[FM17] [Experiment] What drives Youth production?

[Jimbokav1971]

Introduction.

For a few issues of the game now, my saves have generally been more about the development of a club and individual players within that club rather than about winning this trophy or that trophy, (although a little success along the way is always nice too). 

I usually play Full Fat FM with a skin that has an Instant Result button, and the only highlights I look at are usually goals that we score, (unless things are going particularly poorly and I need to do something to avoid getting sacked). My goal is just to keep developing the Youth players produced by the club, and hope that their continued development will see the club progress over an extended period of time. Some of you are probably thinking "where on Earth is the fun in that?", but that's just how I like to play and I make absolutely no apologies for it. 

So I started off playing with San Marino (club), in FM14 and it was in this save that the Youth development bug really bit me. Forget winning the Champions League or the World Cup. The best Day each season was Youth Intake Day and all the others days in the year were only about getting 1 day nearer to Youth Intake Day. Since San Marino, I have played Youth Development saves in Gibraltar, (went very well), Germany, (second season sacking syndrome), Serbia, (Partizan rather than Red Star thank you very much), and I am now managing a little club called HK in Iceland in FM17.

Something that I have often wondered is...... what actually determines how good my Youth Players are going to be?

We all know that it's something to do with the HoYD, and the Junior Coaching, and the Youth Recruitment, and the Nation you are managing in, but how much do each of the above affect things and in what way? How important are they? Also, does it matter what Country you are managing in and what if you have great Facilities but a poor reputation? Going further than that, how might your Intake be affected if you had great Facilities and a good reputation, (because you were successful), but you were playing in the Footballing back-water of Outer Mongolia, (or Greenland was the analogy that I remember first reading about this)? The truth is I just don't know, and it seems that nobody else really does either.

What seems to happen quite a bit in the FM community, is that if something is posted or said in a blog or vlog and is well-presented and articulated, then it is just taken as the gospel truth by loads of readers/viewers, irrespective of the truth behind the statement or claims. Then it is just spouted verbatim as if it is the truth, the whole truth, and nothing but the truth, (and the only truth at that!) lol and it's a bit of a self-perpetuating myth. It's true because so many people think it's true and the reason that so many people think it's true is because so many people say that it's true. It obviously must be true!

The reality is that blogs and vlogs and threads in here are absolutely nothing more than a presentation of an opinion held by one or more people on a particular subject. It could very well be that the opinion is bang on the money and is great, but it could also be that the opinion is a load of old baloney and isn't worth the online paper it isn't written on. (That obviously goes for this thread too by the way). The difference here is that I am going to try and tell you what I think will happen before each mini experiment and then we will actually see the results of the experiment, and then I will offer my own conclusions as to the results of the mini-experiment, but feel free to draw your own conclusions and even disagree with me. If this interests you then just look at the evidence and decide for yourself. 

I've done relatively well with youth intakes over the last few versions of the game, so FM players that I deem as quite experienced have often come to me and asked, "how do you get such good intakes?" The reality is that I don't think I do really, it's just that they look better in comparison to the rest of my squad because I am never going out and buying the next star.

What I have been looking for is a way to do an experiment like this, but at the same time not ruin my own game. I will be honest with you and admit that I'm far too busy, (lazy you say?), to run a blog about FM like so many others do, (I really just can't be bothered), but I love playing the game so I needed something that wouldn't take away from my own HK save in Iceland, because with respect to those of you reading this, if the enjoyment of playing my own save wanes, I'm certainly not likely to be continuing a bloomin FM experiment. The way my mind works though, (and this thread is far more for me than anyone who might read it so apologies for being selfish), while it might be ok for me to do things in the game that manage to generate what I want in terms of end result, I actually want to understand it. I don't mean have a vague notion of little bits and pieces of the game as is the case now. I mean I want to know how every aspect works and how it impacts on the particular way I now like to play the game.

(Ok. Maybe not). 

 

[Jimbokav1971]

How am I going to run these mini-experiments?

For a while now I have been thinking of how I would run an set of experiments like this, because there is no 1 single experiment that is going to answer all the questions I have, or rather there is, but it would take so long to edit a database that i would be bored, and of course there is the reality that many of these experiments may give more questions than answers. I wanted something that would be able to be run alongside my own game, so that I could easily just swap between the two, but which also allowed me the ability to get pretty clear and concise answers while at the same time not be too involved in terms of using the editor and spending tons of hours playing the game without playing the game if you know what I mean.

 
I was lucky this season in that FMT (that's Football Manager Touch), has actually been released as a stand-alone game for the first time, (I think), so I can actually have both FM17 and FMT17 running simultaneously on the same system. It may not work well for that to actually happen too often, (knowing my luck my laptop will blow up), but more likely than running them both together I will play a bit of one and then a bit of the other when I get bored with one or the other. FM17 will be me playing and FMT17 will be me experimenting.

I downloaded the editor for the first time, (I haven't used the editor since the forgotten dark ages of CM something or other), and I will admit that I just couldn't work out how to use it. There were help video's and guides for all this complicated stuff, but I initially just wanted to see CA and PA of players and I couldn't work out how to even do that so I had to look for an alternative. Honestly, I couldn't work out how to do even the most basic things so.....

 

https://www.fmrte.com/

I remember looking at FMRTE back in the day, (I felt unclean and needed a good wash afterwards), and I knew that this would allow me to track CA and PA easily, even if the actual editing might not now be possible. Well it turns out that FMRTE stands for Football Manager Real Time Editor, (no, no I really didn't know that and I refer you to the earlier graphic). I am really not a clever man. 

This was great though. Within a couple of minutes I was able to navigate around a little, (ok ok ok I was still navigating using the FMT17 programme and then pressing the "current FM screen" button, but it works so who cares? Right?)

Now I can see CA's and PA's for the team I have loaded and what's more it seems amazingly easy to edit the basic things that I think I might want to edit to run these mini-experiments. I can instantly edit club facilities and the like and it seems basic and straight-forward. My initial idea had been that I would have to manage different types of clubs in different Nations to get a reasonable return of info from the experiments, but the black and whiteness, (is that a thing?), of being able to use maximum and minimum facility levels now seems just brilliant. It's also going to make it much easier to tweak HoYD and other staff levels, (again to max and min levels), to make it easier to check those things. 

(The options that this has given me now makes my head spin in terms of the huge number of experiments that I could run). It also means that I don't have to worry about being sacked, (yes it is something that I have to worry about and my therapist says that I have done amazingly well to get to the stage that I am now able to spell out the word R.E.G.E.N.S.B.U.R.G. and he insists that one day I may even be able to say it out loud to actual other people).

If I get sacked when I'm on holiday, (or rather if Holiday Manager gets sacked), I can just create and new profile and take over the same job again. Easy!

[Jimbokav1971]

My initial plan was to use FMT17 and the official editor, but my lack of understanding of the Editor, and a lack of certain features and the limitations of others within FMT17, have made me have a re-think and as a result I will be using full-fat FM17 and FMRTE.

Serbia. Background.

For my experiment subject, I have opted for Serbia and Partizan. I think they are a good newgen producing Nation out of the box, without being the best or the worst. My previous experience as Partizan manager suggests that they will be a good initial Test Nation/club with which to compare future Nations/Clubs with, (or more realistically play about with the parameters).

First we need to look at some of the things that I think, (before we start running any tests), might have an impact on the quality of future Youth Intakes. Just for the record, I don't think that Youth Facilities will have any impact on this at all, but I am trying to come into this with an open mind so it would be silly not to double-check because it wouldn't be the 1st time I was wrong in FM.

Club Partizan.
Nation Serbia.
Youth Recruitment 20.
Junior Coaching 16.
Youth Facilities 16.
Status Professional. 
Reputation 6100.

That's not all that has an impact though is it? What about the staff?

HoYD JPA.
HoYD JPP.
HoYD WWY.

Now just for the record, it is assumed that the HoYD is responsible for Youth Intakes, but actually that isn't a given in FM17. We can choose for someone else to be given that responsibility, so I think it's reasonable to expect that the AI can do the same. To remove any ambiguity here, I have given the responsibility to the HoYD.
Before we get to any tests, we have to look at what level of player they have produced previously. I'm not going to mention real players by name because I don't want to spill the beans on the PA of real players, (and it doesn't matter in the slightest who is whom), but it will at least give us an overview. 
These are the highest rated players, (by value), that have been produced by the club and are now playing elsewhere. (Please note that they may well have produced other players with higher PA, but these players are now not valued as much financially and that's how I chose to sort them). 

PA 165
PA 163
PA 160
PA 155
PA 150
PA 141
PA 135
PA 133
PA 133
PA 122

The average of these 10 most valuable players is 145.7
These are in no way the best players that they have produced, they are simply the highest 10 by value.
Now we have to look at who is still at Partizan and see what quality they have, (and obviously I have to check that they are products of the academy. (Again, I'm just going to pick the top 10).

PA 167
PA 165
PA 152
PA 150
PA 146
PA 145
PA 139
PA 137
PA 137
PA 137

The average of these home produced players still at club is 147.5
Now don't think for a minute that this is the average PA of players that will be produced. No way. This is the average of the 10 best that have been produced. It can in no way be used as a comparison against future Youth Intakes. What it can be however is a snap shot of the top end of player that has been produced over a 10-15 or so year period. (Longer than that and the player valuation will start to decline. 
So that gives us a reasonable idea of the best players that have been produced by the Partizan Academy in recent years and also the best players who have been produced and are still at the club.

So what I'm going to do now is holiday until the day before Youth Day as I call it. (I don't know exactly when that is because they are variable now, but I will make regular saves until I find it and then go backwards to the day before the Intake and set a save point there). 
 

[Jimbokav1971]

So the next stage is to run 10 lots of Youth Intakes with the original settings, to see what sort of PA the generated players would have. 

The following table shows 10 lots of PA's. Each column represents a different Youth Intake and they are sorted by PA with the highest at the top.

In fact, I thin I can do better than that.

Here is the same table again, but with a splash of colour. 

Partizan/Serbia with no changes made.

I have also added the cumulative figure for that intake at the bottom, and also divided that cumulative figure by 16 to give an average PA for each intake.

In this instance the cumulative figures range from 1516 to 1750 and the averages range from 94.75 to 109.37.

It's easier to use the cumulative figures and/or averages when comparing figures further down the line rather than trying to compare 2 lots of 16 different numbers. 

[Thebaker]

Can you pinpoint why 3rd season was so good. I have often wondered if form and league position in the months before the intake day is significant.

Also i have noticed that in lower leagues (playing in dafuges challenge etc) i seem to get better players through than when higher up the leagues (judging by the coaches report of their potential ability in terms of league). This is in spite of higher rep and youth recruitment!

[Jimbokav1971]

1 minute ago, Thebaker said:

Can you pinpoint why 3rd season was so good. I have often wondered if form and league position in the months before the intake day is significant.

The simple answer is that there is no reason why 1 intake is better than the other. You have to remember that this isn't 10 years intakes. This is 1 years intake done 10 different times. Every variable was the same, apart from whatever random element is involved.

[Jimbokav1971]

4 minutes ago, Thebaker said:

Also i have noticed that in lower leagues (playing in dafuges challenge etc) i seem to get better players through than when higher up the leagues (judging by the coaches report of their potential ability in terms of league). This is in spite of higher rep and youth recruitment!

The perception that you get better intakes when you are low down the pyramid is simply that. Perception. The reality is that because that is your club at t's absolute lowest ebb, any players that you produce at that stage will be rated against the best player at the club at that time. As the standard of you best player improves, so does the level against which your new youth players are being compared with. The same level of Newgens produced year upon year will appear weaker in comparison to the standard of your best players because there will usually be an upwards trend in top end quality. As the level of competition rises as you promote through the leagues, the standard of your own players improve too, (or else you get sacked), but often the standard of your facilities, staff, reputation and other such variables don't improve in line with your playing squad. It's this that gives the appearance that you are getting poorer intakes, when in fact they are nothing of the sort,  

[Bedese]

This is great, Jimbo 

[Fragile]

I have been lurking your other Iceland thread, and I just wanted to say this is all very interesting work, I consider myself a football manager novice, only really playing since 2012, and racking up less than 700 hours across all these games.  I'm only really just reaching competency, but your focus on personality, and youth has me intrigued (and seems to be working very well for my most recent save).  I also really appreciate some of your anecdotes around the idea that 'greening' every team talk doesn't necessarily produce the best results etc, and particularly around the idea that a lot of fm gospel is really just opinion repeated widely.  There is certainly a lot going on 'under the hood' of this game.

Warning: Stats discussion incoming (I'm normally much more fun than this, I assure you)

I do, however, have more than a passing interest in statistics, I am not a statistician, but I work closely with statistics, particularly the interpretation and likelihood of robustness, as opposed to the generation of.  So, fair warning, if you have no real interest in the use of statistics, this next bit might make your eyes glaze over, and I would encourage you to continue with what you are doing, because you are likely to get somewhere with it.  This mostly concerns the appropriate measures of central tendency, or whether or not you should be using a mean or median to compare different intakes.  It essentially comes down to the difference between parametric and non-parametric data, and the appropriate tests to use on each.

This is partially dependent on the way the PA itself is generated within FM, whether it is a rng value between x and x, or whether it generates a random bonus/deficit to a more fixed central value.  I really have no idea whether it is possible to pull the code apart to figure this out, but I'm assuming since we're here at FM scientist level, we can really only look at the results and try and figure out some idea of how it works.

My gut feeling is that a median will be more representative, and allow you to be more comparative over a longer period, particularly as you start to compare different clubs to each other.  Mean should be used on continuous data, not discrete data, and I think PA probably is discrete, although an argument could be made that the PA is a value rounded from something more continuous, and with enough discrete data, it will start to resemble continuous data, so this isn't a hard and fast rule.  The more pressing thing is that the data should be normally distributed (a bell curve) for a mean to accurately represent the 'middle' of the data.

There are statistical checks you can do on data to find out whether or not you have normal distribution, but the most basic of these is just to chart a frequency table and eyeball it, I took your data (of 10 different intakes) pooled it and charted a frequency table in excel, separating the values out into blocks of 5 PA (So 31-35 there were 5, 36-40 3, 41-45 1, etc).  Eyeballing the data it looks considerably skewed and not normally distributed.  For PA below 101, there seems to be a long low tail, then a real spike at more than 101 which slowly tails off to about 130, then has another long low tail out the right of the graph (note these numbers are also influenced by my arbitrary cut offs of every 5, they're not hard and fast numbers).  I think I have succeeded in uploading a graph, the x-axis labels are slightly mislabelled 35 is 31-35, 40 is 36-40 etc. Essentially It looks like a left skewed bell curve that has been arbitrarily moved to the middle of the graph by extending a tail out to the left longer than it should be.

The effect that this will have, is that the average (or mean) will be skewed to a lower point in general than your median, and running these tests on the data for each intake shows this is the situation, only intake 7 and 9 show an average higher than the median, and in both cases only just higher, with 109.4 vs 109, and 105.6 vs 105.5 respectively.  Intake 5 has the greatest difference with the average being 94.8 and the median being 109.  Most years the difference is more like 6ish (eyeballed, not calculated). Essentially your average is going to be overall more affected by the big outlier data in either direction, whereas the median is not going to be fooled by one or two amazing or poor regens.  You might have an interest in the way these outliers are produced, but if you're looking for 'overall quality' in your regens, median is probably the proper comparator to use.

This goes on to have an effect on the way you measure the spread of the data, but I think I'll leave that out of the discussion unless you are actually interested.  If you only have a passing interest  in this, but wouldn't mind, could you look into including a median and a mean so you can see trends within each and decide which is better representative of the statistics you are trying to achieve.  Or keep doing what you're doing, because at the very least, it is an interesting little experiment.

Thanks.

[Jimbokav1971]

7 hours ago, Fragile said:

This goes on to have an effect on the way you measure the spread of the data, but I think I'll leave that out of the discussion unless you are actually interested.  If you only have a passing interest  in this, but wouldn't mind, could you look into including a median and a mean so you can see trends within each and decide which is better representative of the statistics you are trying to achieve.  Or keep doing what you're doing, because at the very least, it is an interesting little experiment.

Thanks.

Absolutely love this post. 

The truth, (unfortunately), is that in my normal working life, I only have a need for the most basic uses of "averages" and I am actually going to have to look up the exact definition of mean, medium and mode, because school was a long long long time ago and I can't remember which one was which. 

The other thing to say is that I want this set of experiments to be easy and don't really want to do something that is going to feel overly bothersome or a chore. 

Having said that..... I initially didn't use an average at all in the 1st chart and only after posting did I think, "well it shows an average for each Intake, but what it doesn't show is an average of the averages, that we could use to more easily compare results". (Not having checked, I think that the average average is.... well I'm not sure actually). 

Mean I think is the most commonly used average. 

Medium is the middle range in a group of listed numbers.

Mode is the value that appears most often. 

Looking at your chart, the mean is what I posted, the medium is... well there are 28 listed numbers ranging from 35-170 and the mid-point is between 100 and 105, so I think that means that the "medium" average is 102.5 and the mode is the most common number so that's 105. The "average average" however is 101.704.

I've just checked and that seems accurate, but surprisingly, when I searched for average definitions, what kept jumping out at me was the suggestion that for some reason, an average of an average is not deemed as accurate, (but I will admit that I can't understand why). It's what I was thinking about adding to future charts, (because it's nice and easy and makes comparisons easy).

The only thing that was stopping me doing that, (before the accuracy issue was raised), is I don't want a + b + c + d + e + f + g + h + i + j = x / 10 to be a definitive answer to what is a better Youth Intake, because it's not as simple as that. I want there to be a value, (or more accurately a set of values), that allow for relatively easy comparison, but I don't want to simplify it to such an extent that I can say xx intake is better than yy intake because the average average, (it seems to be called a "weighted average" but feel free to correct me if I'm wrong), is higher. 

The reason why I don't want to give a more specific/accurate grading of a set of results is I want you the reader to look at the results and come to your own conclusions. We all have different things that determine what makes an intake successful, and depending on the situation within a save, this can change significantly.

To give you an example of this, let's look at the initial graphic again. 

• First off, the 3rd column jumps out for me because of the top-end ratings. It contains the joint highest PA player and also the 3rd highest PA player. If this was my Icelandic save for example, (remembering that I'm not used to looking at PA figures in the game), then it actually wouldn't be good at all. The players would be too good and either wouldn't reach their potential in Iceland and/or would be unhappy that I had tied them down to such long-term contracts and was keeping them at the club against their will. 
• If you look at the 7th column however, then this is the intake with the highest "average" of 109.37. If this was a Serbian intake then I would guess that these players of 100+ PA might not be good enough to make an impact in terms of me winning a European trophy, but if this were an Icelandic save then all of these 100+ PA players would likely have a significant impact on the senior squad, (and in all likelyhood expect to play International football. 

There will be a "tipping-point" that will vary for each save and will even change within a save over time, and this will be the CA around which a player will need to get in order to impact on the senior squad. Obviously we are concentrating on PA rather than CA, but it can't be ignored completely.

[Ceching You Out]

Very interesting experiment here and I'll be earnestly following. If it doesn't add too much work, checking the perceived PA as the youth come through; would add an extra dimension. Could highlight how often the initial PPA differs from the actual PA.

Excellent contribution by @Fragile as well. That information may prove useful at some point! Plus it's right in my wheelhouse .

[Jimbokav1971]

1 hour ago, Ceching You Out said:

 If it doesn't add too much work, checking the perceived PA as the youth come through; would add an extra dimension. Could highlight how often the initial PPA differs from the actual PA.

I think that's a mini-experiment in itself that will be added to the list. 

[Jimbokav1971]

Experiment 1. Youth Recruitment level edited to Level 1. 

The first thing I am going to change is going to be the Youth Recruitment level. I "think", (please note that this is very different to fact), that this plays a big role in the quality of Youth production. 

The current Youth Recruitment level is 20, so it will be interesting to see what 10 lots of 1 Youth Intake look like when it is changed to level 1. What I'm expecting is a significant drop in PA of the players produced, but just because I am expecting it doesn't mean that's what we will actually see. 

Haha. Well it's just as well that I was prepared to be wrong. 

This is the results of 10 versions of 1 Youth Intake with the Youth Recruitment level edited from level 20 to level 1. 

I wasn't expecting this at all. 

We saw 2 players who were above anything that we had produced before at 180 and 175 PA, but at the same time we have seen none in the 160's or 140's, and a distinct reduction of players in the 130's. 

This probably asks more questions rather than providing answers, but my initial thought is that there are different "triggers" within what we are looking at. The very elite top end players are still being produced, (even above the level we have seen previously), but there is a significant reduction in the better quality players in the 120's, 130's and 140's.

This really wasn't what I was expecting. 

[Fragile]

I was actually not sure whether to even bother mentioning it, but when I ran the stats, it seemed too obvious to ignore.  If you continue to post raw data, we can play with it ourselves.  And if you want me to stop derailing, I can go back to lurking.

I totally agree though, it's dependent on what you need for a save, if you're trying to make a youth save, general quality across the whole intake is obviously going to be hugely important, while if you're trying to get a good sale, one amazing player (aka: the outlier ruining all the stats) is your aim.  And if you needed a good DC it probably doesn't matter how great that amr might become....

Not to keep banging the same drum, but just as an example, I have attached a table of medians and means for the first lot of data posted, ordered by median, (because I'm a median believer )

As you can see the medians are much more tightly bunched 103.5-109.5 int he first intake, compared to 99.5-109.5 in the second, your general quality is lower, and across more seasons it is lower.  However, you still have one season that is producing players at almost the same quality as the original (intake 1), so that's very interesting.  Looking at the raw data, it is mainly the solid, not quite world beaters you stop regening.

I guess from a fluff point of view, this is the kid that is naturally gifted, and you're losing the ones who need to be coaxed along with good staff and good facilities, and from a gameplay point of view, it means even if you're managing in the bottom tiers of some out of the way league, you can still look forward to regen day!

[Jimbokav1971]

@FragileJust to be clear, you are in now way de-railing the thread even a little bit. This isn't a thread where I post and everyone else can shut up. It's for opinion and discussion. The only limits that I am imposing are on myself because I dont have the time do do more. Any insight that someone else can offer will be very welcome, (even if I dont happen to agree).

Was anyone else surprised by that last set of results?

[Soviet]

I was always wondering if loading more than one league would affect the regens production. Will the game stick to the same formula or try to boost the nation you're playing in, generating more regens in that country?

[Toffle]

25 minutes ago, Jimbokav1971 said:

Was anyone else surprised by that last set of results?

Definitely.

Something else really surprising is that you got 8 30's with a top youth recruitment vs 3 30's with the poorest youth recruitment possible.

 

[phnompenhandy]

Very interesting experiment. I've been doing something similar for years too, always making my bespoke fantasy lower league team, filling it with youth and focusing on developing my academy intakes. However, I 'tamper' too much, so I don't think it would be of use to anyone else.

But for the record, I create a team with the pre-game editor, populate it with 21 fourteen-year-olds and one ten-year old (my 'son'), all with CA 1. I make my club rep start at 1 or 2 and all facilities are 1; however, I do stock the backroom with a large contingent of unfeasibly good coaches. We're a bit of a 'cult' - all tightly bonded to the club and me, indifferent to the fame or money (ie all players and staff are on c 20 year contracts for peanuts). This year I've put my first team in Level 12 and U18s squad in Level 22 of DanBHTFC's English file (that I've adapted to include European, British and a few teams from Asia and the Caribbean that have personal associations or cool names).

Once the game boots up I use the official IGE and FMRTE to regulate things the game doesn't do (such as making all my manager attributes 1), and GenieScout to monitor stuff. I've been able to ascertain how much CA can rise in my situation with training facilities at 1 before it hits a glass ceiling until the facilities are upgraded. I've noted exactly how many CA points coaches get on return from courses and so on.

So I'll be interested to compare my findings with yours, Jimbokav.

[KUBI]

In general, if you want to see a significant change in the quality of the youth intakes you need to lower or upper also all the training and youth facilities and the junior coaching at a club.

[Jimbokav1971]

23 hours ago, Soviet said:

I was always wondering if loading more than one league would affect the regens production. Will the game stick to the same formula or try to boost the nation you're playing in, generating more regens in that country?

The general idea of investigating what impacts on primary and secondary Nationalites is  indeed something that I plan on looking into at some point in the future, although loading up an additional structure is just 1 of many aspects in this respect.

I really enjoy different Nationalities coming through so this is certainly something I plan on looking at/discussing in the future.

What are the impact of various types of foreign feeder/parent clubs, staff Nationality, local borders etc etc. When we get on to that I will go with the few ideas I have in my head already, and then see where the results/discussion takes us.

[phnompenhandy]

27 minutes ago, KUBI said:

In general, if you want to see a significant change in the quality of the youth intakes you need to lower or upper also all the training and youth facilities and the junior coaching at a club.

Yes, for those that don't know,

Improving Youth Facilities increases the average PA of your academy intake.

Improving Junior Coaching improves the average CA of your academy intake.

However, increases to their CA will hit a glass ceiling depending on the level of your Training Facilities.

I'm currently finding that no matter how much money I being into my semi-pro club, the chairman will only allow a 1-point upgrade to the Training Facilities every two years.

[Jimbokav1971]

5 hours ago, phnompenhandy said:

Yes, for those that don't know,

Improving Youth Facilities increases the average PA of your academy intake.

Improving Junior Coaching improves the average CA of your academy intake.

However, increases to their CA will hit a glass ceiling depending on the level of your Training Facilities.

I'm currently finding that no matter how much money I being into my semi-pro club, the chairman will only allow a 1-point upgrade to the Training Facilities every two years.

It's not surprising that there is contrasting opinion with regards to this. 

I was under the impression that how it worked was this. 

Youth Coaching = CA.

Youth Recruitment = PA

Youth/Training Facilities = Rate of CA development. 

* Obviously there are other factors involved, but that's the basis. 

[Jimbokav1971]

Continuing on from my last post, while that is what I believed, (and still believe to an extent), I am far more of the opinion now that there are a number of contributing factors that come together to form the building blocks for PA in Newgen. 

While Youth/Training Facilities were on my list, they would have been much lower down than other factors such as Reputation, Youth Recruitment & HoYD ability, Now however I am coming to the early conclusion, (always dangerous), that no 1 thing will dominate things completely and instead it's likely to be a mish-mash of contributing factors. Some might be more important than others, but to get the best intakes you will need excellent ratings across the board. 

[phnompenhandy]

41 minutes ago, Jimbokav1971 said:

Continuing on from my last post, while that is what I believed, (and still believe to an extent), I am far more of the opinion now that there are a number of contributing factors that come together to form the building blocks for PA in Newgen. 

While Youth/Training Facilities were on my list, they would have been much lower down than other factors such as Reputation, Youth Recruitment & HoYD ability, Now however I am coming to the early conclusion, (always dangerous), that no 1 thing will dominate things completely and instead it's likely to be a mish-mash of contributing factors. Some might be more important than others, but to get the best intakes you will need excellent ratings across the board. 

Youth/Junior Coaching = same thing.

I think you are correct. What I'd suggest is that if you have some of the factors in place but not all, the result would be that some years you'll get an outstanding prospect or two, and other years nothing. The more factors you have in place, the higher the chance of good prospects coming through.

But some factors will put a ceiling on what you get. For example, your HoYD might be the best in the world and your training complex 20/20, but if you're based in the highlands of Scotland (like my club is) you have a very shallow pool from which to recruit, whereas the opposite situation would apply to London.

 

You might want to hve a gander at this article, by the way

http://www.passion4fm.com/football-manager-training-facilities-levels/

[Jimbokav1971]

@phnompenhandyI don't think I had seen that particular one, but to be honest they are all much of a muchness. 

"Youth recruitment is the decisive factor of the regens/youth candidates level of ability".

This is a statement made in the above link and it's something that I believed to be the case too. I have just reduced the Youth Recruitment level to 1 in the last experiment and in the very 1st run produced a player with a PA of 180. In 9 more "runs" I then produced players of 175, 157, 139, 135, 134 & 132. This would suggest that the above statement is not true and it's the very reason for this thread. 

Just because someone says it, does not make it true. 

[phnompenhandy]

Sure, it's a fan-written article, not gospel. I'm experimenting, and I'm interested in the results of your experiments too.

What club are you managing that can churn out PA180 - Barnet?

[clarkehwbu]

Very interesting read and will be following (mainly commenting to keep track).

I'd also like to give my 'opinion' which is probably wrong.

Junior Coaching should affect the CA of the individuals in the intake. Youth Recruitment should affect PA of the individuals of the intake. Facilities will then ultimately have a factor in the long run in which these individuals progress and don't hit any 'ceilings' in their development and stunts growth.

You'd assume that the highest possible in all of these would result in the best intakes.

You've then got to take into account other factors (below).

Location - London would have a wider range of possible talent, whereas a place like Bournemouth on the south coast, would have smaller range of possible talent. You've then got to think, does the amount of teams in the area matter? You have loads of London clubs with high reputation and will all be fighting over the same players, some clubs will get the best leaving the others with the 'worse' players. Whereas a club like Leicester, are the only big club in it's city. Unless this is based on region (midlands, south etc) then this could be a mitigating factor into this.

HoYD - I believe this individual can affect this not just through his stats, but also through his personality and also his reputation. I believe HoYD reputation would have an effect, due to the fact that in the real world a player would much rather go to a club with David Beckham as the HoYD who has 5 star reputation as opposed to Joe Bloggs who has 0.5 star reputation with the exact same stats as David Beckham (these are just examples).

Club - Again this is another reputation based theory - Manchester United for example have 5 star reputation and then Bolton have 2.5 stars (for example, probably not exact). Both clubs are pretty local to each other and the player then has the choice of which club he would rather be a youth product of. If both clubs are welcoming with open arms, he's obviously going to pick the higher reputation and 'better' team. My other point on the club being a factor probably only relates to real life as I doubt New Gens have a favoured team before being inputted, but what if this was the 'hidden' factor. If a player who is from Manchester has the choice between two teams (Manchester United & Manchester City for this example), and his favourite team is City he is not going to want to play for their rivals and therefore, may override all these other factors? I've not looked into it too much but do New Gens have favoured teams? If so, do any come through the ranks who have favoured teams that are not the club he's already at?

Randomness - There's also got to be a certain aspect of randomness, as I think you've proved above, with your tests, with the same facilities/coaching you are getting some very different results.

Again this is all my opinions and what I believe factors in, i've never done a 'youth only' save but I do like to promote from within in my saves.

[wilddingoes]

Having experimented with the pregame editor before (in FM15), each country has three major different characteristics that influence youth potential and development.

1. Nation's youth rating: a number from 1-200 (bigger the better) that may determine the "quality" of the players genned in from that nation, probably determining the weighting of the range of PA's. A nation with a YR of 100 might be able to produce a player with PA 190, but a nation with a player rating of 150 would have a better chance (other variables being controlled).

2. Nations are also able to determine similar ranges for each personality attribute. You can make a nation more professional (on average) for example. Personality obviously has a huge impact on the rate of development so this can be a massive boost to some countries. 

3. Football's popularity in the country: This represents the % of elite athletes choosing to play football in your country. Having lower popularity probably lowers the amount of good prospects, but it's difficult to know how

I don't have the newest FM, but if you have the pregame editor, load up a database and search for a nation.  Once you get to the nation's info page you should be able to find the relevant information.

I have no idea how important these stats are, and more importantly, if they are static, but the way the editor portrays them is as integral to the genning process.

The only in-game evidence I can point to is the higher PA of some players from crap nations with non crap secondary nationalities.

[rimisark]

Nation youth rating can do miracles, reputation also (club, league, nation). In later phase of your experiment, you could higher this factors and see what will happen.

[mbowen17]

This is very interesting . Hope you have more findings. 

[nie jem frytek]

I did something similar on FM2015 (the last one I own) and I wanted to share my results.

 

What I did?

I wanted to check what effect on Potential Ability of newgens following attributes have: Head of Youth Development Judging Player Ability, Head of Youth Development Judging Player Potential, Head of Youth Development Working With Youngsters, Junior Coaching, Youth Recruitment, Youth Rating.

Initially I wanted to run much more simulations (15625 or maybe even 46875), but that's obviously way too much work. Instead I ran much more workable 180 simulations. Multiply this by 16 newgens we get every intake and you get 2880 newgens. Should be enough. Important thing to mention is that I divided Youth Rating by 10, so it goes inline with other stats.

 

What I got?

For Potential Ability the most important attribute is Junior Coaching then Youth Recruitment then Youth Rating then Working With Youngsters then Judging Player Ability and Judging Player Potential. The first three are much more important than the latter three.

For Current Ability of newgens the most important attributes are respectively: Junior Coaching, Youth Recruitment, Working With Youngsters, JPA, Youth Rating, JPP. Only standout here is Junior Coaching.

Overall, I'm pretty disappointed with the experiment...

Bonus

 

[godzilu]

Sorry to dig this up, but has anyone tried to see if the scouting setup also affects the youth prospects? Checking the results with no scout setup versus a full team of scouts with 20 20 20 PA CA Y? Edit: Also, I think affiliated clubs should be tested too, they should affect the youth rating at an extent too, since you can get players from other national youth pools, so the chances increase.