[SFraser's Training Schedules for FM10]

You really need to read this thread in it's entirety. I am not going to constantly rewrite detailed explanations of my entire thought process and design procedure because you cannot be bothered to find the previous explanation in a 5 page thread.

Start with this post. http://community.sigames.com/showpost.php?p=4924985&postcount=250

I do not really expect you or anyone to break down everything I want to know or understand at every instance. However, that post really does not answer my question which is why I asked it in the first place.

In that post you do state that determining "Focus" for specific positions is complex. I agree.

I will explain how to err on the side of caution and how to work with these "unknown details" in my worked example to follow.

You just state, after some qualitative analysis:

The ideal overall Focus might be something like the following:

STR: 4

AER: 2

GK: 0

TAC: 4

BAL: 2

DEF: 4

ATT: 2

SHO: 2

SET: 0

Similarly, you have done so for further reworkings. You state you will explain how to work with these "unknown details" and yet I do not really see the "how" here. Again, how do you get these numbers? Surely there is something more than a mere idea that focus on Strength is important for this player and that "[w]e should forget about" some other things. What you have said does not translate to numbers which is why I'd like such explanation. Arguably, "forget" can be translated to training 0 in those categories but that is not so for you.

It may be necessary for me to state that your analysis makes sense and the suggested schedule workings make some sense too. However, something that is quite different would also make sense taking into account of that qualitative analysis. I guess my question is: Why those numbers and not something else? It may be quite an arbitrary exercise, you have stated that the focuses "might" be as you stated, and you have recognised the need to check the progress of training to see if the schedule works but all those provide no real justification for the numbers you have chosen above.

[SFraser's Training Schedules for FM10]

My own approach to Training is to simply find out how to judge slider positions relative to each other. To find out what slider positions per Category produces the same or similar Training effect once Age is accounted for. My first attempt was to increase each Category according to the number of Attributes contained within that Category, and it turned out to produce incredibly accurate results. Only when I was convinced that I was on the right track, that training Attributes per Category and not simply matching Categories was producing significantly superior results in terms of targetted attribute increases, did I release these schedules to the public.

I believe I understand what you are saying here but how would this approach give us the best training? Understandably, an understanding of the levels of training that gives us the same rates will help for further adjustments for more specialised training with regards to position. However, do you do these adjustments? If so, how or, more specifically, how do you determine which categories to adjust and by how much to adjust? If not, then why is training at the same rate in every attribute the preferred training?

For each position their is an U-17 Schedule, a Developing Schedule, a 1st Choice Schedule and a Veteran Schedule. The U-17 Schedule emphasises key attributes for the position at the expense of any nonesense young players do not need to learn. The Developing Schedules are for players between 17 and 24 or those than need a crash course in the Senior Basics. They are Physically Intensive and they focus again on vital areas of the game for that position. The 1st Choice is generally a more rounded, more "Maintain" Schedule aimed also at reducing intensity and maintaining condition and Morale. The Veteran Schedule is aimed purely at players who are starting to decline Physically. It attempts to halt the destructive drop of Stamina while channeling Aerobic CA into Mental and Technical attributes in order to build a Maldini or Giggs.

In looking at your schedules as well as your statement above, there is obviously not a strict focus on training at the same rate in everything. So what's going on? Again, from the initial point of training at the same rate in everything, how do you determine which categories to adjust and by how much to adjust? I'd like more focus on "how much to adjust" because the answer to "which categories to adjust" is inherent in that.

[SFraser's Training Schedules for FM10]

My "theory" is simple. Ignore "Training Lines" ignore "Attribute Weights" ignore all these other theorems that simply do not correspond to what is seen ingame. Take into account the number of attributes per Category, take into account the impact of Age on the rate of improvement/decline of each of the three Attribute Profile Panels.

Are you saying then that you do not, at least directly, differentiate between different positions?

In practice you appear to go outside of your abovementioned simple theory. You do seem to consider which training categories are important in your taking into account the impact of Age. As you have pointed out in your examples, older players decline in their physicals a higher degree than in their other attributes and so you relatively increase their physical training. However, you do not uniformly decrease their training in other categories. In one of your examples, you stated that you chose Tactics and Defending to be relatively unharmed and instead decreased the training of the other categories.

Now how do you come to the conclusion that maintaining physicals but training less in other categories is preferable to declining physicals but relatively greater training in other aspects? In the same way, why is Tactics and Defending preferred to the other categories? How do you come up with how much preference you give to these categories? Are you not applying some standard of importance here?

Without any such standard, would you really have a striker with the same amount of shooting training as a central defender? To my knowledge, which is admittedly quite limited, there is no suggestion that strikers and defenders train or lose shooting attributes at different rates apart from the fact that the lifespan of strikers is smaller. Any effect due to age will most likely not be enough to result in the large difference in training in shooting that people would usually consider appropriate. It probably wouldn't suggest the 50-100% difference that you have in your schedules.

If you are not applying some sort of standard to each category then I'd like to understand what is happening with that example you gave. If you are applying some sort of standard then I can understand you better in that respect. Even then though, I'd still like to understand how you come to how much preference you attribute to each category.

[SFraser's Training Schedules for FM10]

It is you that is jumping to conclusions and making assumptions. You assume that it is important to know the relative weightings in order to make accurate schedules, but you have no evidence or arguement to back this up. It is a mere assumption based on the fact you are aware that attribute weightings exist.

It seems I must state again that I do not intend to say that weightings have any effect on actual training or the speed of training.

We want the optimal training. Certain attributes are more important than others. Certain attributes behave differently than others based on age etc. as you correctly pointed out. Instead of adjusting the basic training schedule based on how many attributes are in each training category by an arbitrary method of deeming certain attributes as more important or affected differently, I would like a more scientific study on this. I believe attribute weightings offers an unbiased opinion on this so I believe studying them is of value. I am not saying they are definitely correct and I do recognise there are other factors too. All I am saying is studying them is of value.

We can look at this another way. The initial starting point for you is that all attributes are evenly weighted in terms of training. We can make a schedule based on that but obviously we need to adjust for position, age etc. or else everybody would have the same training and that's illogical. So we adjust it by how important we deem certain attributes or the training of certain attributes for each position or age of the players. How do we do this? The CA weightings offer one solution to adjusting for position. Again, I am not saying that they are definitely correct or optimal but the fact that it is programmed for certain attributes to be weighted higher means that those attributes are of higher importance to that position. The weightings offer us an unbiased way to adjust schedules instead of the arbitrary method of thinking 'hmm, I think this category is this much more important than another so I'll add a few clicks here and I'll take away a few here' with "no evidence or arguement to back this up".

As for the question on your assumptions, that was in no way taking a jab at you and what you are doing here. I just want to ask what "evidence or arguement to back this up" you have for choosing to follow a way that assumes each attribute is weighted evenly in terms of training rather than the more popular 'Number Line' theorists that assume each training category is weighted evenly. I can see the logical underpinnings of your assumption in that there are fewer attributes in Strength so why should I train as much as Aerobic but, on the flip side, it seems somewhat weird that training would work like that. The idea of training Strength lightly and training Aerobic mediumly leading to the same improvements in each attributes can be uncomfortable. You should understand why I can be a bit skeptical because throughout all the years of FM, clearly the most popular of training schedules have always had Strength and Aerobic at the same amount. I have always used my own schedules based on equilibrium amounts of training with the idea that the equilibrium position is different for each category and for each position. However, surely the great majority cannot get such an integral part of FM so wrong for so long.

I do recognise that you have provided a few examples but I guess what I wanted to ask was "is there any non-empirical basis to your assumption?".

[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'm not saying that weights affect training. However, we are trying to find the best or optimal training schedule and I do believe, because CA is limited and attributes have their weights that make up CA, that taking into account how weighty attributes are is the best way to go. I view CA as the overriding factor and I view training as a way to mold CA into attributes. Without taking into account the weights of attributes, how can one optimally mold CA into attributes?

You stated in the OP that individual attributes do have their complications based on position (as well as age):

The complicated part comes from the fact that certain attributes require more CA to improve than others, based on Age, Position etc. There are some rough and ready rules but no exact figures so there can be no exact science to training.

The Strength and Tactics Categories for a Centreback will require more CA to increase than the Ball Control Category. If he is young then his Strength Category naturally receives more CA anyway, if he is old his Strength Category naturally tries to shed CA. This means that you will always have to consider Strength Training much higher than others to achieve the same rate of increase, but this is not true when a player is very young nor very old.

So apart from these complications, you assume that attributes have exactly the same weight with regards to training. Is there any basis for this assumption? This is much the same in comparison to "Number Line Theory" proponents who assume that each training category is weighted equally.

With that said, I do recognise that those aren't the actual weights but are rather the rankings of weights. However, there must have been a way to get those rankings which would be from some sort of raw data closely related to actual weightings. In fact, actual weightings are unnecessary in this study. What is required are relative values to give us an idea of the proportions to make an ideal schedule.

In any case, rankings still gives us a better idea of which attributes the game deems important to the positions. Which attributes are important are a necessary consideration for any schedule rather than a rudimentary schedule based on how many attributes are in each training category. I do recognise that you have made adjustments based on importance of attributes but a less arbitrary study on this is valuable in my opinion.

[SFraser's Training Schedules for FM10]

If you really want to train attributes at the same level, I don't think looking at the rudimentary numbers of how many attributes are within a training category is good enough. I believe weighting the training according to the weights of all skills in each training category as they are for each position will work better. These can be found here http://community.sigames.com/showpost.php?p=3275601&postcount=670.