A Closer Look at Training.

[ron.e]

Lyssien - I'm finding it impossible to find my training day for crystal palace. I've tried putting one player on a different training regime with just ball control for example and just kept checking everyday but nothing happens. Is there another or definate way of finding out your training day? on another save game with blackpool i somehow worked out that my training day for them was thursday and they are also in the championship. Do you know if training days(where arrows change in training) are for same leagues or for countries? Thanks

[SFraser]

Sorry not quite catch what you exactly mean. What's your point in detail regarding these two pictures? Thanks.

Specificly that while a high rate of CA gain should be aimed for it is clearly not a player development "strategy" and makes no account of the actual improvement or development of a player. Under certain ingame circumstances players whose CA increase may "develop" into inferior players.

CA increase is a background mechanic that underlines player development but does not control it. Ingame events may not alter the rate of CA increase by much, if at all, but they have a profound influence upon the bias of CA distribution and are the key factors involved in the development and improvement of a player.

Thus while you may be correct in stating that ingame actions have no impact on CA gain, you are clearly wrong to state that ingame factors have no impact on player development. As a managers influence over Ambition, Determination and Professionalism is limited to mentoring, fines and is generally zero once a player has passed mentoring age then said CA gain is not a "strategy" but a background mechanic whereas short term ingame trends define the shaping and weighting of CA distribution.

The hidden mechanics of natural CA gain have been brought to light, the next stage for this topic is to discover the major ingame factors involved in attribute distribution and the relationships between various attributes and CA in terms of weighting, in order to understand what actions and what behaviours produce what responses and what must be sacrificed to achieve what goals.

Clearly no one wants to improve the CA of Cristiano Ronaldo by 2 and his Natural Fitness by 1 if it comes at the expense of a net 15 point attribute loss to his mental and technical abilities. Thus the key to development (once your model professionals mentor your youngsters) remains understanding ingame behaviours, distribution of attributes and their relative weights.

Lyssien - I'm finding it impossible to find my training day for crystal palace.

Take one player that is happy with his training schedule and increase either his workload or his fitness training above what he is happy with. The day when his training happiness changes is the day your training is updated. It may take you a few weeks to hunt down the specific day of this change.

[catafan]

Specificly that while a high rate of CA gain should be aimed for it is clearly not a player development "strategy" and makes no account of the actual improvement or development of a player. Under certain ingame circumstances players whose CA increase may "develop" into inferior players.

CA increase is a background mechanic that underlines player development but does not control it. Ingame events may not alter the rate of CA increase by much, if at all, but they have a profound influence upon the bias of CA distribution and are the key factors involved in the development and improvement of a player.

Thus while you may be correct in stating that ingame actions have no impact on CA gain, you are clearly wrong to state that ingame factors have no impact on player development.

That depends on what the definition of "development of player" is. I have always been meaning that player development is player CA-increase and attributes-shaping based on the CA-increase in my posts. For me how to achieve a high rate of CA gain IS the "player development strategy", and I believe it is for many other people. If for you "player development" means more than that it is absolutely no problem but please do not regard our different definitions for the phrase "player development" as the same thing.

You can always talk about that something else (although I don't clearly know what it is by now from your posts) is more important than simple CA-increase and attributes-shaping without emphasizing the same phrase with confusing definitions. Otherwise please state clearly that what your "player development" exactly mean. That's very helpful for further discussion (avoiding many meaningless disputes right?). Thank you.

Clearly no one wants to improve the CA of Cristiano Ronaldo by 2 and his Natural Fitness by 1 if it comes at the expense of a net 15 point attribute loss to his mental and technical abilities. Thus the key to development (once your model professionals mentor your youngsters) remains understanding ingame behaviours, distribution of attributes and their relative weights.

First of all the two pictures you've posted are from two different games which does not represent the developing profile of the same player. In fact Cristiano Ronaldo's Natural Fitness shouldn't have increased in your game featuring the right side picture.

Secondly the picture on the right you've posted does show some attributes loss compared to his original state. I don't have your saved game so I can only guess the reason could be one of the following:

(1) it is normal that player's attributes temperorarily drop due to recent poor training or fatigue, but that drop will only exist before the Attributes-CA synchronization comes which happens once a month in the game.

(2) If a player's dualfootedness or position(s) have been modified, the way CA is distributed into attributes is modified accordingly, i.e. if your C.R. has his leftfoot increased or some other position modified, some of his attributes probably have to drop to fit the CA.

(3) The lost attributes are compensated for by some other increased attributes. Please notice that attribute increase may be not visible (e.g. Pace=88 to 92 displays only 18).

Cheers

[SFraser]

I think we should move on from bickering over whether CA increase is the only method of evaluating player development and whether ingame short term trends influence development, as the relationship between ingame events and distribution of attributes cannot be ignored.

Clearly there is no need to argue over definitions as CA gain and attribute distribution are two sides of the same coin. What you have raised regarding CA development is important, what I have raised regarding observations of correlation between ingame activity and attribute distribution is supported by yourself.

There is a clear concensus amongst Training testers that match performances produce an apparent improvement in a player, but I have raised the point previously that specific in-match activities correlate with specific attribute changes. I will state now that it appears I was wrong in stating match performances influence CA gain, but this leaves a correlation between performances and attribute change weighting.

There are 4 clear solutions to the question of correlation between match performances and attribute change. 1) Specific actions in matches weight those attributes more favourably. 2) High performances in specific matches increase the rate of attribute redistribution. 3) The previous 2 are both true. 4) The correlation between match performances and attribute redistribution is false.

Through methodical ingame observation, reading the posts of other training testers, and involving myself in zero high speed "holiday" testing schemes devoted to observing CA increase alone I will rule out option 4. You are welcome to argue this conclusion but unless you explain it sufficiently you will receive no concession of the point.

As I have only observation of an ongoing save rather than data from multiple tests I will state that I cannot say one or another where 1-2-3 are correct, but I must raise the point that tests on this question must involve day by day methodical analysis which makes them difficult to test in the usual means.

One could assume that Training and Match performances are similar in their relevance to attribute distribution; that Match performance and experience is a high intensity, high workload, maximum rated coach training session. It would be interesting to see if season long statistics from matches correlate in way to an increased weighting of the relevant attributes.

However to return to our original problem, CA gain is not an indication of the development of a superior player, merely the indication of a greater limit for combined attribute weights. To make proper use of this limit of attribute weights the manager must know not only the relative weights of each attribute but also the actions that influence their redistribution.

[catafan]

However to return to our original problem, CA gain is not an indication of the development of a superior player, merely the indication of a greater limit for combined attribute weights. To make proper use of this limit of attribute weights the manager must know not only the relative weights of each attribute but also the actions that influence their redistribution.

Right on spot. CA only restricts the sum of weighed attributes, not indicates a good shape of attribute distribution. That's why I think how to shape player attributes with given CA is important too. So let's move onto the attribute shaping part.

In my theory (you can argue about that later of course) appropriate training schedule and good coach is the major way to shape player's attribute distribution (here the attributes are those included in various training categories). Some attributes like Aggression, Bravery, Influence which cannot be trained or tutored, are modified by ingame events although that happens at a relatively low frequency (e.g. crisiticising player for red card could reduce his Aggression). Other attributes like Determination and hidden attributes can be tutored and some of them can grow itself (e.g. Determination grows if Professionalism is high enough). However I have never found any indication that ingame events and on-pitch performances help improve particular attributes (I mean attributes that can be trained or tutored) before hearing you said so. Could you please make this point more detailed with some examples?

Cheers

[RobinGoodey]

SFraser - you said a couple of posts back that fining a player can alter their hidden mental attributes? Which ones? And how do they alter?

Cheers

[James8]

Lyssien - I'm finding it impossible to find my training day for crystal palace. I've tried putting one player on a different training regime with just ball control for example and just kept checking everyday but nothing happens. Is there another or definate way of finding out your training day? on another save game with blackpool i somehow worked out that my training day for them was thursday and they are also in the championship. Do you know if training days(where arrows change in training) are for same leagues or for countries? Thanks

I find for me it is the day my game begins on. I have an EPL save which began on a Tuesday. Tuesday is my training change day. Makes sense as everything is therefore calculated in week intervals from Day 1 of your save start date.

[catafan]

Lyssien - I'm finding it impossible to find my training day for crystal palace. I've tried putting one player on a different training regime with just ball control for example and just kept checking everyday but nothing happens. Is there another or definate way of finding out your training day? on another save game with blackpool i somehow worked out that my training day for them was thursday and they are also in the championship. Do you know if training days(where arrows change in training) are for same leagues or for countries? Thanks

Hi Ron.e, one important thing I want to say is that the training update day that Lyssien has posted is for the update of Player happiness and Training amount which happens every week, but not for the arrow change which happens every month in the so-called Attributes-CA synchronization. And I believe different players or leagues have different update days (for both cases).

Cheers

[ron.e]

Hi Ron.e, one important thing I want to say is that the training update day that Lyssien has posted is for the update of Player happiness and Training amount which happens every week, but not for the arrow change which happens every month in the so-called Attributes-CA synchronization. And I believe different players or leagues have different update days (for both cases).

Cheers

I found out that my training day for crystal palace in the championship is thursday. The way i found out was one of my players happiness went from unhappy to happy and also 2 other players in the same training group, their arrows changed as well so in my game, player happiness, training amount and arrow change are on the same day.

[catafan]

I found out that my training day for crystal palace in the championship is thursday. The way i found out was one of my players happiness went from unhappy to happy and also 2 other players in the same training group, their arrows changed as well so in my game, player happiness, training amount and arrow change are on the same day.

Sorry I may have misunderstood your meaning. If you mean arrows like this

then you are absolutely right about the updating date.

I thought it was the arrow change like this:

Sorry about that.

Cheers

[catafan]

A word to S.Fraser (also to everyone who has read my previous posts), I have to say sorry that I did under-estimate the importance of Workrate in CA-development.

The previous test data that I hold have included a Workrate=16-17 in every set of result. Such a Workrate(16-17) alone does not make any detectable increase in CA but recent tests tell me that if combined with other factors (Professionalism and Ambition) Workrate does play a considerable role in CA-increasing, although less important than Professionalism. More detailed results like the ratios of the relative importance of all these factors to each other will be given later when I finish all the tests (really time-consuming...).

Cheers

[Ampalaea4]

Hello all.

First of all great post!really good work on explaining how training works...

Second thing is that in order to do a succesfull development on a player we must know relative weights of each attribute in his position he plays..

I mean if we have a MC then passing will need more ca development to increase than marking...So even if we have his attacking trainer higher than his defending doesnt mean that he will raise his passing-creativity faster than his marking.

Did I understand this part wrong?

But we dont really know the actual weights of each attribute,cause this is a secret from SI and only researchers know...

So how we can make a succesfull training schedule if we dont know the "expensive" and the "cheap" attributes of each position?

[catafan]

Hello all.

First of all great post!really good work on explaining how training works...

Second thing is that in order to do a succesfull development on a player we must know relative weights of each attribute in his position he plays..

I mean if we have a MC then passing will need more ca development to increase than marking...So even if we have his attacking trainer higher than his defending doesnt mean that he will raise his passing-creativity faster than his marking.

Did I understand this part wrong?

But we dont really know the actual weights of each attribute,cause this is a secret from SI and only researchers know...

So how we can make a succesfull training schedule if we dont know the "expensive" and the "cheap" attributes of each position?

I guess what you could do is to increase the corresponding training amount for the attributes which you want to increase more.

I mean, constant (but no necessarily frequent, once or twice per seaon will do) adjustment to your training schedule.

[PrinceEgypt]

Has anyone got a squad of players and has a seperate training schedule for each and every single player? Im thinking of doing this for my Rushden team. But has anyone tried it and is it effective?

i.e. What are the chances of a defender (young and mid-age) who is yet a long way from reaching his PA and has a tackling of 10, reaching his PA and eventually a tackling attribute of 15+ if he had his own seperate schedule and his tactical training was set to intensive?

[catafan]

What are the chances of a defender (young and mid-age) who is yet a long way from reaching his PA and has a tackling of 10, reaching his PA and eventually a tackling attribute of 15+ if he had his own seperate schedule and his tactical training was set to intensive?

If he can reach his PA eventually (say CA increases by 40 points), I am sure his tackle could be 15+ given an appropriate schedule (maximum defending training I recommand).

[lam]

I doubt the "number-of-clicks theory" proposed here, although I haven't run any tests recently. The reason is that development depends on condition, natural fitness, injury proneness, professionalism etc, directly or indirectly. It also depends on PA-CA difference, on ..................

I agree with Lyssien.

I think the variations in training sliders allow you to tailor the bar chart. As indicated by SFraser (great post by the way)

Bar 1 = Heavy loss

Bar 2 = Moderate loss

Bar 3 = Maintain

Bar 4 = Moderate gain

Bar 5 = Heavy gain

It stands to reason (which is a polite way of saying its untested) that if you are just slightly below Bar 2, then one click in the training slider would see you raise to level with Bar2.

*EDIT* Looking at the bars, I beleive there are 40 different levels within the chart working on the usual game assumption that base/lowest = 1 setting, which in this case would be 2 points - this was confirmed by using a ruler to measure what I considered to be the smallest differential. I appreciate that this is not entirely scientific 8O)

I have two players that I feel have attributes at a level where I wish them to maintian themselves rather than look for gains or losses.

These particular attributes have bar charts that are in a position that looks like 4 points above line 3, which is considered to be the maintain level.

I have adjusted their sliders in relation to each catagory by 2 clicks to the negative.

I am hoping that this reduces the catagory down to a maintain level which in turns free's up some training time elsewhere.

Ill report back when I can.

LAM

LAM

[lam]

Take one player that is happy with his training schedule and increase either his workload or his fitness training above what he is happy with. The day when his training happiness changes is the day your training is updated. It may take you a few weeks to hunt down the specific day of this change.

Do training days change for each player or is it one day for the whole club?

LAM

[Macon]

For me, it's Sunday for everyone (in League Two)

[tcoleman1]

i was wondering if anybody could give me some tips or help on how to create player specific training schedules? e.g. what is the best players to make them for, how you get the best increase in stats from training?

[catafan]

i was wondering if anybody could give me some tips or help on how to create player specific training schedules? e.g. what is the best players to make them for, how you get the best increase in stats from training?

It depends on what kind of attributes-profile do you want for your player.

For example for DC we generally need high Marking/Tackle/Concentration, Heading, Positioning and Jumping as priorities, plus some decent attributes like workrate, strength, decision, teamwork...etc. Therefore normally we give a DC High amount of Aerobic, Defending and Tactic training, Medium amount of Strength and Ballcontrol training, light amount of Attacking and Shooting, and probably zero amount of SetPieces.

But if you have a DC who already has good physical attributes (say 18 in jumping and 15+ in the rest) but lacks in marking and positioning, you can reduce the amount of fitness training to medium and increase the Tactic and Defending training to intense. It's all up to you and what your players look like.

[phnompenhandy]

i was wondering if anybody could give me some tips or help on how to create player specific training schedules? e.g. what is the best players to make them for, how you get the best increase in stats from training?

In addition to catafan's points, I'd stress that age is a significant factor. Players don't develop their physical attributes much after age 23-24; I have schedules that are more physical for U23 players, then move them to a different one that reduce str/aer to maintain those levels and increase the techincal/tactical ones.

[Lyssien]

Do training days change for each player or is it one day for the whole club?

LAM

Whole club. It's Sunday for me in the greek superleague.

I never tested whether it's the same day for all clubs in the same league, same nation or in every different savegame of the same league. Two people have reported that their "update day" is Thursday, so I guess it's not Sunday for everyone.

As I said, you can find your update day by doing a small test. You can do that in two different ways: Either give extremely hard training to your players and check everyday their training happiness to find the day it changes to "unhappy with training workload". Or heavily alter your players' schedules (e.g. reduce everything to zero training and push just one regime to the right extreme) and check everyday to see when their training levels (bars on the left of the training overview screen) change to reflect their new schedule.

[DocSander]

Whole club. It's Sunday for me in the greek superleague.

I never tested whether it's the same day for all clubs in the same league, same nation or in every different savegame of the same league. Two people have reported that their "update day" is Thursday, so I guess it's not Sunday for everyone.

As I said, you can find your update day by doing a small test. You can do that in two different ways: Either give extremely hard training to your players and check everyday their training happiness to find the day it changes to "unhappy with training workload". Or heavily alter your players' schedules (e.g. reduce everything to zero training and push just one regime to the right extreme) and check everyday to see when their training levels (bars on the left of the training overview screen) change to reflect their new schedule.

Or, third way, which keeps most of the player's training schedule intact, spot a player in your squad with the same training level for two different areas (f.e. attacking and shooting). Then change one of these areas and check everyday when a change in levels occurs.

BTW in my game (Holland, Eredivisie) it's Friday. And this is not the first gameday as has been suggested. So to me this day seems not te be determined by whichever day is game day one in your save, but perhaps league/nation/club determined...

[lpnumb]

Excellent thread. It really gives some insights on the training regime.

Reading it and observing the development of my current future potentials I have come to this conclusion. CA gain is based more on match experience and match performances(in some way like in a role-playing game), and then the training schedule is used to shape the CA gain the way the manager likes it. I mean no matter how hard you train a given player if he is not playing any matches the increase in CA ability will not be so significant. Furthermore I've never seen a "balanced" world-class player so the hidden mentality stats are probably modifying the PA(I mean limiting it to some extent). I'm not sure about my conclusion so feel free to comment on it.

One more question about "burning" youngsters. I found a very hot prospect for my Bayern Munchen side(Konstantin Rausch 18 yrs D/WB/M L) from Hannover and signed him for 800k. I decided to keep him in the first team as a backup probably giving him around 10 matches for the first season. But the game decided to s***w me over and Lahm and Massimo Oddo got injured for a month and two months respectively. So I had no choice but playing him in the first team. He played 8 games for 7 weeks and put some decent performances(average rating of 7.10). In the beginning he was improving almost all his stats(even in categories he wasn't training), but now he is getting worse at almost everything(except for Pace and Acceleration). So what's your opinion is he "overburnt" or the initial gain was just from the difference between facilities and coaches.

One more thing. All of the current wonderkids and world-class players in real life football are playing regular first team football since relatively young age(somewhere around 18), so this makes me think that if the player has the right mentality(high Professionalism, Determination, Pressure handling etc.) and high enough Stamina and Natural Fitness to retain his fitness throughout the season he'll be able to start his first team fixtures involvement earlier than in his twenties.

[SFraser]

lpnumb:

There are two possible solutions to your second question regarding your young player replacing your injured first team player. The first possible solution is that your players performances increased the rate of redistribution of his attributes and you are now seeing the losses that correspond to the earlier gains. The other solution is that your players recent perfomances are below his previous performances and the excess gains achieved through high performances are subsequently lost through poor performances. There is a third solution and that is that both the previous two are occuring.

The final point you raised is a question of micro management of condition, and personality versus opponent. A youngster can be played regularly if he is often substituted either on or off, he will attain high ratings if he is started in the right games and subbed at the right times.

[tigerhgrrrrrr]

So I read that 120 total notches for a traing program is OK but surely that is too many? It hikes the overall intensity to Heavy. Can anyone confirm this "120" number, or the correct (I assume) lower number that should be stated?

[SFraser]

Why is "Heavy" too intense? I know that myself and a few other contributors to this thread use maximum intensity training schedules for all players throughout the entire season, lowing the intensity only when it impacts the rate of condition recovery.

[tigerhgrrrrrr]

Why is "Heavy" too intense? I know that myself and a few other contributors to this thread use maximum intensity training schedules for all players throughout the entire season, lowing the intensity only when it impacts the rate of condition recovery.

Oh I thought into heavy was too great a risk of injury, if its OK though that would make sense of 120 notches. Also there is a graph (in another thread?) that shows a severe tailing off of training effectiveness from mid-normal training intensity onwards, but it could be that extra that makes all the difference. Shall give the 120 boost a go though.

[XenoExclusive]

SFraser.

Yea, like tiger refers to, do you find boosting it to maximum that much of a benefit? Taking into consideration the graph?

If yes, although you set it to maximum overall, do you find setting an individual category to maximum detrimental? IE maximum aerobic?

Cheers mate. Btw, fantastic thread!

[catafan]

There is nearly no difference between 109 notches (maximum medium training) and 119 (maximum heavy training) regarding CA-improvement. The value presented in that graph is the average of two parallel data, one of which for 119-notch is actually even lower than that for 109-notch, indicating that the benefit of this slight increase could be well cancelled by random factors. Besides, high training amount brings high injury rate and possible player unhappiness (if you mind). So IMO 109-notch is well enough for the overall training workload.

As for the training amount in an individual category, I don't think a maximum amount will hurt, except that maximum fitness training can gradually increase the hidden attribute Injuryproneness. So it is not recommended to leave the fitness training at maximum load for a very long time (unless you don't mind a high injury rate). I think a-couple-of-year intense fitness training for teenagers who is in the process of fast CA-boosting will do.

[XenoExclusive]

Thanks for the great reply catafan.

But what you are saying is, is that high overall doesn't improve CA. But that's not what training is for no? Training is for shaping (Not sure if its the same as the old -3/+3 thing). Thanks for the info about high fitness training increasing Injury Prone stat, great to know!

So you recommend using about 109, and setting polar catergory work loads? IE a striker should have Zero defence etc, and near maximum attacking and shooting for example?

And as previously stated, if fitness (aerobic in particular) doesn't change too much after 23, would that last notch of heavy be a good idea for fitness training for my U23/24 players? All positions?

Cheers

[catafan]

IE a striker should have Zero defence etc, and near maximum attacking and shooting for example?

Yes something like that. The amount in specific catetory depends on how much Passing/Creativity and Shooting/Composure/Longshots you want for your striker.

My strikers generally have this schedule:

But for some strikers who don't shape their attributes very well, I apply specific schedules for them. One of my striker has good attributes at large but lacks in some Strength and Offtheball, so I give him such a schedule:

You can do the similar thing to your striker.

And as previously stated, if fitness (aerobic in particular) doesn't change too much after 23, would that last notch of heavy be a good idea for fitness training for my U23/24 players? All positions?

Cheers

Sure you can give them heavy aerobic training (there's actually no 'heavy' for that, do you mean 'High'?), or even maximum (intense), depending on how desperately you want a high Pace or Jumping or other aerobic attributes. From my experience aerobic attributes hardly increase after the age of 21, so I don't recommend too much aerobic training for the 22 or older (mid-High the most IMO).

One important thing is that if you can't have your youngsters' CA improved, a training schedule alone won't shape their attributes very fast (i.e. attributes shaping with invariable CA is very, very time consuming).

Cheers

[XenoExclusive]

Beautiful! Thanks for that catafan.

Some great information!!

[SFraser]

I wouldnt read too much into the training analysis by Catafan outside of obvious hidden attribute changes. Match Engine events and the opposition involved have a significant influence over the distribution of specific attributes even though the exact relationships are unknown.

[Aytumious]

I wouldnt read too much into the training analysis by Catafan outside of obvious hidden attribute changes. Match Engine events and the opposition involved have a significant influence over the distribution of specific attributes even though the exact relationships are unknown.

Is there any evidence for match engine events having a significant influence over the distribution beyond a subjective guess?

[SFraser]

Is there any evidence for match engine events having a significant influence over the distribution beyond a subjective guess?

The single largest piece of evidence is the increase of specific attributes that are not weighted through training, both relative attributes that have no workload in training and absolute attributes that cannot be trained.

This post shows that training alone cannot explain attribute distribution weighting as untrained attributes are increasing. This post shows an increase in untrained attributes in one player and a decrease in untrained attributes of another player that has played half as many first team games. Both are young players with significant differences between CA and PA.

The first post shows a striker with previously 18 for Free Kick taking increasing to 19, and the second post shows a Left Back with previously 13 for Long Throws increasing to 14. Neither player increased in both attributes, neither player is training set peices, and both players logically take their respective set peices during a game. 18 Free Kick Striker taking free kicks and 13 Long Throw Left Back taking left side throw-ins.

Obviously there is some function that is favouring attribute redistribution in set peices that is not training. Obviously there is some function favouring particular set peice attributes for particular players, and obviously there is something favouring Free Kicks and Crosses and not Corners for a Free Kick Taking player that does not take Corners and plays on the wing.

Now this kind of data cannot be ignored, although clearly the greatest attempt to do is being made. There are variations of specific attribute distribution rates within training attribute groups that cannot be accounted for by training, preferred position and attribute progression through match exercise but can be easilly explained by including match activity as an additional factor in attribute distribution weighting. The game already calculates data for each attribute involved in each action for each player throughout the course of a match. There is no reason why the relationship between activity and attribute weighting should not exist and indeed outside of the match engine its self it most certainly does exist, and assuming its existence provides a solution for fine discrepancies between specific details of attribute weighting, and corresponds to the observations made by multiple training testers on precisely this issue.

Clearly I have just outlined a logical explanation using existing game mechanics to answer the questions I myself posed rhetorically to draw attention to discrepancies in specific attribute weighting, with evidence to support the claim of discrepancies and influences beyond training and methodical progression of attributes due to CA increase and player preferred positioning alone. What exactly makes you think it is subjective and how precisely would you account for the discrepancies between the gain of untrained set peice attributes shown in my 8 screenshots?

[catafan]

As far as I know, the increase of un-trained attributes is simply because the CA increases significantly (say 4-6 points) on the attributes-CA synchronization, in which case the increase of those trained attributes are not enough to fill up the CA and therefore a global increase for all attributes is applied. You may not see all attributes have green arrows because the increase could be not big enough to ensure an obvious change on player profile page. For example, a change from 88(displayed as 18) to 92(displayed as 18) will not show green arrow on the attribute, but 70(14) to 74(15) will.

This can easily be proved: use an editor to increase a player's CA by 30 points (or more if you like) without changing his attributes, and wait for the attributes-CA synchronization to come (this happens once a month). Eventually on the synchronisation day you will see all attributes increase by nearly the same points (those attributes that have CA-share of course, so Aggression, Determination, Flair and NatureFitness will not change) even if you have given the player ZERO training in every category.

Cheers

[SFraser]

Yes I am aware of those points above. I am talking about ingame features of attribute distribution produced without editors, such as this for two AMR/L under the age of 25 over the course of one year:

You can clearly see that the rate of improvement for each attribute in the same training category varies significantly in an individual player, and that these rates vary from player to player. There is atleast a 10% difference in the ratio of anticipation gains to off-the-ball gains in player 2 in comparison to player 1 and it may even be as high as 100% difference in the ratio between anticipation gains and off the ball gains depending on the actual values triggering the attribute display. Clearly there is some mechanic outside of training that is working to redistribute gains amongst attributes for individual players which is not negligable and may be as influential or indeed more so than the training system.

[catafan]

Yes I am aware of those points above. I am talking about ingame features of attribute distribution produced without editors, such as this for two AMR/L under the age of 25 over the course of one year:

You can clearly see that the rate of improvement for each attribute in the same training category varies significantly in an individual player, and that these rates vary from player to player. There is atleast a 10% difference in the ratio of anticipation gains to off-the-ball gains in player 2 in comparison to player 1 and it may even be as high as 100% difference in the ratio between anticipation gains and off the ball gains depending on the actual values triggering the attribute display. Clearly there is some mechanic outside of training that is working to redistribute gains amongst attributes for individual players which is not negligable and may be as influential or indeed more so than the training system.

Editors are indispensable tools for finding out the underlying mechanism in this game, aren't they?

First of all let's consider training alone. The attributes in the same training category (e.g. Finishing,Composure,Longshot in the Shooting category) are not ensured to increase by the same points on the updating day (the same day as attribute-CA synchronisation). That is, on the synchronisation day, Finishing and Longshot may increase by 1 (ingame value, equal to 0.2 point in the player attributes profile page) whereas Composure remains unchanged. This process can easily be monitored if using FMEditor which displays the precise ingame values for each attributes (1-100) rather than the 1-20 in FMrte. Therefore attributes in the same category, whatever training amount is set for it, do not improve/decrease at exactly the same rate, especially in a short term run like 2-3 years.

Secondly as for the attributes changes between different players, the actual increase also heavily depends on CA-change, not only on training amount in the corresponding category.

The above two facts could well explain why even if you started with two cloned players, 6 months later you get two different attributes profiles. Certainly I cannot exclude the possibility that other events trigger the increase of particular attributes, but for me the phenomena you have presented are not backing up that point.

Cheers

[SFraser]

The two facts would not explain how the ratio of change between two attributes is signficantly different between two players. It would explain that attributes in the same category do not increase at the same observable rate, and it would explain that two different players do not gain the same quantity of increases, but it would not explain why even allowing for maximum values for the smallest observed attribute change and minimum values for the greatest observed attribute change that the ratios of change between both players are significantly different from one player to the next.

In short if you allow for the closest possible sets of values for each observed attribute change in player 1, a minimum of six points in Off The Ball and a maximum of 9 in Anticipation that gives a maximum ratio of 1.5:1 anticipation to off the ball with any other possible values bringing the ratio closer to 1:1 and under.

In the second screenshot the closest value set that would produce the closest ratio is a maximum increase in Off-The-Ball of 4 points in the first 11 months and a minimum increase of 11 in anticipation. This produces a minimum ratio of 11:4 anticipation to off-the-ball or 2.75 anticipation to 1 off the ball, with any other possible values increasing that ratio that 3:1 and beyond.

The smallest possible variation in the ratio of the gain of anticipation to off-the-ball in the two screenshots is almost 50% more anticipation per off-the-ball for player 2 than player 1, using the most conservative possible estimates for major increases and the most liberal estimates for minor increases to artificially assume the smallest difference in ratio. If you use any other possible values to calculate these ratios, the difference will increase. The minimum difference is already 50% between the ratio of gain of the same attributes in the same category in different players. Player 2 is gaining atleast 50% more anticipation per off-the-ball increase than player 1. That is the minimum possible difference displayed in those screenshots.

[catafan]

In short if you allow for the closest possible sets of values for each observed attribute change in player 1, a minimum of six points in Off The Ball and a maximum of 9 in Anticipation that gives a maximum ratio of 1.5:1 anticipation to off the ball with any other possible values bringing the ratio closer to 1:1 and under.

In the second screenshot the closest value set that would produce the closest ratio is a maximum increase in Off-The-Ball of 4 points in the first 11 months and a minimum increase of 11 in anticipation. This produces a minimum ratio of 11:4 anticipation to off-the-ball or 2.75 anticipation to 1 off the ball, with any other possible values increasing that ratio that 3:1 and beyond.

I don't know why you have to compare that two ratios between these two players.

For player-1, in a whole year the Anticipation has increased by 1 (1-9 ingame points) and the Off-the-ball has increased by 2 (6-14 ingame points). That's a very normal situation which I believe everyone could have noticed in their games. More points have been given to Off-the-ball by random factors during the last 12 months.

For player-2, Anticipation has increased by 3 (11-19) and Off-the-ball by 1 (1-9), for me his Off-the-ball is simply unlucky to gain more points in the past 12 months compared to his Anticipation. If you check the last month only, Off-the-ball increased by 1 but Anticipation by zero---maybe luck is back .

The biggest difference between attributes increase in one month I've seen is 4 ingame points, in which case the Stamina has increased from 72 to 76 in one month while Strength remains at 72 (for the same player of course). If that situation happens for 3 times (3 months) the difference between the two attributes could be 12 in-game points, not to mention 12 months.

I still do not exclude the possibility your have raised for attributes shaping, but I think we need more solid evidence for that.

Cheers

[SFraser]

I don't know why you have to compare that two ratios between these two players.

To prove the highly influential discrepancy that is equal to all other factors in determining attribute weights you have refused to acknowledge up to now.

For player-1, in a whole year the Anticipation has increased by 1 (1-9 ingame points) and the Off-the-ball has increased by 2 (6-14 ingame points). That's a very normal situation which I believe everyone could have noticed in their games. More points have been given to Off-the-ball by random factors during the last 12 months.

The biggest difference between attributes increase in one month I've seen is 4 ingame points, in which case the Stamina has increased from 72 to 76 in one month while Strength remains at 72 (for the same player of course). If that situation happens for 3 times (3 months) the difference between the two attributes could be 12 in-game points, not to mention 12 months.

If it is indeed "random factors" and "luck" that are producing a minimum of 50% variation in the ratio of attribute distribution between attributes in the same training category then what is the point of the training system? Why would a minimum of 50% influence on the weight of attributes for distribution be left to "luck" in this game? And particularly if it is "luck" that is producing a minimum of +50% of the weight of attributes for redistribution, and not player action based weighting, then surely every player in the game should slowly develop into an average of all roles and positions irrespective of training distribution.

I don't discount some element of "luck" but I would expect that if the results I have shown were based purely on luck and random factors then the distribution of attributes in certain players like Centre-Backs would have been raised as a bug.

[catafan]

If it is indeed "random factors" and "luck" that are producing a minimum of 50% variation in the ratio of attribute distribution between attributes in the same training category then what is the point of the training system? Why would a minimum of 50% influence on the weight of attributes for distribution be left to "luck" in this game? And particularly if it is "luck" that is producing a minimum of +50% of the weight of attributes for redistribution, and not player action based weighting, then surely every player in the game should slowly develop into an average of all roles and positions irrespective of training distribution.

I don't discount some element of "luck" but I would expect that if the results I have shown were based purely on luck and random factors then the distribution of attributes in certain players like Centre-Backs would have been raised as a bug.

First of all, difference of only one or two points between two attributes is not that huge, is it?

Secondly, if that difference in your player is caused by other non-random factors, the trend should continue, which means 2 or 3 years later your player-1 and player-2 should have much bigger gap between Anticipation and Offtheball, but I doubt it will be that case. The opposite results of their attributes increase (one has Anticipation increase>Offtheball increase but the other has the contrary) seem to back up the random factor theory.

Thirdly, training schedule tends to shape attributes, but does not ensure the precise attributes profile you want. Giving intense amount in Tactic training, more points will be put in the attributes in this category and less will be put in other categories, that is the point of training system. Following your logic, if other events can make that "50%+" biased distribution, what the point of training system then?

As for the attributes distribution in Centre-Backs, what makes you think it is a bug?

Cheers

[SFraser]

First of all, difference of only one or two points between two attributes is not that huge, is it?

The answer to that question is contextual and beside the point. There has been proven to exist significant attribute redistribution weighting that cannot be explained by the training distribution system, or by inherant attribute weights subject to CA increase. This additional influence you ascribe to "luck" and "random" factors which have yet to be explained and which if purely random would undermine the current mechanics of training.

Secondly, if that difference in your player is caused by other non-random factors, the trend should continue, which means 2 or 3 years later your player-1 and player-2 should have much bigger gap between Anticipation and Offtheball, but I doubt it will be that case. The opposite results of their attributes increase (one has Anticipation increase>Offtheball increase but the other has the contrary) seem to back up the random factor theory.

They back up no "random factor" theory as you have not produce a "random factor" theory other than words "random factors" to explain results you cannot explain through editors. The only "theory" that is supported by the data is additional influence on attribute weighting from a source that is not training and is not inherant attribute weights.

Thirdly, training schedule tends to shape attributes, but does not ensure the precise attributes profile you want. Giving intense amount in Tactic training, more points will be put in the attributes in this category and less will be put in other categories, that is the point of training system. Following your logic, if other events can make that "50%+" biased distribution, what the point of training system then?

The answer to that question depends on the logic of the system producing the additional influence. If it is "random" then it undermines the current training system. If it is a causal system that produces logical consequences from data then it is a system of attribute weighting that is additive to the training system but gains its data from elsewhere. The logical source of that data would be the match engine, which would not show up in player profile editor tests and indeed is pretty well hidden.

As for the attributes distribution in Centre-Backs, what makes you think it is a bug?

I do not think it is a bug. I said if the additional influence on attribute weighting for CA distribution was "random" and the calculated minimum variation of +50% in the above post was symptomatic of the influence of "random factors" throughout the game then one would expect to see over time the redistribution of attributes in a highly generic fashion with the majority of players displaying generic and undefined attributes such as equal values for Heading-Flair-Technique-Dribbling-First Touch.

In short a large influence of "randomness" on attribute distribution would ultimately even out most attributes in most players. This would be viewed as a bug. I would be extremely surprised to find any actual mechanics of randomising attribute weighting beyond semantics regarding complexity and relationship.

[catafan]

Originally Posted by SFraser

There has been proven to exist significant attribute redistribution weighting that cannot be explained by the training distribution system

Why is that? I think it can be well explained by random distribution in the training system. As I said before, in one month the difference in increase between two attributes in one category could be as much as 4 ingame points, it is not surprising at all that after 12 months the difference is 12 points or more. For me, these results are completely from randomness.

It seems you don't think random factors can or should make that much difference which for you indicates other factors from the match engine. The problem is, the random factors in the training system do exist in this game, but the match engine issue you have raised still remains as a conjencture. So I said we need solid evidence for that point. IMO only the biased attributes distribution in the same category in your two pictures is not convincing enough to prove your theory (it seem you think it is convincing enough so let other people give their opinions on this).

The answer to that question depends on the logic of the system producing the additional influence. If it is "random" then it undermines the current training system. If it is a causal system that produces logical consequences from data then it is a system of attribute weighting that is additive to the training system but gains its data from elsewhere. The logical source of that data would be the match engine, which would not show up in player profile editor tests and indeed is pretty well hidden.

I don't think that randomness undermines the current training system. A schedule alone cannot and should not ensure a precise attributes shaping. Random factors must be involved to distinguish between the perfect shaping result and the normal, otherwise every player will be shaped into a uniform attributes distribution, which is unrealistic.

I do not exclude the possibility of the existance of your "causal system", but I hope there is more solid prove for that. Otherwise nobody could acctually exploit that "causal system" to build their players, and therefore what you are talking about will not make any sense.

one would expect to see over time the redistribution of attributes in a highly generic fashion with the majority of players displaying generic and undefined attributes such as equal values for Heading-Flair-Technique-Dribbling-First Touch.

Assuming there are only random factors in training system determining attribute distribution, it should not be equal values for Heading-Flair-Technique-Dribbling-First Touch in the end, and it should be equal additional gain for these attributes. e.g. A DC who starts with a Heading=15 and FirstTouch=10 will not end with Heading=FirstTouch=18, it will be more like Heading=18 and FirstTouch=13 plus or minus some error brought by random factors.

I personally have a feeling that DC may naturally gain extra points on his Heading more than other attributes in the same category. But it is nothing more than a personal guess without any solid evidence as backup. I seriously dont recommend people to accept that idea because personal feelings without solid evidence to support are highly likely misleadings in the end, although sometimes it could luckly be true.

Cheers

[vindapop]

having done a search and read through your interesting article, im still confused as to why most of my players have extremely low training levels compared to some who are very high. Surely it cant be just mental attributes??

[SFraser]

having done a search and read through your interesting article, im still confused as to why most of my players have extremely low training levels compared to some who are very high. Surely it cant be just mental attributes??

From what I have seen ingame the differences in training levels are primarilly down to workrate, but that is also subject to other variables I am not entireally sure of.

[DocSander]

OK, I figured out "Rest Days" and injury probability, too.

The game only sees the schedule the player has when it becomes Sunday, and it assumes he is going to have this schedule for the whole following week. Therefore, if a player loses condition abruptly (e.g. because he played a match) we have no way of reducing his workload, because his weekly schedule has been set on Sunday. So the game offers us another way: "Rest days"! So "Rest days" are the way to reduce the schedule we have already set, in order to reduce injury probability when condition is low!

Looking into this 'rest days thing' at the moment...

If I understand you correctly this serves as a way to reduce the probability of an injury occuring during training with a low condition (f.e. post-match).

My question then is, does it also influence the recovery rate of players with low condition levels and/or whose fitness is displayed as "(very) tired after his last match"?

From my own experience this seems not to be the case, or do you (or others) have a different view on this?

[Yonezzz]

OK, I've decided I want to start creating my own individual schedules, but after readin through the OP three or four times I'm struggling to get my head round some of the things, so apolgies, but if someone could clear a few things up for me, I'd be hugely grateful.

- With the unhappiness limits, are strength and aerobic linked together to form an overall 'fitness' limit, or are there two individual limits (be they different or not) that need to be looked for?

- How are the bars on the training levels graph calculated? I mean, I understand that each category is based on relevant intensity of the schedule but altered based on the individual player, but what do the size of the bars mean in relation to the graph lines? If it's above or below the top line, what does this mean? Just that it's more or less than that individual player would 'naturally' expect or be prepared to undertake? If one category is well above or below the top line, is that something to worry about? What would happen if you put all sliders down to zero, would you see no bars at all? Identical bars? Bars that differ slightly from person to person?

- As for the overall training level: as I understand it, based on the player's personal traits, I can tweak his schedule to keep the same overall schedule workload, but this will give a different Overall Training Level, right? And is a low training level better or worse than a high training level? And by high or low, I guess I mean in relation to the top graph line behind the bars. Is that the optimum or maximum?

- For one of my players, aerobics and tactics are on the exact same level on the training levels graph. However when I click those two categories, their yearly progresses differ slightly. They're identical for months 4-12, but aerobic is higher for months 1-3. What does this mean? That he became a 'different' player during month three and consequently the training he did in those months is based upon a different scale?

Sorry for these questions, but as I said, I'm struggling to work get my head round a few of these things so would appreciate a reply.

[SFraser]

- With the unhappiness limits, are strength and aerobic linked together to form an overall 'fitness' limit, or are there two individual limits (be they different or not) that need to be looked for?

If a player thinks that both Strength and Aerobic are too high he will complain about the level of "fitness" training. If he thinks only one is too high he will complain about "strength/aerobic" training.

How are the bars on the training levels graph calculated? I mean, I understand that each category is based on relevant intensity of the schedule but altered based on the individual player, but what do the size of the bars mean in relation to the graph lines? If it's above or below the top line, what does this mean?

This is a tricky one. Based on how I have come to understand training I would assume that they exist simply to provide some kind of reference system for the user and do not infact indicate anything specific other than a means of judging accurate ratios. However untill the exact mechanics of training are known one cannot rule out the possibility that these levels indicate some precise element, function or plateau of the attribute distribution system itself.

As for the overall training level: as I understand it, based on the player's personal traits, I can tweak his schedule to keep the same overall schedule workload, but this will give a different Overall Training Level, right?

Correct. Whereas the Training Level of the Training Category of for example Ball Control is closely dependant upon the Intensity of the Workload in the Category of Ball Control, the Overall Training Level does not so simply depend upon the Overall Training Workload.

Infact the Overall Training Level is a function of Overall Schedule Intensity and the number of Categories being trained. As the number of Categories being trained is reduced the Overall Training Level will drop dramaticly, irrespective of the Overall Workload. The lack of clarity of information presented on the screen leads to the phenomena whereby Goalkeepers with schedules heavilly focused on key Goalkeeping Categories like Tactics, Aerobic and Goalkeeping appear to be significantly underperforming in Overall Training Levels. The key to understanding the display in to think in terms of the attributes each Training Level represents. Training Categories like Set Peices and Attacking have almost no relevance to a Goalkeeper and to a lesser extent Categories like Ball Control are significantly less important than Tactics and Goalkeeping. In terms of Overall Attribute training a Goalkeeper will indeed be doing far less than most other players in the club.

is a low training level better or worse than a high training level? And by high or low, I guess I mean in relation to the top graph line behind the bars. Is that the optimum or maximum?

The way the system works in general is that as a Players Ability increases each attribute in his profile will have a certain bias that determines what percentage of the overall increase is added to that attribute. The Training system allows the manager to increase or decrease the bias of the respective attributes in those Categories.

If we imagine for a moment that the Training system alone determined the entire bias of all attributes and no other factors were involved whatsoever then increasing a single Training Category would add 100% of all Ability Increase directly to the attributes in that Category. However there are many other significant factors that influence the bias of any particular attribute, and in short the training system allows the manager to decide the bias of increase of attributes that the training system allows.

If the training system accounts for 50% of the bias of attributes overall, and other factors account for the other 50% of the bias then the manager can redistribute that 50% bias to whatever Categories he chooses. There is evidence to suggest that the higher the Overall Workload the greater the effect training has on influencing the bias of the attributes being trained.

In short the greater the Training Level the greater the bias of those attributes for added ability points, the greater the Overall Workload the greater the influence of Training on the overall bias. Ofcourse to increase the Overall Workload you must train several different Categories and this will reduce their respective bias in comparison to each other, while increasing the effect of training.

- For one of my players, aerobics and tactics are on the exact same level on the training levels graph. However when I click those two categories, their yearly progresses differ slightly. They're identical for months 4-12, but aerobic is higher for months 1-3. What does this mean? That he became a 'different' player during month three and consequently the training he did in those months is based upon a different scale?

I can give you multiple possible answers to this question because in short I do not know the exact reason. To give an answer that is better based on fact I would require significantly more information on your player.

1. The time of year may affect the Progress of Aerobic training.

2. It is possible that the higher an attribute gets the lower its natural bias becomes, making it increasingly difficult to maximise attributes.

3. His fitness, conditioning and morale may have an impact on his Aerobic/Strength training.

Sorry for these questions, but as I said, I'm struggling to work get my head round a few of these things so would appreciate a reply.

It's no problem. That's why this thread is here. The general principles and mechanics of Training are pretty well ironed out in this thread, but there are many specific details that remain guesswork and hypotheses.