Summary of recent findings for optimal play in FM24

Youth intake stuff

Strap in.

tl;dr
  • Your youth intake players are created on preview day. Their names, ages, nationalities, positions and a hidden "quality tier" are decided then and never change.
  • Their CA and PA are not decided until intake day. The preview is the Head of Youth Development judging a practice roll of each player, not the real thing.
  • PA is driven mostly by junior coaching and your nation's youth rating, with youth facilities a distant third. Training facilities play no part.
  • A hidden nation-wide PA quota then moves players up or down. In England it pushes many players up, but it's first come, first served: once the country's quota for a PA band is used up, later clubs miss out.
  • CA is mostly set at roughly 40–55% of the final PA, adjusted for age.
Where this comes from

I spent a long time pulling apart FM24's game code to trace exactly how youth intakes are made. I then checked it against a real save (Aston Villa) by watching the preview arrive in December and the intake land on 15 March. Everything below matches what the game actually did. Where a number is an example, it uses Villa's values from that save.

A few caveats:
  • This is FM24. Other versions may differ.
  • I've described the rules that apply in a normal save at a club in a big footballing nation. There are a handful of special cases for particular countries, clubs and US academy leagues that I mostly skip (because I've been at this for ages).
  • Anything random is a dice roll. When I say "1 in 20", that's the actual roll in the code.
Part 1: The timeline

WhenWhat happens
93 days before your nation's intake window opens, at 09:00 (mid-December for England)The game creates every youth player for clubs with a human manager, and sends you the Youth Intake Preview
The intake window (mid-March for England, spread over several days)Each club in the nation gets its own day. On your day, the players created in December become real players and their CA and PA are rolled.
End of the windowClean-up, and the game tops up the pool of unattached youngsters.

AI clubs don't get anything in December. Their whole intake is generated from scratch on their intake day.

Part 2: Preview day — creating the players

On preview day the game builds a "blueprint" for each incoming player. A blueprint contains:
  • name, nationality (and any second nationality), place of birth and date of birth;
  • positions (which positions they're natural in);
  • a hidden quality tier.
It contains no CA, no PA and no attributes. Those don't exist yet.

How many players

The number depends on how short of players your nation is. The game compares the number of players it wants in the nation with how many it has.
  • England in my save wanted 4,860 and had 4,539, so it was 321 short. A gap that big gives league clubs 14–16 players.
  • A smaller gap (150–299 short) gives 10–12, and smaller still gives fewer.
  • Reserve and affiliate teams get 3–6. Top-division clubs always get at least 7.
Villa got 16.

The hidden quality tier

Each regular player rolls a tier. This is the most important hidden number in the whole process.

ChanceWhat it means later
Elite0.5%Best PA odds, best CA range, and it skips the quota (see Part 5)
Good18%Strong PA odds
Standard54%Middling PA odds
Weak27.5Poor PA odds
Gap-filler(see below)Fills empty positions; PA kept low

Some adjustments apply:
  • In lower leagues, Elite and Good players can be swapped to Standard.
  • At low-reputation clubs, Weak players have a 1 in 3 chance of becoming Standard.
Gap-fillers are added afterwards to cover any position group your intake has nobody in. They always get the poorest PA treatment.

Villa's 16 were: 4 Good, 4 Standard, 3 Weak and 5 gap-fillers.

Part 3: The preview itself

To fill in the preview, the game takes each blueprint and makes a temporary practice copy of the player. It then runs the full intake-day CA and PA generation on that copy (Parts 4–6). Your HoYD rates each practice player. The real CA and PA are rolled again on intake day.

This is the key point: the preview grades a practice roll from the same dice the real players will use. It is a good guide to the odds, not a promise.

Step 1: Your club's benchmark

The game works out a benchmark (B) from your senior squad:

B = average over your best 13 players of (CA + PA) / 2

A stronger first team means a higher bar for youngsters to clear.

Step 2: Judge each practice player

This section got too long, so I've posted it here.

Step 3: Grades per position group

Players are grouped by their natural position: GK, full-backs, centre-backs, DM, central mids, wide mids, AMC, wingers, strikers and wing-backs. For each group:

score = 4 × (good players) + (average players) − (poor players), kept between −4 and +4

Score−4, −3−2, −101, 234
GradeFEDCBA

A group with no players shows "–".

Examples.
  • Full-backs with one good and two average players: 4 + 2 = 6, capped at 4 → A.
  • A single good player on his own also gets an A.
  • Centre-backs with one average and one poor player: 1 − 1 = 0 → D.
  • A lone poor goalkeeper: −1 → E.

Step 4: The star rating

Take the average R of your top 6 practice players and call it T. Then:

Code:
if T > B:  rating = 20                                  (5 stars)
else:      rating = 11 + 36 × (min(T/B, 1) − 0.75)       (T/B no lower than 0.75)
stars = (rating − 10) / 2

Next, a little HoYD noise is added. It depends on the HoYD's judging ability (1–20):

HoYD judgingNoise on the ratingNoise on the player count
16–20noneexact
6–15±1 (half a star)±1
1–5±2 (one star)±2

Example. B = 120 and the top six average T = 110:
  • T/B = 0.917, so rating = 11 + 36 × 0.167 = 17, which is 3½ stars.
  • A HoYD with judging 10 could show 3, 3½ or 4 stars.
If the top six averaged above 120, you'd see 5 stars before any noise.

A HoYD with better judging gives you a more accurate preview. Nothing I traced suggests he changes the players themselves.
 
Last edited:
Part 4: Intake day — PA

On your club's intake day each blueprint becomes a real player. Name, age, nationality, positions and tier come from the blueprint. Everything ability-related is rolled fresh.

Step 1: The club's youth score

youth score = ( YF/5 (rounded down) + YR × (JC/20) × 0.5 ) × 1.5
  • YF = youth facilities (1–20).
  • JC = junior coaching (1–20).
  • YR = your nation's hidden youth rating (0–200). England's is 135.
  • For a player born outside your nation, YR can be boosted by youth recruitment: YR × recruitment/20, used only if it's bigger.
Villa example. YF 18, JC 14, England's YR 135:

(18/5 rounded down = 3, plus 135 × 0.7 × 0.5 = 47.25) × 1.5 = 75.

Look at the weights:
  • Each point of junior coaching is worth about 5 points of youth score in England.
  • Each 5 points of youth facilities is worth only about 1.5.
  • Training facilities aren't in the formula at all.
The score is then nudged by a small random amount (−2 to +1).

Step 2: Lucky boosts (Elite and Good tiers only)

Elite and Good players get extra chances to have their youth score bumped up:
  • Top-flight lift. In the top division, rand(100) < youth recruitment raises the score to a minimum based on national reputation (between 45 and 80).
  • The "wonderkid lottery".At a club in England, France, Germany, Italy, the Netherlands, Portugal, Russia, Spain, Türkiye or Ivory Coast, with a score of at least 31 (This is the main route to a genuinely world-class prospect):
    • 1 in 21: the score becomes a random number from 0–119, if that's higher;
    • otherwise 1 in 5: a random number from 0–69, if that's higher.
  • Continent bonuses.
    • South America: 1 in 20 → at least 60–89.
    • Africa: 1 in 32 → at least 35–64.
    • Europe gets none.
  • Argentina, Brazil and a couple of specific clubs have their own extra bonuses.
Step 3: The youth score picks a PA range

The score falls into a band. The band plus the tier decides the odds of landing in each PA range. The game rolls from the best range down; the first hit wins. The bands are 0–9, 10–24, 25–39, 40–54, 55–74, 75–89 and 90+. Here are the top three:

Tier170–200150–180130–160110–14090–12070–10050–8030–60
Elite (90–99)6%47%47%
Elite (100+)12%44%44%
Good (90–99)5%32%63%
Good (100+)10%45%45%
Standard8%31%31%30%
Weak/filler74%11%15%

The worst result for an Elite or Good player in this band is 130–160, compared with 110–140 in the band below. Oddly, a Standard player here has a 30% chance of the lower 90–120 range, which can't happen at 75–89.

Youth score 75–89

Tier170–200150–180130–160110–14090–12070–10050–8030–6010–40
Elite3%32%33%32%
Good2%25%37%36%
Standard100%
Weak/filler3%7%58%22%10%

Youth score 55–74

Tier150–180130–160110–14090–12070–10050–8030–6010–40
Elite17%28%55%
Good13%22%65%
Standard100%
Weak/filler4%63%24%9%

Then a few promotion rolls:
  • A short ladder of 1-in-N rolls (1 in 20, 1 in 6, 1 in 3, then 1 in 5, 1 in 4 …) can move a player up one or more ranges. It mostly lifts Weak players from 30–60 or 50–80 to 70–100.
    • Part of the ladder only runs if rand(50) + 50 ≤ the nation's youth rating. That always passes in England.
    • Part only runs if rand(15) ≤ junior coaching, so better junior coaching means more promotions.
  • Youth recruitment jump. A player in the 70–100 or 90–120 range, with a youth score of at least 61, jumps straight to 130–160 if rand(90) + 12 < youth recruitment.
    • Villa (recruitment 17): about 5.6%.
    • Recruitment 20: about 8.9%.
  • A sanity cap usually (19 in 20) stops a player rising more than two ranges above where the first roll put him.
Step 4: Pick the PA value inside the range

ChanceLands in
1 in 4the top 10 points of the range
1 in 4the bottom 10 points
1 in 2the middle

Then +0 or +1 is added. Values of 180 or more are usually re-rolled (half the time from 180, 49 times in 50 from 190).

Example. A Good player lands in 130–160:
  • 25% chance of 150–160;
  • 25% chance of 130–140;
  • 50% chance of 140–150.
 
Last edited:
Step 5: Gap-filler cap

Gap-fillers whose PA comes out at 121 or more are pulled back to 120–130.

Before the quota, Villa's chances of PA 140+ were about 64% for a Good player, 2% for Standard and 0.2% for Weak. Villa's real intake had 10 of 16 players at 140+, including Standard and Weak ones. That's because of the next part.

Part 5: The nation PA quota (the big hidden one)

What the quota is


When you start a new game, FM counts how many players of each nationality sit in each 20-point PA band (0–19, 20–39, … 140–159, 160–179, 180+). That snapshot becomes the nation's target for each band. The game then keeps a live count:
  • new players add to it;
  • retirements and deletions take away from it.
By the time of the first intake, plenty of players have retired, so many bands are below target. England in my save, the day before Villa's intake:

PA band40–5960–7980–99100–119120–139140–159160–179180+
Target326602105814601016288556
Live1304038861317894275577

What it does to a new player

Your club's junior coaching sets a floor and ceiling for how far the quota can move a player:

Junior coachingCan't be pushed belowCan't be lifted above
15–2080–99180+
12–1460–79160–179
9–1140–59140–159
1–820–39120–139

  1. If the player's band is over-full, he's pushed down one band at a time until he reaches a band with room, or the floor. "Over-full" means the live count is well above target, roughly 10–30% over.
  2. If he wasn't pushed down, the game tries to lift him up, one band at a time, until he reaches the ceiling. Each step happens only if:
    • the next band's live count is below target − 5; and
    • he survives a "stop here" roll that depends on his position:
Player is a naturalChance of stopping at each step
Striker or attacking midfielder1 in 12
Defender1 in 8
Midfielder, wing-back or other1 in 10
Goalkeeper1 in 5

3. If his band changed, his PA is re-rolled in a range around the new band (for example 130–160 for the 140–159 band).

There are some exceptions:
  • Elite players skip the quota entirely.
  • Gap-fillers can be pushed down but never lifted.
  • The quota for your club's nation is used, even for a foreign-born player. Villa's foreign-born goalkeeper got lifted to PA 158 by England's room, even though his own nation's band was already full.
Worked example (Villa, junior coaching 14, so the ceiling is the 160–179 band).

A Standard central midfielder rolls PA 124, in the 120–139 band. England's 140–159 band has 275 against a target of 288. That's below 288 − 5 = 283, so there's room.
  • He survives the 1 in 10 stop roll and moves up to 140–159.
  • Next step: 160–179 has 57 against a target of 55, which is not below 50. He stops.
  • His PA is re-rolled in 130–160. He ends up with something like PA 147.
The catch: first come, first served

There's one quota per nation, shared by every club whose intake falls in that window, and it fills as clubs are processed.

England's 140–159 band had room for exactly 8 more players (283 − 275) for the whole country. Villa happened to be one of the first clubs processed and took 5 of those 8. A club processed later that week would have found the band full. Its Standard and Weak players would have stayed around 60–130.

My simulation for Villa of the chance of PA 140+:

TierNo quotaQuota with roomQuota with the 140–159 band full
Good64%88%—
Standard2%66%2%
Weak0.2%52%1%
Gap-fillerunchangedunchangedunchanged

This is the single biggest effect on a typical intake, and most of it is luck of the draw on your intake day.

The preview's practice players also go through the quota, using the room available on preview day. Three months later that room may be very different. That's another reason a 5-star preview can lead to a disappointing intake, or the other way round.
 
Last edited:
Part 6: Intake day — CA

CA is worked out in two stages.

Stage 1: A rough CA from the club

CA score = (4 × junior coaching + youth facilities + first-team reputation / 100) / 2

Villa: (4 × 14 + 18 + 7000/100) / 2 = 72.

That score picks a CA range by tier:

CA scoreEliteGoodStandard/fillerWeak
90+105–13565–10045–7530–50
75–8995–12560–9040–7030–45
55–7485–12545–8035–6025–40
35–5465–9540–6030–5020–35

Club reputation counts here. A bigger first-team reputation means better starting CAs.

Stage 2: CA is reset from the final PA

For most players, once PA is settled (including any quota lift), CA is recalculated as a percentage of PA.

1. Base percentage.

Final PANormal club"Strong club"
Over 145usually 50% (sometimes 52 or 55, rarely 45)52% or 55%
100–145usually 40% (sometimes 45 or 50)45–55%
Under 10020–29%20–29%

A "strong club" is one flagged as big in the database, or whose first team is in the top two divisions. Even then the bonus only applies with chance CA score / 125. For Villa that's 72/125 = 57.6% of the time.

2. Age adjustment.

Age1415161718
Adjustment−5−2.5−50+5

Yes, 15-year-olds really do get a smaller penalty than 16-year-olds.

3. Target.

target = PA × percentage / 100

If the target is above the rough CA, then CA = target + a random number from −5 to +4.

4. Low-PA players keep their rough CA. For players under 100 PA, the target usually comes out lower than the rough CA from Stage 1, so they keep it.

5. CA always ends below PA.

Examples (Villa, real players from the save).
  • Good tier, age 16, PA 161, normal-club roll. 50% − 5 = 45%, and 45% of 161 = 72, giving CA 67–76. Actual: 68.
  • Weak tier, age 15, PA 157, strong-club roll of 55%. 55 − 2.5 = 52.5%, and 52.5% of 157 = 82, giving CA 77–86. Actual: 81.
  • Weak tier, age 16, PA 158, the rare 45% roll. 45 − 5 = 40%, and 40% of 158 = 63, giving CA 58–67. Actual: 59. That's about a 1% chance per player.
  • Gap-filler, PA 75. 20–29% of 75 is below his rough CA, so he keeps the rough CA from Stage 1. Actual: 55.
Rule of thumb: a high-PA youngster arrives with CA at about 40–55% of his PA.

Part 7: Putting it together — Villa's real intake

TierAgeCAPA
Weak165918
Standard1669146
Standard1676160
Good1668161
Weak153173
Good1561145
Standard1653114
Good1667141
Weak1581157
Good1556145
Standard1675153
Gap-filler155575
Gap-filler154781
Gap-filler153765
Gap-filler155562
Gap-filler153699
  • 10 of 16 had PA 140+. That's practically impossible without the quota.
  • England's 140–159 band went from 275 to 283 and stopped at exactly target − 5.
Part 8: What this means for you

I've posted the practical upshot of this guide here.
 
Last edited:
Things I've not dug in yet to (because this took hours) but want to:
  • Where does a player's nationality and personality come from? I've heard for years it's HoYD but now I have my doubts.
  • What influences when you get your intake compared to other teams in your country?
Note that this investigation occurred in a save with only the EPL loaded on a small database (to make sinming quicker) so actual quota numbers will be higher in most cases (this is why it says every team gets 14-16 from a pool of ~300: I only have 20 teams loaded.)
 
Youth intake stuff
Thanks so much

I've been wondering about this for years

So SI weren't lying when they said Youth Facilities affect newgen PA, but neither was I in saying it makes no discernible difference.

When isolated:

Youth Rating 1>135 = 100.5 (approx, depends on junior coaching)
Junior Coaching 1>20 = 96.19 (approx, depends on youth rating)
Youth facilities 1>20 = 3.8

That is the effect on score, which then needs to be translated to PA.

If my math is correct, what you can boil youth facilities down to (roughly) is 1 > 20 = +1 median PA:

Youth facilities 1 > 20 is very roughly ~30% chance of (double the % chance of highest PA newgen)
Translates to 5% shifted up 20 PA
Translates to 0.8 players gaining 20 PA
Translates to 1 newgen gaining 16 PA
Translates to +1 PA for each newgen (from total of 16)

For comparison, after 10 intake samples, the median PA has a confidence of -/+5 PA at best, so I think it's fair to say this would have been indistinguishable from random noise without reverse engineering the code.

Since in practice it bumps a single newgen's PA by a dozen or so, on the occasion it does, it's not completely useless but the effect is still quite miniscule.

In testing I approximated the effect on PA as 40% junior coaching and 25% youth rating. If that still holds true, then youth facilities would be 1.0-1.5%.

Youth Facilities 1>20 seem to result in 66% chance of extra 5-10 CA
0.66 x 7.5 = 4.95 CA
From memory, the CAs I would typically see were all around ~60, so Youth Facilities have ~8.25% effect on CA, which is about what I found (~10%)

It would be good to see the complete set of tables for youth score

Also you've potentially explained what the hidden nation factor is probably constituted by, but to verify it would be good to get data on Singapore say. That way it can be compared to my formula where factor X for England is 62 and Singapore is 25. If its lines up, then it confirms what it is.

In regards to where newgen nationality & personality come from I think I have some info on that, but I'd have to look it up again. Of course it's better to just look at the code. But I'll just say that I don't think HoYD has any effect on nationality and personality, other than he influences 1-2 newgen personalities, and I think from playing myself before that he really does.
 
This is amazing.

Next, a little HoYD noise is added. It depends on the HoYD's judging ability (1–20)

Presumably that's the judging player potential rating for the HoYD? (i.e. not judging player ability)

I'm now also wondering

What determines attribute spread of CA - I remember reading somewhere that the game has archetype player templates it uses, I wonder if that's a real thing
Does it make any difference what the HOYD preferred formation is. Maybe that woudl eg set the probability of which templates newgens are built off
 
The FM24 GK testing is on hold, and will be for a while now, so I'll just post the results I have so far from the larger sample testing I began before:

LBNJlcK.png


Win rate includes draws as 1/3 of win value.

Reflexes, Agility, Acceleration matter. Pace & Stamina matter a little bit. High Command Of Area strangely seems to be a bit of a negative. Technique seems slightly negative. Others listed don't really matter.

What determines attribute spread of CA - I remember reading somewhere that the game has archetype player templates it uses, I wonder if that's a real thing
FM used to have nation attribute templates, for attributes and personality. They removed them after Quartz did an article on the 'racism problem' of their data. I think the article(s) was talking about black EPL players getting low professionalism and higher aggression and whatnot, which is due to volunteer researchers inputting that data for each unique individual, but nation attribute templates seem to got swept up and hidden away in the closet as part of SI's panicked response.

I thought that after that, it was simply lobotomized where personalities would be generated randomly. But as someone pointed out to me, in actual fact its not random (testing showed this is true), and so personality templates still exist, they're just hidden. So I think it's likely the same is true of attribute distribution. They would have changed though, so looking back at old FMs might give you a rough idea, but its probably not the exact same data.
 
I tested a few training schedules for FM24 using LightningFlik's harness, now that it's fixed

I assessed the results using AI with my FMST26 weightings for outfield, so it's not going to be perfectly accurate, but it shouldn't be far off. The score is basically how good the player actually ends up being after 24 months of training, rather than CA.

There's all sorts of factors you can change up. Rather than give specifics, let's just say this is a realistic EPL-level example, combined with 130 MatchXP/month. Quickness focus & double intensity is used, and the player is set to ST only. Therefore these results apply to ST only. I chose ST as I feel it's the most likely position for a newgen we're going to use.

272.13 [Handling][Shot Stopping][Attacking][Physical][Chance Conversion][Aerial Defence][Ground Defence][Distribution]
263.57 [Physical][Match Practice][Attacking][Defending]
262.64 [Handling][Shot Stopping][Attacking][Physical][Chance Conversion][Aerial Defence][Ground Defence][Distribution][Match Practice]
260.05 [Physical]x2[Attacking]x2
259.54 [Physical]x2[Attacking]x2 - with 2x 'Team Match/Match Practice' in simulation (note: separate from MatchXP)
258.58 [Handling][Shot Stopping][Attacking][Physical][Chance Conversion][Aerial Defence][Ground Defence][Distribution][Defending]
256.95 [Shot Stopping][Attacking][Physical][Chance Conversion][Aerial Defence]
252.69 [Physical][Match Practice][Attacking]x2
248.72 [Quickness][Match Practice][Attacking]x2
241.86 [Match Review]
230.80 [Rest]

Selected CA comparison:

119.44 | 272.13 [Handling][Shot Stopping][Attacking][Physical][Chance Conversion][Aerial Defence][Ground Defence][Distribution]
119.69 | 263.57 [Physical][Match Practice][Attacking][Defending]
116.83 | 260.05 [Physical]x2[Attacking]x2
88.678 | 241.86 [Match Review]
88.676 | 230.80 [Rest]

After 60 months of training, full rest was still down at 106 of 144 PA.
[Physical]x2[Attacking]x2 reached and stayed at 129 of 144 PA at 45 months.
Top 2 reached and stayed at 129 of 144 PA at 41 months.

Not sure how long holiday break is in FM, maybe 2 months. Leaves 10 months/year training. That means players will be ready with the top schedules at age ~20-21 roughly.

The reality could be different somehow, but the math suggests full rest would take ~12 years to be ready for the big time (~130 CA), so never. But even then, certain key attributes such as dribbling (56 > 41) and concentration (46 > 30) have declined over 60 months and will continue to decline. So you end up with a Div 2 player.

For [Physical]x2[Attacking]x2 dribbling increased slightly (56 > 65) while concentration stagnated (46 > 47).
For the top schedule, both increased slightly: dribbling (56 > 66), concentration (46 > 57).
 
Does this take in to account where the attributes end up (i.e. how good is the striker considered to be after development) or does it just look at CA?
 
Does this take in to account where the attributes end up (i.e. how good is the striker considered to be after development) or does it just look at CA?
The first score (i.e. 272.13) is the weighted performance % score of the player, like you would see in Genie Scout say. Using the attribute data in the csv file after 24 months of training.

The CA values are separate and I've shown separately.
 
The first score (i.e. 272.13) is the weighted performance % score of the player, like you would see in Genie Scout say. Using the attribute data in the csv file after 24 months of training.

The CA values are separate and I've shown separately.
Could you also add a test using one of the game's default routines to see how a player would grow if you loaded the game and touched nothing?
 
Came up with a superior training schedule while trying to fix the pitfalls of [Physical]x2[Attacking]x2

261.81 [Physical][Attacking][Defending From the Front][Ball Distribution]

Compared to [Physical]x2[Attacking]x2:

Slightly higher performance score
100% > 80% Outfield intensity
100% > 70% GK intensity
Concentration fixed - (46>47 before) > (46>59 now)
Pace and Acc is same (very slightly better in new one)

Could you also add a test using one of the game's default routines to see how a player would grow if you loaded the game and touched nothing?
Need more details. Do you mean do % score as well? And should I do the setup I'm using right now, or do you just want me to start from scratch and try with a random player? What's the purpose?
 
Need more details. Do you mean do % score as well? And should I do the setup I'm using right now, or do you just want me to start from scratch and try with a random player? What's the purpose?
I mean compare these meta schedules with one of the schedules the game gives you by default. I'm curious to see how handicapped a person would be if they ignored training.
 
I mean compare these meta schedules with one of the schedules the game gives you by default. I'm curious to see how handicapped a person would be if they ignored training.
I'll do one, but you know HarvestGreen already did this with a few. I remember one or two of them were actually pretty good - not far from the top. The main downside was higher unnecessary intensity. Most of them are crap though.

Unfortunately the AI I'm using decided to fail just now, and I'm out of free chatgpt uploads right now. I'll post the result when it's working again.

Edit: 'Gegenpress' default schedule, normal intensity, no focus - 256.3. I chose this one because I think it was one of the best default ones, but can only recall vaguely.
 
Last edited:
Youth Intake Minutiae
  • Newgen nationality comes from the club, nation, region and city data (and also possibly the club vision)
  • Positions are determined by gaps in the squads plus random secondary positions
  • Personalities have a 1 in 7 chance of having personal attributes pulled towards the HoYD's (3 * HoYD's attribute + player's attribute) / 4
  • Your players can get a bonus from other members of staff. This is something I didn't realise last night and it invalidates what I said about the HoYD not influencing players, only scouting them.
Youth-staff bonus

Your youth staff can add a small random bonus to the youth score. It only applies when all of these are true:
  • the player is Elite or Good tier;
  • your first team plays in one of the top two divisions;
  • your nation is one of the game's leading football nations. It needs a hidden football-importance rating of 13 or more, and England passes. A few nations the game flags as "major" qualify with 9 or more, but then only for 1 player in 4, at random.
Staff score first

Code:
staff score = ( 2 × (HoYD's JPP + HoYD's WwY) + the JPP of each other youth staff member )
              ÷ ( 4 + the number of other youth staff members )
              (rounded down)
  • JPP = Judging Player Potential and WwY = Working with Youngsters, both 1–20 as shown on the staff profile.
  • The HoYD counts as four people. Both of his attributes are counted twice. Other youth staff count once each, and only their JPP is used.
  • "Other youth staff" means coaches (regular, fitness, goalkeeping, set piece) contracted to any of your teams below your nation's youth age limit. England has a youth age of 20, so U18, U19 and U20 coaches would count. Coaches with multiple jobs count, as do player-coaches.
  • U18 Manager, Assistant Manager and performance analysts do not count.
  • With no HoYD, the score is just the other youth staff's average JPP. With no HoYD and none of those staff, the score is 1.
Villa example.
  • HoYD with JPP 18 and WwY 19: 2 × (18 + 19) = 74, counting as 4 people.
  • Six other youth staff with JPP 11, 7, 12, 11, 11 and 9: total 61, counting as 6.
  • (74 + 61) ÷ 10 = 13.5, rounded down to 13.

Then the bonus:

Code:
cap   = the smaller of: the staff score, or the youth score ÷ 5 (rounded down)
bonus = a random whole number from 0 to cap − 1, each equally likely

Villa example. Staff score 13; youth score 75 ÷ 5 = 15. The cap is 13, so the bonus is 0–12 (average 6).

I watched the same intake twice. Villa's Elite and Good players went from 75 to 76, 77, 80, 81, 83, 85, 85 and 86 before the nudge.

The youth score ÷ 5 limit means a weak youth score caps the bonus whatever your staff are like. For example, a youth score of 40 limits the bonus to 0–7.

What it's worth
  • For Villa, nothing. The best case is 75 + 12 + 1 (nudge) = 88, which is still in the 75–89 band. The PA odds in Step 3 depend only on the band, so the bonus can't change them.
  • It matters when your youth score sits just below a band edge (55, 75 or 90).
Example. An English club with junior coaching 16 and youth facilities 15–19 has a youth score of (3 + 135 × 0.8 × 0.5) × 1.5 = 85.5, rounded down to 85. With a staff score of 12, the bonus is 0–11.

The player reaches the 90+ band if the bonus plus the nudge is at least 5. That's 26 of the 48 equally likely combinations, so 54%. Without the bonus, 85 can never reach 90 except through the lottery in Step 2.

For a Good player, that moves the worst possible PA range from 110–140 up to 130–160.

One English club in my save had exactly this: youth score 85 and staff score 12. 11 of its 16 Elite and Good rolls reached 90+ before the nudge.
  • A great HoYD barely moves the staff score, and a poor one is worse than none. With Villa's other staff.
HoYD (JPP/WwY)Staff score
20 / 20(80 + 61) ÷ 10 = 14
18 / 19 (Villa's)13
No HoYD61 ÷ 6 = 10
1 / 1(4 + 61) ÷ 10 = 6
  • Each other youth staff member pulls the score towards his own JPP, so staff with poor JPP drag it down.
  • The preview's practice players get this bonus too.
Clubs in view-only or unloaded nations (nations whose leagues aren't playable in your save) use junior coaching instead of the staff score. If the club is a reserve or affiliate side, its parent club's junior coaching is used. That figure is cut so that youth score + junior coaching is at most 70. The cap and bonus then work as above.

Examples.
  • Youth score 54, junior coaching 12: cap is the smaller of 12 and 54 ÷ 5 = 10, so the bonus is 0–9. I saw a US club go 54 → 62.
  • Youth score 60, junior coaching 14: junior coaching is cut to 70 − 60 = 10, so the bonus is 0–9.
  • Youth score 85: junior coaching is cut to 0, so no bonus. I saw this at a French club.
 
Last edited:
Part 8: What this means for you

The practical upshot: what to build and who to hire


Facilities do the heavy lifting. Staff add a little on top. In order of impact:

1. Facilities
  • Junior coaching is king.It's the biggest single factor in the youth score, worth about 5 youth score per point in England. It also:
    • gates the promotion rolls;
    • sets how high the quota can lift players. 15+ unlocks lifts all the way to 180+; 12–14 stops at 160–179;
    • counts four times in the CA score.

      If you can only upgrade one thing, upgrade this.
  • Youth recruitment matters more than it looks.It drives:
    • the top-flight lift for Elite and Good players;
    • the jump straight to 130–160 (5.6% per eligible player at 17, 8.9% at 20);
    • the boost for foreign-born youngsters.
  • Youth facilities help, but much less. They add about 1.5 youth score per 5 points, and a little to CA.
  • Training facilities don't affect the intake at all.
2. Your Head of Youth Development: what to look for
  • Judging Player Potential (JPP) and Working with Youngsters (WwY), equally. These set his part of the youth-staff bonus (Part 4, Step 1b). Each counts twice. On his own, his staff score is (JPP + WwY) ÷ 2, rounded down.
Examples.
  • JPP 18, WwY 19: 2 × (18 + 19) ÷ 4 = 18.5, so 18.
  • JPP 20, WwY 20: 2 × 40 ÷ 4 = 20.
  • JPP 10, WwY 12: 2 × 22 ÷ 4 = 11.
  • JPP of 12 or more also gives you the best preview the game allows for 15- and 16-year-olds (Part 3, Step 2). Above 12 it adds nothing more to the preview, but it still counts in the youth-staff bonus.
Example. For a 16-year-old with PA 158 and projection 137, a JPP 12+ HoYD rates him 147. A JPP 5 HoYD rates him 140–143.
  • Judging Player Ability (JPA) of 16 or more if you want exact stars and an exact player count (Part 3, Step 4). 6–15 can be half a star out, and 1–5 a whole star. JPA doesn't affect the position grades or the real players.
  • His personality. 1 time in 7, six of each new player's hidden personality attributes are pulled three-quarters of the way towards his own.
Example. His attribute 12 and the player's 4: (3 × 12 + 4) ÷ 4 = 10.
  • Nothing else about him changes who arrives.He doesn't choose their nationalities or positions.
    • Nationality comes from your club, nation and region. A board-vision objective to play or sign players from a particular nation can swap some players to that nation.
    • Positions are steered by how many players your squad already has in each position group, plus random rolls. With one goalkeeper or none, the intake leans heavily towards a keeper.
3. Your youth-team coaches: fewer and better beats more
  • Only four jobs count: Coach, Goalkeeping Coach, Fitness Coach and Set Piece Coach. They must be on a youth team, which in England means U18 to U20. Only their JPP matters.
  • The staff score is an average, so more staff doesn't help. Each counted coach pulls the score towards his own JPP. He raises it only if his JPP is above the current average, and lowers it if it's below.
Villa example.
  • HoYD alone (JPP 18, WwY 19): 18.
  • Add the six U18 coaches (JPP 11, 7, 12, 11, 11, 9, total 61): (74 + 61) ÷ 10 = 13.5, so 13. Every one of them is below 18, so each one lowers the score.
  • Without just the JPP 7 coach: (74 + 54) ÷ 9 = 14.2, so 14.
  • When hiring a youth-team coach for the intake, look for JPP at or above your HoYD's (JPP + WwY) ÷ 2. Anyone lower costs you a little bonus.
  • Staff attached to the U21s or the first team don't count either way, nor do staff in any other job.
  • Don't overdo it. This covers the intake only. Your youth-team coaches also train your youngsters, which I haven't looked at. The bonus is small, so don't let a good trainer go just to lift it.
4. Who doesn't matter for the intake

Your Manager, Assistant Manager, Director of Football, Technical Director, Head of Player Development, scouts, analysts, physios, doctors and sports scientists play no part in any of the intake steps I traced.

What it's worth
  • The youth-staff bonus only applies to Elite and Good players, at clubs in the top two divisions of a leading football nation (Part 4, Step 1b).
  • There's no point in a staff score above your youth score ÷ 5 (rounded down). The bonus uses whichever is smaller. Villa's youth score of 75 makes that 15.
  • It only changes PA odds when it carries a player over a band edge: 55, 75 or 90, or 100 for Elite and Good players in the top band.
Example: youth score 85 (the English club from Step 1b). These chances are before Step 2's lucky boosts.

| Staff score | Bonus | Reaches 90+ | Reaches 100+ |
| --- | --- | --- | --- |
| 12 | 0–11 | 26 of 48 = 54% | never (max 85 + 11 + 1 = 97) |
| 17 (e.g. a 17/17 HoYD alone) | 0–16 | 46 of 68 = 68% | 6 of 68 = 9% |

At 100+, a Good player's chance of 170–200 doubles from 5% to 10%.

Example: Villa, youth score 75.
  • With a staff score of 13, the best case is 75 + 12 + 1 = 88. No chance of 90+.
  • At 15 or more, the bonus is 0–14, and 90+ needs a bonus of 14 and a +1 nudge: 1 in 15 × 1 in 4 = 1 in 60 per Elite or Good player.
  • Junior coaching does far more. Each extra point adds about 5 to the youth score for every player, whatever their tier.
 
Last edited:
Next, a little HoYD noise is added. It depends on the HoYD's judging ability (1–20)

Presumably that's the judging player potential rating for the HoYD? (i.e. not judging player ability)
Both. It's complicated. I've rewritten the section on how players are evaluated:

Part 3: The preview itself

Step 2: Judge each practice player


The HoYD gives each practice player a rating R on the same 0–200 scale as CA and PA. R is not* simply the player's PA. It's a blend of two numbers:
  • the practice player's true PA;
  • a projection: a guess at how good he'll become, worked out from his current ability and age alone (explained below).
How much of the blend is PA depends on the HoYD's Judging Player Potential (JPP) and the player's age:

Code:
w = HoYD's JPP (1–20), nudged by between −2 and +1, then capped by the player's age
R = (w × PA + (20 − w) × projection) ÷ 20, rounded down

The nudge is fixed for each HoYD and player pair. It comes from internal ID numbers, so think of it as random.

Player's ageMost weight PA can getShare of R that's PAJPP that always reaches it
15 or under735%9
161050%12
171260%14
181470%16
191780%19
20 or over20Up to 100%none (20 ca be nudged to 18)

So even a perfect HoYD can only "see" 35% of a 15-year-old's PA. The other 65% of the rating comes from the projection, which is built from his CA. The same ID quirk means about 1 player in 100 gets the most weight his age allows whatever the HoYD's JPP, and about 1 in 100 gets a weight of just 1.

The projection. The game assumes the player keeps improving until he's 28. He improves faster when he's young and when his CA is already high:

Code:
yearly growth  = CA × ((40 − age) ÷ 20)⁴ ÷ 25, but at least 3
raw projection = CA + (28 − age) × yearly growth, rounded down

The raw projection is then trimmed if it's above your league level:

league level = 8 × (league reputation ÷ 10, rounded down) + 40

League reputation is an internal number; the Premier League's was 185 in my save, so its level is 8 × 18 + 40 = 184. For each player, the game then lowers that level by a random 1 to 5 steps, each step being (level − 40) ÷ 20. For the Premier League a step is 7.2, so the lowered level is 176, 169, 162, 155 or 148.

Anything above the lowered level is only partly kept: 85% of it if the lowered level is 140 or more, 55% if it's 120–139, and 35% if it's below 120. The projection is never below the player's CA and never above 200.

Examples (all real practice players from Villa's preview; the HoYD's JPP was 18, so w was 7 for 15-year-olds and 10 for 16-year-olds):

  • Age 15, CA 53, PA 121. Yearly growth = 53 × (25 ÷ 20)⁴ ÷ 25 = 53 × 2.44 ÷ 25 = 5.18. Over 13 years that's 67, so the projection is 53 + 67 = 120. That's under the league level, so it isn't trimmed. R = (7 × 121 + 13 × 120) ÷ 20 = 2407 ÷ 20 = 120.
  • Age 15, CA 87, PA 159. Yearly growth = 87 × 2.44 ÷ 25 = 8.50, × 13 = 110, so the raw projection is 197. His lowered level was 169: he's 28 over, and 85% of 28 = 23.8, so the projection is 169 + 23.8 = 192. R = (7 × 159 + 13 × 192) ÷ 20 = 180, comfortably "good".
  • Age 15, CA 48, PA 142. Growth 4.69 × 13 = 60, so the projection is 108. R = (7 × 142 + 13 × 108) ÷ 20 = 119. A 142-PA player is rated like a 119 because his CA is low.
  • Age 15, CA 50, PA 56. The projection is 113, so R = (7 × 56 + 13 × 113) ÷ 20 = 93. A 56-PA player is rated like a 93.
  • Age 16, CA 69, PA 158. Yearly growth = 69 × (24 ÷ 20)⁴ ÷ 25 = 5.72, × 12 = 68, so the projection is 137. R = (10 × 158 + 10 × 137) ÷ 20 = 147. A HoYD with JPP 5 would use a weight of 3 to 6 instead, giving R = 140 to 143.
Each player is then labelled:
  • Good if R > 159, or if R ≥ B × g, where g = 1.4 − 0.4 × (B − 50) / 110, kept between 1.0 and 1.4.
  • Poor if R < B × p, where p = 1.0 − 0.2 × (B − 50) / 110, kept between 0.8 and 1.0.
  • Average otherwise.
Example. With a benchmark of B = 120:
  • g = 1.4 − 0.4 × 70/110 ≈ 1.15, so a player needs R ≥ 138 to be "good".
  • p = 1.0 − 0.2 × 70/110 ≈ 0.87, so R < 105 is "poor".
With a weaker squad (B = 80), the bars are 1.29 × 80 ≈ 103 for "good" and 0.95 × 80 = 76 for "poor". In proportion to the squad, the bar for "good" is higher at a smaller club. In absolute terms it's far lower.
 
  • Personalities have a 1 in 7 chance of having personal attributes pulled towards the HoYD's (3 * HoYD's attribute + player's attribute) / 4
  • Your players can get a bonus from other members of staff. This is something I didn't realise last night and it invalidates what I said about the HoYD not influencing players, only scouting them.
Interesting, so HoYD/youth coach JPP and HoYD WwY = 2.5x effect of Youth Facilities on newgen PA, if I've read it correctly. So quality staff are a better investment than youth facilities, even for newgen PA.

Surprising that U21 staff and U18 manager don't count for it.

And it in fact is roughly 2 in 16 players that get the HoYD's personality. And I was thinking on past experience that it was around 3/4 of HoYD's personality, because I remember it was quite obvious who had been influenced.

It would upset players, so it's not generally feasible, but I wonder if you sent your first team squad to the reserves and promoted players with positions you didn't want, on intake day, would you get positions you want? I guess its method for picking the squad probably looks at the best at the club though.
 
Part 6b: Intake day — the attribute spread (and the HoYD's coaching style)

Step 1: Every player gets an archetype


The game holds a list of player archetypes (I'll call them templates). Each template has:
  • a position group it's for (e.g. central defender or winger);
  • a role: the role the real players it was built from suited best in that position;
  • a height band: under 177 cm, 177–185 cm, or 186 cm and over;
  • a weight: the number of real players it was built from, which is how common it is;
  • a region: either one nation or one continent;
  • a multiplier for each of the player's attributes, e.g. ×1.3 Pace and ×0.6 Heading for a quick, small winger.
A new player can only get a template that matches one of his natural positions and his height band. The region rules are in Step 4.

Step 1b: Where the templates come from

There's no fixed list in the game files. The game builds the templates when you start a new game, from the real players in your database, and stores them in your save. So two saves with different databases, or with different leagues loaded, have different templates.

1. Which regions get templates. Every nation with a playable league in your save gets its own set, built from players of that nationality. Every other nation (view-only or not loaded) is pooled by continent: one set for each continent, built from the players of all those nations together.

Example. You load England and Spain as playable, and nothing else. England and Spain each get their own templates. Scottish, French and Irish players all go into the shared Europe set. Brazilians and Argentinians go into the South America set.

2. Which players count. A player is used if:

- he's 20 or over (I'm fairly sure this check is his age);
- he isn't a newgen himself (normally there are none yet when the game is set up);
- his CA is at least a minimum for his nation. The minimum is 1.25 × a nation figure I haven't identified, rounded down. It is raised to a floor that depends on the nation's reputation (my reading of the field): 10 for the lowest, then 25, 30, 40, and 50 for the top nations.

Example. If a nation's minimum works out at 50, a 24-year-old with CA 48 is left out, and a 31-year-old with CA 120 is used.

3. Each player becomes a "shape". The game works out a level from his CA: CA ÷ 20 + 6, rounded down. Each attribute is divided by that level, then raised to the power 1.2. That stretches it a little: values above 1 go up and values below 1 go down. Finally it's kept within limits that depend on how much the attribute matters for his best role in that position:

How much the attribute mattersLimits
One of the role's key attributesat least ×0.85, no upper limit
One of the role's preferable attributesbetween ×0.7 and ×1.25
Matters for the position, but not for the rolebetween ×0.5 and ×1.1
Doesn't matter for the position at allno limits

("Key" and "preferable" are my reading of the game's two levels of role importance. I believe they're the ones the role screen highlights, but I haven't confirmed it.)

Example. A winger with CA 140. His level is 140 ÷ 20 + 6 = 13.

  • Pace 17: 17 ÷ 13 = 1.31. To the power 1.2, that's ×1.38. If Pace is a key attribute for his role, it stays ×1.38. If it's preferable, it's cut to ×1.25. If it only matters for the position, it's cut to ×1.1.
  • Heading 8: 8 ÷ 13 = 0.62. To the power 1.2, that's ×0.56. As a key attribute it would be raised to ×0.85, and as a preferable one to ×0.7. Otherwise it stays ×0.56.

A player who is natural in several positions (except sweeper) gives one shape for each, using his best role in each position.

4. Similar players are merged. For each shape, the game looks for an existing template with the same position group, role, height band and region. It scores how different they are. For each attribute that matters for the position, it puts both values into bands relative to the average shape of that position group in the region. The bands are: up to half the average, up to 0.7×, up to 0.9×, under 1.1×, under 1.3×, under 1.5×, and above. If the two values are two or more bands apart, that adds 1 point, or 2 points for the position's most important attributes. If the closest template scores 19 or less, the player is merged into it: each multiplier becomes the running average, and the weight goes up by 1. If not, he starts a new template with weight 1.

Example. The group average for Pace is ×1.0, and a template's Pace is ×1.25 (the "1.1 to 1.3" band). A new player with ×1.45 is in the next band up, so no point. A player with ×0.85 is two bands down, so 1 point (or 2 if Pace is very important for the position).

If he's merged into a template built from 3 players with Pace ×1.20, its Pace becomes (1.20 × 3 + 1.45) ÷ 4 = ×1.26, and its weight becomes 4.

5. Rare templates are thrown away. Once a region is built, any template from 4 or fewer players, or from under 3% of the players in its position group in that region, is deleted.

Example. A nation's central-defender templates were built from 400 players in total. 3% of 400 is 12. A template from 11 players is deleted, and one from 12 is kept.

6. The list is sorted. First by region (all the continent sets, then the nations), then by position group, then by role. Within the same role, the template with the most players comes first. The order matters in Step 4.

There's also a routine that tops the list up later: it adds templates for any nation that has become playable and has none yet, and rebuilds every continent set. I haven't worked out what triggers it in-game. My guess is adding leagues to a running save, but that's unconfirmed.
 
Back
Top