Sealionborn

This is the second post in a series of posts detailing my ideas for how the FM games can be improved going forward. My first post was this one, in which I detail my ideas for player progression concerning both overall ability and specific attributes. This second post will deal with the general training system in FM, with the goal of creating a system which provides better information and emphasizes player freedom without hamstringing the AI. The general training system currently in FM is built around training modules, with a maximum of three modules being used each day. I very much like this system in concept and I will be refining and modifying it rather than removing it wholesale, my issues are with it's implementation, not the concept itself. The biggest issue with the modules is the fact that they don't do what they say they do, which is just about the worst flaw a system in a simulation game can have, aside from not working at all. It's been well documented by Evidence Based Football Manager that the "Overall" training module does not have an appreciable effect on physical attributes, despite claiming to improve them. In my anecdotal experience the Teamwork attribute almost never improves, despite being listed in almost every module i use when constructing my training schedules. This makes the system difficult to engage with on a deeper level, as the player has no way of making informed choices when all the information we're given is flawed at best and outright false at worst. The first part of my suggested improvements is therefore a simple one: make the modules list EXACTLY which attributes it benefits. The module should give an equal amount of attention to every attribute listed in the same focus, and do nothing for any other attributes. This way the information presented is reliable and easy to understand, letting the players make actual informed decisions about training. My second suggestion is that the players should be allowed to design their own modules. Allowing the players to design their own training modules would let them mould their teams to suit a specific playing style; or the player could create well-rounded training modules that create more versatile players and teams. The custom modules should allow the players to choose primary and secondary focuses, affected position groups (All players, Goalkeepers/Outfielders, Goalkeepers/Defenders/Attackers), as well as how much focus each of these groups get. My third suggestion is that the intensity of training modules should be flexible. In FM today a "Physical" module has an intensity of 35, there's no way to get it lower or higher. I propose a system where players decide the total intensity of each day of training, say Monday is a 60 and Tuesday is a 40, and then divide that total between the modules assigned for that day. This should help the AI as well, seeing as the default training schedules can be pretty erratic in terms of training intensity, with some days having an intensity of ~70 while others are in the 20's. My fourth suggestion is a small one; let the players use as many of one training module as they want in a week, I really don't see a reason to stop players from doing this. If the Players want to streamline training the game should just let them, forcing players to engage with a system they find tedious just makes them more likely to automate it by leaving it to their staff, which isn't desirable in my opinion. If you have any critiques or additions you want to make to my ideas please let me know.

This is the first post in what will be a series of posts detailing ways I would like to see the Football Manager games improved. I have written about some of these ideas in previous posts, but I hope this will be a more cohesive and well structured improvement on those posts. When I refer to attribute numbers in these posts I will be using the 1-100 scale that's actually used to calculate a player's current ability, not the 1-20 numbers we see displayed. Just divide the numbers I give by five and round up to get the displayed value. The first issue I would like to discuss is player growth/development. This is the core of FM to me and many other players -- there's nothing quite like seeing a youngster from your academy blossom into a world class player after all. While the current growth system is decent I feel like it depends too much on seemingly random factors. I like to create players using the pre-game editor and play out their careers at my club. As I did that I found that even near-identical players would develop in radically different ways, even when given effectively identical training and match experience. As an example: One player would grow their work rate by 4, while the other grew it by 1 in the same period after having the same general and individual training schedule. This suggests that there must be a significant amount of randomness in how players develop. I think this takes away from the fun of the game; why pay attention to training if the results are more or less random? I have an idea for a system which removes the element of random chance while still providing varied outcomes. This system is based around two kinds of "points" I call Experience and Priority. Experience points are responsible for improving attributes whereas Priority points decide where the Experience points are allocated. Experience points would largely be earned through match experience, though training does contribute, especially for players who train well. Priority points are earned almost exclusively through training, though some attributes like Leadership and various Set Piece attributes should also improve when used in a match setting. Every week (ideally calculated at 00:00 every Monday) the player's total Experience should be distributed across their attributes according to how many priority points each attribute has. If improving an attribute would cause the player to exceed their Potential Ability score that attribute should be frozen, all excess Experience should be distributed to the remaining viable attributes until progression becomes impossible. Of course the issue with this system is that it would be extremely predictable and therefore boring. There needs to be some dynamism to it and I have some ideas. The simplest is making some attributes require more Experience to grow than others, even if they all start at the same number. Attributes like Bravery are difficult to improve IRL and that should be reflected in the game. Other attributes should be made more or less difficult to grow depending on other attributes and the player's body. For example: A 190cm tall player should have an easier time getting to 80 Jumping Reach than a 180cm tall player would. Some attributes may be "tethered" to each other. It would make sense for a player with high Technique to have an easer time improving their First Touch than someone with worse Technique. This could also stop "free" attributes from growing out of control. By "free" attributes I mean the attributes which don't contribute to a player's Current Ability score. If the growth of Flair is limited by Dribbling and Technique we shouldn't see a bunch of old centre backs with 15+ Flair because that was an attribute they could still improve without exceeding their Potential Ability rating. The more complex idea is what I call the "Development Archetype" system, I'll just be referring to them as Archetypes for brevity. The Archetypes give a boost to Priority point generation for various attributes, it does NOT affect total Experience gain. The Archetypes are inherent parts of the player's DNA and they cannot change over time. An Archetype is made up of two parts: The Positional Archetype and the Stylistic Archetype. The Positional Archetype designates a player as a natural at a specific position. A player who can only play DM may generate as a "Defensive Midfielder" or "Midfielder", these Archetypes boost the priority of different attributes to varying extents. Every position should have at least two Archetypes attached to it. Think "Winger" and "Wide Forward" for the AML/AMR positions, one more focused on wide play and another more suited to cutting inside. These Archetypes may even boost the same attributes, but the extent to which every attribute is focused on will be different. This should prevent things like strikers massively improving their Marking and Tackling while Finishing is stagnant, even when their training schedule is well balanced. The Stylistic Archetype gives further focus to a smaller group of attributes. These attributes are responsible for a specific part of the game, say press resistance or elusive 1v1 dribbling. These groups should consist of at most 5 or 6 attributes. Individual attributes like Dribbling should be represented in multiple attribute groups. Every attribute should be a part of at least two groups, the more the merrier. I'll discuss the groups in more detail in the upcoming training focused post as they form the backbone of much of my thinking there as well. Combining the two Archetypes should form a descriptor like "Press Resistant Midfielder" or "Tricky Winger" letting a manager know what kind of role the player would have an easier time developing into. The full descriptor should only be visible at 100% player knowledge, but the Positional Archetype should be visible for any scouted player. By combining the Stylistic and Positional Archetypes it becomes possible to create varied developmental outcomes without relying on randomness. If two players start with the exact same attributes but entirely different Archetypes they will develop into very different players, even with the same training schedule. If I train two players with identical starting attributes as Box to Box Midfielders their whole careers, but one player is an "Aggressive Defensive Midfielder" and the other is a "Press Resistant Midfielder", they will naturally develop into different players in a predictable yet compelling way. The Aggressive player might have a 50% priority boost to Tackling and Aggression, while the Press Resistant player might have a 50% boost to First Touch and Balance. In my opinion this system would be a massive improvement to the way development works at the moment. It's a rational system in the sense that it doesn't rely on randomness, but it still provides varied outcomes. It's a more responsive and engaging system which would work both with the current training mechanics and with my upcoming suggestions for changes to this system. Please tell me what you think of my ideas, constructive criticism is very much welcome.

I assume it's being replaced rather than removed, and I wouldn't be too surprised if the changes were functionally insignificant. I assume the change will be largely focused on UI functionality, navigability, and aesthetics.

I assume you mean a financial penalty from the owners? I think most owners would hesitate to kneecap an already struggling club by outright withdrawing money. But I could see the owners imposing certain limits on a manager like "No contracts worth in excess of X per year" or "Max transfer value of Y" if the club is struggling, so as to limit financial problems in the future if they think they might sack the manager.

Personally I would like this system to use the same 1 - 20 scale as attributes and positions. This way we can model a player becoming more comfortable with a trait as he works on it. Maybe players could even pick up traits by osmosis from playing a certain role extensively? I feel like this would be a more realistic way for traits to form, as players habits usually form without direct involvement from the coaching staff. If a player has been playing DLP for years, and he has the stats for it, it would make sense for him to develop the "dictates tempo" trait right?

My understanding is that the training focuses the Assistant suggests are based of what role the Assistant believes the player to be strongest in. So if I go to the individual training screen and click the "Ask Assistant to set individual training" button, he would set players to train in their best role and set individual focuses based on that role. I think it would be better if the individual focuses were determined by the role the player is being told to train. So if I want my centre back to become a Ball Playing Defender and set that as his training role, the individual focuses should help him grow into that role, not whatever role the Assistant thinks he is best at. To add to this i suggest splitting the "Ask Assistant to set individual training" button into two options: "Ask Assistant to assign for everything" which does the same thing as the current button, and "Ask Assistant to assign focuses" which only assigns focuses. This should make it easier to shape players to fit our tactics without having to switch focuses around manually, saving us a lot of busywork without fundamentally changing the training system.

The way training is set up currently leads to a lot of seemingly random or inexplicable results. I've seen just about enough of my strikers finishing and movement growing by 2 across an entire career, while their positioning, marking, and tackling grow by 6. All of this in spite of me running exactly 0 defensive sessions ever. As such I have taken it upon myself to design a training/progression system which would actually be responsive to the manager's input, while still being dynamic enough to avoid becoming completely predictable. Before we start I want to clarify a point: when I say an attribute progresses by a point I do not mean it progresses from 15 to 16, rather I mean it progresses from 75 to 76, as under the hood FM stats are actually rated from 1 - 100 not 1 - 20. The game simply divides the actual attribute by 5 and rounds up to get the number displayed, this is how your coaches can say a player has improved significantly in an attribute despite the displayed number remaining unchanged. Personally I would prefer if the game displayed the actual number rather than the 1 - 20 approximation, but that is beside the point. Part 1: Progression. One of the issues I have with the current system is that players seemingly progress at unpredictable intervals. Personally I would prefer if players progressed all at the same time. I propose 00:00 every Monday. Paired with the more significant changes I suggest below this will help to create a more predictable and satisfying progression system. As for how players grow, I suggest a system based on experience points gained through training and match experience. The training load and match experience -- modified by the quality of the player's performance in both arenas throughout the week -- contribute to the pool of total experience gained. This pool of exp is then allotted to the player's attributes according to a system of weighting which I will describe later. Once an attribute has acquired the necessary exp it will increase by one point, say from 60 to 61. When a player's CA matches their PA ceiling their attributes will freeze at the level-up point if levelling up would bring the player's CA above his max PA. Any excess exp is then passed to the next attribute in line until all attributes are frozen. This stalemate is only broken by occasional attribute decline brought on by holidays, age, or injury, allowing attributes to progress again. If a player's attribute declines due to injury -- say acceleration falls after an ACL tear -- this should be counteracted by that attribute receiving a proportional weighting bonus once the player recovers, representing the player trying to get back to their best. Lastly, before we get into training I should note that not all attributes are made equal, some attributes are simply harder to improve than others, especially as a player ages. Acceleration and Pace might be more in line with the requirements of other attributes when the player is still a wonderkid, but as time passes it becomes more difficult to improve these attributes, until eventually all you can do is keep pace with the decline. This probably goes without saying, but improving an attribute from 80 to 81 should require more exp than it took to improve that same attribute from 79 to 80. Part 2: General Training. The core part of training your players is General Training defined by the various training modules we use to make our training schedules. As much as I love the idea of this system in theory, in reality it doesn't really work as implemented. As showcased by Evidence Based Football Manager on YouTube and others, the modules don't do what they are supposed to. Notably the Overall module doesn't progress physical attributes the way its description suggests, despite carrying a higher workload and risk of injury to suggest physical work is being done. This makes the modules unreliable, and makes adjusting training to work more on a specific aspect of the game difficult and unsatisfying. After all it is hard to trust that running more Defending sessions would actually matter when my players are already becoming solid defensive contributors based solely off of a diet of Physical, Attack, and Possession sessions. At the same time this training regime is not reliably progressing my players' attacking or physical attributes, which sometimes leaves me wondering if training is a total placebo. The first part of my solution to this problem is extremely simple: make it so that every attribute listed on a module is given the same amount of weight, I.E. if Teamwork is one of ten attributes listed it will receive one tenth of the attribute weight. This should ensure that training actually produces the desired results. The second part of the solution is to allow the managers to design their own modules. Designing training is a HUGE part of management and I feel like the current system lacks both depth and engagement. If the player could design their own modules, including everything from which attributes are affected, the training load it incurs, and which groups of players get the most focus, this would allow the player to craft the training regime they truly want -- like a real coach -- rather than having to make do, as is often the case currently. Part 3: Weighting and Individual Training. With all this groundwork done in regards to progression and general training it is finally time to talk about individual training and the Weighting system. What the weighting system does is taking the exp gained from training and match experience and decide how much of that exp goes to each attribute based on how much weight it has. Unlike total exp weight is not modified by training or match performance, it is a fixed value. If a training module awards 1000 weight units and affects 10 attributes, each attribute would receive 100 units of weight. If this training module was the only experience the player got that week then all of its exp would be shared evenly among these 10 attributes with identical weight. If we consider all the results one can get from the near infinite array of possible training schedules it's easy to see how this can become a complex yet responsive system. It would allow managers to create well rounded teams, as well as teams geared toward a particular tactical style. Certain actions during a match will add weight to an attribute as well. Taking set pieces should add weight to the respective attribute, acting as captain or vice captain in a game should likewise add weight to the leadership attribute. This should go a long way to compensate for these attributes being left out of most training modules. As for the weight of attributes affected by match experience, this is decided by the player's individual Position/Role/Duty training. If a player is set to train as a Libero (Support) this will result in a slightly different weighting than Libero (Defensive). This weighting is based on the weighting the game uses to determine which Position/Role/Duty the player is best suited for. This weighting is also used for any training module whose affected attributes is listed as "Individual Roles". If a player training to be a Regista is played at CAM their attributes would still be weighted according to the Regista role. If no role is being focused on the weighting will be decided by what position -- not role -- the player is played in. For training it will use the weighting of the position assigned as the player's "playing position" if no Position/Role/Duty is specified in individual training. As the only function of Position/Role/Duty training is to modify the weighting of attributes it should not increase a player's workload. Actually telling a player to work on a specific aspect of their game -- say Defensive Positioning -- should increase exp gained from training as well as add weight to the relevant attributes. As such it does increase training load. The same logic applies to telling a player to work on their weak foot, as this adds the weak foot on the list of weighted attributes. I would also suggest making it so the individual training suggested by the backroom staff is based on which Position/Role/Duty the player is being trained in, I.E a player being trained as an Inside Forward should not be recommended crossing training. Part 4: Adding some spice. I feel like the changes outlined above would create a far more responsive and engaging system for managers to deal with, but it does run the risk of becoming too predictable. If every player developed the same it would quickly become just as stale as the current system (though still superior IMO). To counteract this without introducing simple RNG to the game i propose a mechanic I call "Developmental Archetypes". A developmental archetype modifies the total exp required to increase certain attributes, making it so players are more likely to develop in certain directions, as they simply have an easier time improving those attributes. My idea is that each player will have two of these archetypes, one primary and one secondary. All archetypes should be available as primary or secondary, though a player cannot have the same archetype twice. The archetype chosen as the secondary archetype will have half the effect it would have as a primary archetype. As an example of how this system would work let's compare prime Salah and prime Messi, two players who ostensibly played the same position: Right Winger as an Inside Forward. A player like Salah might have a primary archetype called "speedster" which boosts their Acceleration and Speed, plus a secondary archetype called "pressing monster" which boosts Teamwork, Work Rate, and Stamina. Meanwhile Messi might have a primary archetype called "virtuoso" which boosts their Dribbling, First Touch, Technique, and Vision, plus a secondary archetype called "shifty" which boosts Dribbling, Acceleration, and Agility. This way two players with similar starting attributes can develop dynamically even with the same exact training regimen, making the game far more compelling. There should also be archetypes which negatively effect certain attributes, after all Adel Taarabt isn't turning into a team oriented pressing machine no matter what training you give him. Though exclusively negative archetypes shouldn't exist: A "soloist" like Taarabt makes up for much of what they lack in Teamwork, Work Rate, and Stamina through technical skills and Vision. I imagine the primary developmental archetype being known for any player you've scouted, whereas the secondary archetype is only known for the players at your club. These are my suggestions for how to improve the player progression in FM. After all, seeing players develop is the core of the FM experience, so it's disheartening to see the system be as non-responsive as it is currently. If you have any critiques or additional suggestions please add them to the conversation. Edit After some consideration I think it would be better if the Developmental Archetypes modify the amount of Weight an attribute receives rather than how much exp it takes to improve the attribute. Say a player with the "shifty" primary archetype earns 130% of base weight units for Dribbling, Acceleration, and Agility, instead of giving these attributes 70% of base exp cost. My original suggestion would let players with relevant archetypes develop faster in addition to directing their growth, effectively letting some players earn more exp. This runs the risk of making some archetype combinations flat out superior in a way I find undesirable. I think connecting archetypes to the weight system instead of exp makes it more of a seamless addition than the original suggestion.

At present we can use the pre-game editor to see how every Attribute is weighted for each Position in the game, as well as a player's score in the various Roles they are eligible for. However we can not see how the Role Scores are calculated, which is a point of frustration for me as I enjoy creating and signing players with specific Roles in mind. Of course the Key/Preferred Attributes highlighted by each Role give a decent idea of what a player needs to perform the Role, but it is not always intuitively obvious based on these Attributes why a player is rated higher in one Role than another. It is by way of example hard to figure out why the game sees Jude Bellingham as a Ball-Winning Midfielder rather than a Box-to-Box or Mezzala, considering his range of Technical Attributes, Vision, and Off Ball Movement. If I had access to the specific weighting of these Attributes it would go a long way towards helping me understand precisely what I should be looking for when scouting for, or creating a player with a specific Role in mind. I don't understand why this information should remain obfuscated when Positional Attribute weights and Role Scores are now readily available -- in the end this merely constitutes a more granular view of what the Key/Preferred attributes already tell us, as well as a look at how the sausage is made in regards to Role Scores.

My understanding is that under the current system, if I change a player's training from the default position training to for example Inside Forward (Attack) this causes all attributes considered as Key/Preferred to get an equal amount of extra focus. I am unclear as to what attributes get attention under the default training mode, as far as I know there might be no real difference between DR and AML as far as default position training is concerned. My suggestion is that the system should be changed to a more nuanced system based around the attribute weighting for positions and roles. Under this system, default position training gives the attributes focus proportionally to the attribute weight for the position, for example AML would heavily focus on acceleration, agility, and pace -- while relatively neglecting finishing and off ball movement. When changing a player's focus to a specific role, say Inside Forward (Attack) this should change the weighting from the positional default to the weighting for the role, putting more emphasis on finishing and off ball movement than what is there by default. Moving a player form positional default training to a specific role should not increase a player's workload -- unless they are being trained in an unfamiliar role. Alongside this system i would suggest that individual attribute training should be modified to work with this system. To be specific I would prefer the current focus groups be broken up -- if I have a player with underwhelming pass accuracy -- but with good technique and vision -- I should not have to train him in all three attributes at once, this leads to player being unhappy with training as they don't see passing as a problem area in their game, and overall it makes individual training unintuitive and inconvenient. I think we should be able to focus any attributes we want in individual training, possibly with the limitation that we can only focus three attributes at a time, as this aligns with most of the focus groups anyway. In this new system individual training would add weight to the attributes being focused on, say 6 or 12 in total (just to have it cleanly divisible between both 2 and 3) and increase the total training load. In my opinion this system would give managers -- both human and AI -- more control over their player's development and allow us to shape players according to a system of play, much like clubs like Barcelona and Ajax have done historically.

Currently the editor allows us to see the weighting of attributes for positions as well as a player's score for every relevant role. My issue is that we don't get the same insight into how the role scores are determined as we do for how CA itself is calculated. I suppose there is an argument that this would demystify the game too much, but I think it would be helpful to know what the games internal criteria for each role is, as the Key/Preferable attributes are too vague and don't always correlate with what scouts and assistants deem a players best role to be.