[SFraser's Training Schedules for FM10]

It's been that way for a while, although I'd not realised it was so exact. So attacking on 2 and shooting on 3 should move to 4 and 6 respectively as you drag workload, yes?

Correct

[SFraser's Training Schedules for FM10]

I was using the training module last night and thought I'd post about an observation I made.

If you create a new schedule, set one category slider to notch 4 and another to notch 1, and then drag the Overall Workload slider to the right you will notice that the slider set to 4 will move a further 4 notches before the slider set to 1 moves by an additional notch. This is true for all the category sliders.

I think this is very interesting because this supports the category Unit assumption. The mechanics of Overall Workload, as observable to the user, works to increase the training schedule by the set category ratios.

This may be a known feature, I'd be surprised if SFraser hadn't noticed this when developing this approach to schedule design, but I haven't ever seen it noted in this forum. It's another reason for me at least to feel confident in using this approach and these schedules.

[SFraser's Training Schedules for FM10]

I had hoped that providing enough detail would enable people to design their own schedules and produce them here in this thread for others to download. I am not sure why, but this ambition for a "schedule library" has seriously failed.

DocSander and I are desperately waiting for your feedback on our joint 'project' which is designed to facilitate the training library idea. If you are really serious about it then let's get together and finish it off.

[SFraser's Training Schedules for FM10]

WF = Wing Forward (AMR/L)

CF = Centre Forward (FC)

[SFraser's Training Schedules for FM10]

I've often wondered this too. There is a certain amount of 'slack' when moving the workload slider before the category sliders start to move. There is also an uneven spreading out of the category sliders relative to each other when you increase the workload slider which I have never been able to explain either.

I really don't know :confused:

[SFraser's Training Schedules for FM10]

I've been using this schedule for a couple of seasons. It is amazing!

I have a question about midfielders and passing. My deep-lying playmaker Sandro only has 14 in passing, and I really want to improve that area. Almost every other stats than passing is improving on his Developing CM-schedule. What do you reccomend me to do?

Increase his ATT training by 2 notches. This adds an extra level of Focus as we have described it. Give it ~3 months (without injuries) and check to see if there has been an improvement. If not, repeat this process.

And post your results here please

[SFraser's Training Schedules for FM10]

It's an interesting theory and would explain why the general schedule has all the categories set to the same value (i.e. a "uniform" increase in stats when ignoring age and injuries).

But this doesn't happen in my experience. I will happily stand corrected if anyone can confirm that this is what they see when using the default General schedule.

[SFraser's Training Schedules for FM10]

@ GADONGTA

I'm not sure I understand what you are saying. Could I ask you to re-phrase and maybe include a worked example of how you might formulate a training Baseline on the Workload contributions of each category.

I get confused by Training Levels. I have never seen a clear, unambigous explanation of what the graphs represent. Because of this, to the best of my understanding, SFraser has devised these schedules based on his observations of outputs (attribute changes) from the inputs (category slider notches). The training graphs have been largely ignored as a schedule design tool. Perhaps you could start by giving your interpretation of their meaning.

[SFraser's Training Schedules for FM10]

But what are the underlying evidence that 1 notch is equals to training 1 attibute. It's only still an assumption right?

Correct. It is an assumption and will remain that way until we can see the game code and know for sure what is going on. That is never going to happen mind.

Therefore, there can be no conclusive evidence. Confidence that the assumption is correct comes from predicting and then seeing the desired results. The results that have been reported thus far are all positive and show that the assumption is more or less sound.

The fact that attributes are grouped into categories makes it impossible to target single attributes individually. However, this very fact helps to reinforce the 1 notch per attribute assumption. We know all the attributes are weighted and thus 1 notch per attribute would be errorneous if we had the level of control to affect each attribute individually. But, the very fact that the assumtion is spread over a training category containing multiple attributes means that any errors are smeared and the end result is possibly something very close to 1 notch per attribute. This is a further reason to be confident in using the 1 notch per attribute assumption. Obviously, we can be more confident in this assumption for the Tactics category containing 5 attributes than we can in the Attacking category which only contains 2 attributes. Again, category sizes are outside of our control but the proof, as always, is in the pudding - the results being reported so far are very good.

I'm pleased to see the assumptions being challenged in a constructive manner

[SFraser's Training Schedules for FM10]

What I say from training didn't come from my imagination but from what you can read of the Marc Vaughan hints & Tips dated 2007 when they introduce the training like we can see in FM2010...here is the link to download guide in Word format : http://www.sigames.com/downloads.save.php?id=379&file=471

Jim, is there a 2010 version of this guide?

[SFraser's Training Schedules for FM10]

For myself, i think It's not the training intensity, but the training "reaction" of the player to the intensity (mixed with quality of coach and personality)

Yes, well I think we're saying the same thing. A player's personality and the ability of your coaches are involved in determining the 'response' to the training intensity you set.

Example : I put all sliders to 13 notches that it means a load of training for each area to a possible gain of attributes. How do you explain that after three months of training and match exercise without injury for a player, graph area for defense is at the bottom ? because he didn't react with defense training.

This example is indicative of a poor personality, most likely unprofessional. I don't have too much experience with these player types as I usually try and rid my squad of them asap. I have never seen a situation like the one you describe, I'm sure you've checked that you have a coach assigned to the Defending category and that he is not overloaded.

And you talk about undesirables attributes, but I think it's not good to think like that. Perhaps all attributes count for a player during a match....you can't imagine that the Match engine only used the 7 or 8 attributes that you can see in the tactic view....All attributes is used in the multiple formula to translate how is going the player in the match engine...

To resume, even Defense training is useful for an attack player or a midfield....

For me position training is not the good thing to maximise chance to develop your player to his best.

Set Pieces are of no use to a DC, as an example. And I would imagine that these are never used in the ME unless you assign your DC to take free kicks or corners. Furthermore, the knowledge we have is that a position and age based approach is a good starting point for training schedule design. Then there is scope for adding a few additional flavours to these baselines.

I wouldn't be comfortable in assigning individual schedules based on the player's own response. In essence the player is telling you how to train him whereas, as the manager, you should be giving the instructions based on how you wish to see him perform in your tactic.

[SFraser's Training Schedules for FM10]

No sorry, this just isn't true.

The graph you are referring to represents the training intensity modified by the quality of your coaches and the personality of the player. It is not an indication of how training should be set up.

If you do this you will most likely see players developing positionally undersireable attributes. I know this because I have trialled the exact method you have described myself.

[SFraser's Training Schedules for FM10]

Those are not the correct attribute weights, and from my investigations into training and discussions with other people about training I have found no evidence that attribute weights are actually involved in training.

Attribute weights seem to me to be like CA and PA values. Important "under-the-hood" values with minimal gameplay involvement. The fact that a Centrebacks Tackling attribute takes up 10x more CA than his Finishing attribute is completely irrelevant from a gameplay point of view, and rightly so.

The only time attribute weights have any relevence is when you are using editors to design players yourself.

It would be a mistake to force players to learn attribute weights in order to carry out training. It is a completely unrealistic mechanic used only to balance the game. Making it an integral part of training rather than designing code to hide it from gameplay would significantly negatively impact the game.

I second this!

Furthermore, the link posted is by Jumbalumba is misleading. The numbers in that table are the weighting RANKS not the weight values themselves.

SFraser, have you checked your email recently?

[SFraser's Training Schedules for FM10]

Thanks Prozone. (Bristol lad like me I see.. red or blue?)

Red... for Man Utd... originally from London! :o

[SFraser's Training Schedules for FM10]

There's something in the pipeline which I think you're going to like... watch this space!

[SFraser's Training Schedules for FM10]

Please don't take my Excel sheet as gospel. I put that up for demonstration purposes only, the numbers for GKs I suspect are not quite right. I see it is being used to create new schedules which I'm very happy to see but if the numbers are wrong then it will obviously create undesired results. I'm going to take down the link to avoid further confusion.

[SFraser's Training Schedules for FM10]

when i test my shedules how long player need to be on it so i can decide its working or not.

I would say a 3 month period, uninterrupted by injury, would be a good starting point. Three points on the Training Progress graph should allow you to see a trend.

[SFraser's Training Schedules for FM10]

Are there going to be a downloadable collection og schedules?

Probably. SFraser and I are working on something at the moment so please be patient.

In the meantime, all the information you need to update the original release is already in this thread.

[SFraser's Training Schedules for FM10]

Should you need some help of any kind, please just let me know...

Some weeks ago I've chatted with SFraser about the possibility of creating this kind of 'all encompassing' training tool, kind of in line with the Team Talk Optimizer (shameless plug ) that I've launched before!

It's good to see this thread really forming the thought process on this approach to composing training schedules now. Big thumps up for all contributors for holding onto a 'focussed' discussion...

Anyway, just let me know if you need a hand on this one...

Would be a waste of time to compose the same kind of tool independent of each other...

This is great to know - thanks. I will be in touch I'm sure.

[SFraser's Training Schedules for FM10]

Does this imply that under the current schedules a young centre backs finishing and tackling (just selecting a random unimportant, and important attribute respectively) should improve at roughly the same rate? (Just when I think I have a handle on things, my brain throws another question into the mix). By your explaination I think it would, but I have been using the schedules for a while and have always observed predominantly the most effective stats increasing more, so I assumed the schedules had already been weighted after being balanced.

My understanding is that SFraser has tried to achieve balance in all cases. So yes, in theory all attributes should improve at the same rate if development was 100% controlled by training alone. But, in-match events for example provide an additional effect on attribute growth and I'd speculate that this may provide an additional boost, particularly to the key positional attributes, which would account for your observations.

I strongly suspect that achieving a perfect state of balance is impossible. The controls at our disposal are not fine enough to affect each attribute individually. Even if the balance is slightly off, your observations would suggest that the difference is in favour of the positionally applicable attributes which is no bad thing.

[SFraser's Training Schedules for FM10]

@ProZone Thanks very much for your quick and clear response, I just have some more questions in relation to the points you raised. I hope you dont mind

How is 'Role Bias' distinguished from 'Focus'? It was my understanding that Focus was already biased towards roles, just that under SFrasers original schedules the roles were more loosely defined.

Also what do you define a "balanced" schedule as? For me a balanced schedule would be the same for every role (thus more akin to 'baseline'), in that it implies even improvement across all attributes.

Im embarassed to say I havnt seen any of your previous work on overall workload, but it would be my inclination to maximize workload subject to unhapiness with workload on an average player (unless you were going to individualise schedules). I have no idea what governs unhappiness with workload but I would assume professionalism, work rate etc.

Your understanding of balance is correct. Focus biases a schedule to achieve balance, nothing else. SFraser's schedules have not been designed for particular Roles. He has designed them for position and age and they are all balanced.

I think this highlights the need for a clear write-up, but I'll attempt to give a brief summary now.

Baseline - the number of slider notches/clicks is set to be equal to the number of attributes trained in a particular category. This provides each training category with what we'll call 1 training 'unit'. For outfield player's I believe we now have:

1 STR unit = 3 slider clicks

1 AER unit = 5 slider clicks

1 GK unit = 0 slider clicks (not interested in this for outfield players)

1 TAC unit = 5 slider clicks

1 BAL unit = 4 slider clicks

1 DEF unit = 3 slider clicks

1 ATT unit = 2 slider clicks

1 SHO unit = 3 slider clicks

1 SET unit = 5 slider clicks

Focus - what SFraser has tried to produce are balanced schedules. What this means is that, in general, all attributes increase by the same amount or all attributes decrease by the same amount. I say 'in general' because there are factors outside the control of training (match events being an obvious example) which affect this improvement or decline. Let's forget those for now.

The development of each player is affected by the position he plays, his age and other hidden character traits. However these hidden character traits are outside of our control so we are left with the ability to manipulate schedules for age and position only. These two factors need to be accounted for in all schedules we design and are the key reason why balance is not the same as the Baseline. Hence, SFraser has produced positional schedules each of which has 4 age variations (Youth through to Veteran).

The tricky part to understand is that development is controlled by CA distribution. Let me simplify this with the following example statement. 'It is harder to improve ATT attributes for attacking players because for attacking players the ATT attributes have higher weighting coefficients than for defensive players and therefore require higher amounts of CA for improvement.'

The following Focus values are for SFraser's First Choice AM schedule:

STR: 3

AER: 3

GK: 0

TAC: 3

BAL: 3

DEF: 2

ATT: 4

SHO: 3

SET: 3

What do we see? The ATT category is given additional Focus to ensure that the attributes contained within are developed at the same rate as the attributes in all the other categories. The DEF category has a reduced Focus to achieve the same effect. Thus, this AM schedule has been balanced. Focus biases a schedule to achieve balance.

Role Bias - now we have a balanced schedule for a player of known position and known age we can use this as a foundation and further bias the schedule to promote the attributes we would like to improve, or neglect the attributes we are not interested in, depending on the Role we want him to play.

Don't worry too much about my Workload comment. For FM2009 I managed to calculate the Overall Workload total from the individual training category slider positions. It's something I mentioned to include in a 'Training Design Tool' so as people don't design schedules which they can't implement in the game due to there not being enough Overall Workload available.

[SFraser's Training Schedules for FM10]

If 'Natural Fitness' could be improved through training, it wouldn't be natural!

Indeed!

But what else are we supposed to assume if it's included in the strength Training Category attributes? If the training module could be tidied up a bit so as to be less ambiguous then it would benefit us all.

[SFraser's Training Schedules for FM10]

You are completely on the right track with your comparison of attributes to categories and this I have done myself, but the Training attribute display is at best "untidy" and at worst misleading.

Absolutely! If I may be so bold and any of the SI guys responsible for the Training module are reading this then I'd like to politely and respectfully ask them to pull their socks up! A little more care and attention to detail, plus a clearer description of displays in the manual (any one of the various manuals at that), would remove a lot of ambiguity.

I like that alot Prozone.

Prozone this simple and clear explanation of the relationship underlying these schedules could be easilly used to define "half increases" for example 2.5 : 3 or 4 ratios of improvement, and should make the whole thing easier to use and discuss.

I heartilly approve of any attempts by you to construct any excell based application for constructing balanced schedules, and would certainly like to find a way to format these ideas and explain them clearly to readers of this thread.

As for the issue of Attributes involved in Training, I have recently come to know specific information on certain types of issues that was completely unknown to me before but that we both discussed between ourselves as being vital information and the "next step" of investigation to clear up a great deal of training issues. This information I will not be releasing to the public directly but I will be using it.

I can say that Natural Fitness, Reflexes, Flair do not require CA in an outfield player and are not involved in Training.

These are all great points and I will return to this thread and these points very soon. I shall first have to go and consider what can be done with this new, and excellent, perspective.

I'm glad you like the idea and I agree this will enable a greater degree of flexibility in designing training schedules.

I have described in my previous post how I think this approach can be extended too.

I will take responsibilty for producing an Excel based 'Training Design Tool' of some sort. It will probably look like the demonstrator I posted earlier but I could build in the Overall Workload contributions I solved for FM2009 if you remember those.

I agree that this needs to be written up clearly now, I'm happy to take that discussion offline.

[SFraser's Training Schedules for FM10]

@ Grisales

Don't despair! I think that, since the Tactics Creator has been introduced, training has become the most complicated area of the game to understand - perhaps it always was! I have been looking at it closely for many months now and I still can't answer many of the questions I have.

Your first point is correct and SFraser has said that there are revisions to be made to the original category baseline no.s used. These are in the process of being corrected. I did point out that my sheet was for demonstration purposes only.

Your second point is correct also. Certain categories can be biased more than others simply due to the fact that they contain fewer attributes. That's just the way it is.

I have long wanted to develop Role specific schedules too, and perhaps this is where I should come clean...

The framework which I proposed above can be easily adapted to do this. Whereas I proposed that Baseline x Focus = Balanced Training Schedule my belief is that this can be modified one step further, thus:

Baseline x Focus x Role Bias = Role Specific Training Schedule

This whole process has relied on someone to understand the requirements for, and produce, a balanced schedule which can then be biased to favour the attributes required for each Role. The Role attributes are known and have been used in the development of my own Squad Analysis Tool (apologies for the plug :o). SFraser appears to have discovered the elusive balance, not simply for one case but for each position and for each age group. In all honesty this is a remarkable achievement

The limiting factor, as always, with training schedules is the Overall Workload limit but we'll tackle that one when we get to it!

[SFraser's Training Schedules for FM10]

I may be wrong ProZone, but I'm pretty sure that I read some time back that Natural Fitness was a fixed attribute and could not be trained, and personally have never seen it go up or down in any long-term game I've played. Not sure about Flair, as it's my personal opinion that it should be a mental attribute that could be encouraged, (as a player ages he becomes more confident in trying things, if that makes sense) though whether it is actually viewed that way by the designeers is another matter.

As you can see above, I came to my own conclusion. I'm not sure if there is a way of knowing for absolute certain either way. It's unlikely we'd get a response from SI either based on my previous requests for clarification.

SFraser has been looking at this for some time now so let's await his response.