesser leagues not included above, but they are all probably not worthwhile loading. Some nations are of interest but need to be edited in and were therefore not assessed, such as Japan and Paraguay.

Colombia, South Korea and Switzerland are surprisingly solid, even though they don’t produce the absolute best newgens. France Div 2 is better than the majority of 1st divisions. Argentina Div 2 is too weak to include given its large number of teams dragging down performance. Ukraine, Russia and South Africa do quite well, but their non-EU status make them necessary to dispose of. Chile, Peru, Greece, Romania and Norway were notably disappointing. Romania is a curious case where there was one really good club (130 PA) and not much else, dropping it out of consideration.

All testing done with FM19, so future versions will be a bit different, but shouldn’t be much different.

Number of teams in an active division has a strong correlation with processing time. I suspect the minor discrepancies are due to greater player numbers loaded, and perhaps level of transfer activity.

Division selection is a matter of taste, and as us patricians know, some people don’t have any. So I decided to make the ‘normal’ setup something typical of what I’d personally select, and favoring divisions with good newgens when expanding my selection.

The most reliable measure of newgen quality is the median PA of the youth intake. You can generally expect the margin of error to be +/- 10%. My calculations use the average median PA of the best club in that division, based on real samples and not theory. I will explain and evidence all this in a future article.

The figures in brackets are simply the average median PA of the best club in that division, divided by the number of teams in the division, giving a quality vs. performance ranking. For example, Germany Div 1 has 18 teams, and the best median PA produced by a team was 126 PA (both Leverkusen & Stuttgart). 126 / 18 = 7.

The essentials
Germany Div 1 (7.000)
England Div 1 (6.900)
Brazil Div 1 (6.900)
France Div 1 (6.750)
Argentina Div 1 (5.192)
Spain Div 1 (TBA)
Italy Div 1 (TBA)
Portugal Div 1 (TBA)
Austria Div 1 (TBA)

The optional
Belgium Div 2 (11.750)
Croatia Div 1 (11.000)
Switzerland Div 1 (11.000)
Switzerland Div 2 (10.700)
Slovenia Div 1 (10.000)
South Korea Div 1 (9.500)
Ireland Div 1 (9.300)
Slovakia Div 1 (8.750)
Romania Div 1 (8.643)
Denmark Div 1 (8.357)
Scotland Div 1 (8.333)
Greece Div 1 (7.928)
Belgium Div 1 (7.875)
Northern Ireland Div 1 (7.333)
Germany Div 1 (7.000)
Netherlands Div 1 (6.833)
Bulgaria Div 1 (6.786)
Serbia Div 1 (6.687)
Norway Div 1 (6.687)
France Div 3 (6.444)
Czech Republic Div 1 (6.388)
France Div 2 (6.350)
Sweden Div 1 (6.250)
Japan Div 1 (6.222)
Poland Div 1 (6.111)
Brazil Div 2 (6.050)
Colombia Div 1 (6.000)

Included for comparison
Germany Div 2 (5.666)
Portugal Div 2 (5.611)
Germany Div 3 (5.250)
Netherlands Div 2 (5.150)
England Div 2 (4.666)
USA Div 1 (4.652)
Argentina Div 2 (3.875)
Portugal Div 3 (1.500)

A caveat is that I have not yet taken into consideration whether a division with a number of decent clubs might outweigh a division with only one good club — for instance, France Div 1 has a lower peak PA than England Div 1 (135 vs 138), but 2 more clubs that are 110PA+.

USA Div 1 is an interesting case as it relevant in terms of transfers, merchandizing and general importance, but is tough to include given it has a high number of teams and non-EU status. This is one case where an exemption might need to be made.

Despite having 72,000 more staff and the same number of players, the custom database is slightly faster than Normal + all ‘England’ players loaded. Adding all ‘players of English nationality’ (+16.2% players) only caused a 4.26% slowdown compared to ‘normal 16 leagues’. Number of staff appears to make zero difference to processing time, and I’ve seen this reflected in other testing I’ve done. The slight time improvement I suspect is due to having removed a few players from active leagues, rather than say having to access a smaller database.

For those interested, using ‘Best 32 Leagues’ with only Div 1 from England (instead of Div 1–6) produced a time of 9:44 (31 October 2018).

There were only very minor differences in save load time, that may be partially attributable to stopwatch error. Time to save might be a different story, but I did not bother to test.

Players seem to increase save size by roughly ~1800 players per MB. Staff increase it by ~4400–5100 per MB. This can matter for people who frequently save long games.

Summary:

• Processing time largely determined by number of active clubs
• Active leagues is a decent but lesser measure of processing time
• Number of players has a minor effect on processing time
• Number of staff has no effect on processing time
• Belgium, Croatia and Switzerland are excellent nations to load
• USA is a dilemma, while most lower divisions are best left out