FREVKY said: It certailny is smaller than before, what I meant is that ST formula still covers jumping reach twice. I just wanted to point out that square root of "X to the power of 2" is "X", so there's no need for making it more complex than it should be from mathematical point of view. Expand Yes but it's a general solution. MRL also has "Acc Agi" combo, interestingly without Agi itself as a solo attribute.
FREVKY said: Updated the spreadsheet with your corrected rating evaluation for positions you mentioned. So now for striker it stacks up jumping reach, as it is taken into equation twice (most and least important coefficient - for the latter sqrt((jumping reach)^2) equals just jumping reach. Expand What do you mean? The effect of jumping reach should be much smaller than it was before.
Although honestly I think for now it would be best to just remove the double attributes for the time being. Too much confusion and desperately trying to make it work for very minimal effect.
MrSphinx said: A screenshot of these, for example. I am unclear which training to do on which day Expand Give players 2 days of rest after the game and then schedule 1 quality session per day until the next match. If you run out of days then combine 2 sessions on same day that are least demanding according to intensity bar. At the end fill the leftover slots with recovery.
Orion said: In linear regression you basically check every single feature and select the one that fits to the data the most. Then you exclude that one and check the model with this first feature + the second one. In polynomial regression you allow model to look for 'single' feature but also combination like feature1 x feature2 or feature1 x feature1. If it fits the data better than 'single' feature it will pick that feature. If 'single' feature fits the data better than 'combined features' it will keep the single one. In this model I allowed searching for quadratic equations. And for some cases it turned out that at some point, like for the Forward, Jum x Jum fits the data better than any single feature that was left in the pool. But for example Jum x Jum gives worse fit that just Jum. Expand Yes but the model was working with adjusted numbers. Let me give an example. Average JR in the league is 12, a striker has 17. The way you explained stuff in the original post and later on basically says you take 17-12=5 so you do the calculations with 5. And isn't it possible that quadratic fits well for adjusted number (5) but not the actual number 17? Squaring numbers like 5 is a lot different that 17. Above 8 it basically dominates every other attribute basically double dipping (or triple since it's also the best attribute for a striker anyway).
Edit: Or to put it another way. Comparing JR contribution in above scenario if striker has 17 or 18 JR. Taking numbers at face up value you'd get roughly 1,27 vs 1,40. So a difference of 0,13 which would roughly be the effect of +6 pace. Using adjusted numbers of 5 and 6 makes the scores 0,49 and 0,55 respectively, a difference of 0,06 so half of the previous difference. Still equals +3 pace which feels a lot.
Orion said: The training data was transformed to show player attributes not as 1-20 but I calculated average value for each attribute for every league. Then for every player I changed his attribute value as (actual_attribute_value - average_attribute_value_in_player_league). This way attributes can also be negative - if player has lower attribute value than average for his league. And that was the data that I used to training the model. So model says 'hey, take this 'new attribute value' for the player and multiply it by his coefficient'. Because of this in theory we should check in what league you are playing, check average attributes values for the players in that league, calculated this new 'difference' attribute and then use coefficient to calculate player 'rating'. To make it easier, and because we compare players that will play in the same league, we will just use original attribute values of players. This may introduce some additional error in the rating calculation, but since this makes calculation so much easier we accept this flaw of the method. Expand
Btw how does this work with double attributes? Like Jum^. If the model uses adjusted values then squaring the full number seems odd and potentially overshadowing everything else.
Example: Related attributes: Positioning, Bounce, Statistic type: Key headers: stat_typeÈ
So key_headers (headers that become goals?) the attributes used by the engine is Positioning and Jumping Reach only? I guess bounce I just a bad translation.
You guys are really cutting up the whole game 😅 Expand
I have been playing Football Manager for a few years now, but I am still struggling to master youth development in the game. While I have had some success with scouting and signing young talent, I often find that my youth academy players don’t develop as well as I hope.
I am currently managing a mid-tier team, and I’ve been investing in facilities and coaching staff, but I still feel like I am missing something. Here are a few specific questions I’m hoping the community can help me with:
- Training Schedules: What’s the best approach for creating personalized training schedules for youth players? Should I focus on specific attributes early on, or aim for balanced development? - Mentoring Groups: How effective are mentoring groups, and how should I set them up to ensure my young players benefit the most? - Game Time: How do you balance giving youth players enough game time without negatively impacting your team’s performance? Is it better to loan them out or keep them in the squad? - Hidden Gems: Are there any tips or tools you use to identify high-potential youngsters in your youth intake?
I’d love to hear any strategies, experiences, or tips you have on this topic. Whether it’s about specific tactics, team instructions, or hidden features in FM, I am all ears!
Thanks in advance for your insights!!! Expand There are some things you can do with development, but the most important part is signing the right kind of players. That means generally a good/great starting point for their age, a good personality and obviously high potential. You can find some training schedules on this forum. Many threads started by harvestgreen22 contain tons of different options and analysis. I'd warn that most of those schedules produce weird players, so perhaps some general principles. You don't actually need that many sessions per week to reach quite close to the cap that the game seems to allow. 3-4 sessions per week on double intensity get you almost there. In the end you can even leave the schedule to your assistant - it should be good enough. One key part of training seems to be the intensity and I'd never leave that to assistant/on default. Either use settings from harvestgreen22 or maybe ZaZ https://fm-arena.com/thread/12971-training-and-resting/ (those under "current test"!). That way you keep your players fresh, with less injuries while still maximising growth.
Mentoring is kind of useless. It's very hard to fix a bad personality. Much better strategy is to simply sign good personalities, when you can afford to.
After players reach about 19 years of age they desperately need game time in order to progress so at that point you either need to give them solid minutes in your team or find a good loan for them.
ClaudeJ said: I might be misremembering, but I think it's been known for years that training just assign ability points the player earn playing matches into a specific group of attributes. And the more reputations the team and the league have, the more ability point are earned.
I must admit I don't understand the hype here. Expand
Kind of, yes. But the extent you can funnel these "earned points" is new. Creating 20/20 players on demand was unheard of.
bigloser said: In my experience it's a lot more. I'm surprised the league rep didn't matter in HG's testing.
B teams not in loadable leagues(like German Regional) usually do not develop much after 18 even if you micro manage the roster to make sure guys get phantom games. This may be CA dependent because the especially bad players that I move to the 2nd team to prevent them taking time from better prospects in the youth team tend get up arrows for a year or so before stopping.
Getting consistent 20-30 minutes of game time in blowouts for the first team gets more development than youth teams.
This may be selection bias since I'm more likely to force game time for high PA players. Expand
Like I said maybe in another thread. Issue with such B teams is that they are semipro. So players don't train full time and they will hardly develop.
juliius said: I hope some of you guys know this, because i don't. How much value do the u18 and u21 games have in comparison to first team football, in terms of generating player growth? I seem to remember first team football being more valuable. Am i wrong? Expand
Youth games aren't much better than friendlies in terms of development
rossbalch said: So here's a question, it seems like league reputation doesn't matter. But what about first team, vs say a "B" team. If you set it up so both the First Team and the B team compete in the same league, do the first team players develop better than the First Team players? Expand The issue with B teams is that they are usually semipro
Positional tests are much harder to do. For starters they require 10-100x more runs per single test to pick up the signal, on top of 10x because there are multiple positions. There is also a question of player roles potentially adding more runs needed. For example stamina on a WB(A) might be 3x more valuable than on a IFB(D) even though they technically play the same position. Alternatively it might be possible to keep the number of runs down if we can find other metrics to measure.
Additionally some attributes like jumping reach are relative. If the whole team has 8 JR, having a 17 JR winger might have a huge effect. Whereas in a team full of 17 JR players there might be no difference between a 8 or 17 JR winger
Zippo said: I forgot to mention that in this test we capped the max value of Jumping Reach to 17 instead of 20.
We did that because we were concerned that increasing Jumping Reach to 20 would create a very unrealistic environment, like gnomes play against giants, which might distress the result of the test by giving extra undeserved importance to the Jumping Reach attribute. Expand Could you make it a bit more visible in the first post, it's there but unless you read super carefully it's very easy to miss.
If we exclude obviously OP acceleration and pace, then jumping reach is still one of the best attributes. Lets not get carried away by how broken the first two are.
debelizec19 said: Dear @Yarema thank You very much for explaination. So it is very important to have rest and recovery as many as possible in 2 match week + at least one training(quickness/attacking/match p.)? Expand Some would argue you can put an entire week of rest when you have 2 games in the same week. I like to structure around it and still fit in 2-4 quality sessions and then fill the remaining slots with rest or recovery. I only put rest when I can do 3x rest in a single day preferably the day after the match, otherwise I put recovery.
Key thing is to keep your players fresh enough so they can actually play in both matches, no use destroying them with training to gain 1-2 more CA per season at the cost of results.
CBP87 said: beats offside trap for AF and IF, plays killer passes for AM and cuts inside for IF. I also make sure I had a left footed IF on the right hand side and a right footed IF on the left Expand
Honestly I find footedness is very individual. Some players perform better on same side as their dominant foot and some on the other. Haven't figured out a way to know other than simply playing them and noticing where they perform better.
What I'm comfortable saying is that it's certainly not the end of the world if you play a right footed player as IF on right. I feel it was much more of an issue in previous FMs.
debelizec19 said: Dear @keithb please can you explain what the trainig should look like for matches on tuesday and saturday(and other combinations) or give example? You mentiond that we need to be careful when we have matches on different days in week. So now I am little bit confused about training schedules on different days. I am using [Quickness]+[Match Practice]+[Attacking]+[Recovery]x7+[Double Intensity]+[Additional Focus Quickness]. Hope you understand what I am trying to understand Expand
Just put the most impactful sessions as far after the match as you can, at least 2 days away, maybe 1 if you can super rest. And then fill the other slots with recovery. It can vary a bit whether you are home or away for either of the two matches. The day after the match should always be either full rest if home or recovery if you are travelling.
For example you have matches on tuesday and saturday, and you value quickness session the most in your schedule, then put it on friday - that way you have 2 days of rest before and your whole team should be able to participate. And you can put for example attacking on thursday, some of the team might be too tired to participate (and that's OK) but majority should be fine.
Excellent stuff. Would it be possible to add some hidden attributes as well after the initial batch is done? Like weak foot, consistency, important matches and so on
It doesn't differentiate by role
Yes but it's a general solution. MRL also has "Acc Agi" combo, interestingly without Agi itself as a solo attribute.
What do you mean? The effect of jumping reach should be much smaller than it was before.
Although honestly I think for now it would be best to just remove the double attributes for the time being. Too much confusion and desperately trying to make it work for very minimal effect.
Give players 2 days of rest after the game and then schedule 1 quality session per day until the next match. If you run out of days then combine 2 sessions on same day that are least demanding according to intensity bar. At the end fill the leftover slots with recovery.
In polynomial regression you allow model to look for 'single' feature but also combination like feature1 x feature2 or feature1 x feature1. If it fits the data better than 'single' feature it will pick that feature. If 'single' feature fits the data better than 'combined features' it will keep the single one.
In this model I allowed searching for quadratic equations. And for some cases it turned out that at some point, like for the Forward, Jum x Jum fits the data better than any single feature that was left in the pool. But for example Jum x Jum gives worse fit that just Jum.
Yes but the model was working with adjusted numbers. Let me give an example. Average JR in the league is 12, a striker has 17. The way you explained stuff in the original post and later on basically says you take 17-12=5 so you do the calculations with 5. And isn't it possible that quadratic fits well for adjusted number (5) but not the actual number 17? Squaring numbers like 5 is a lot different that 17. Above 8 it basically dominates every other attribute basically double dipping (or triple since it's also the best attribute for a striker anyway).
Edit: Or to put it another way. Comparing JR contribution in above scenario if striker has 17 or 18 JR. Taking numbers at face up value you'd get roughly 1,27 vs 1,40. So a difference of 0,13 which would roughly be the effect of +6 pace.
Using adjusted numbers of 5 and 6 makes the scores 0,49 and 0,55 respectively, a difference of 0,06 so half of the previous difference. Still equals +3 pace which feels a lot.
Because of this in theory we should check in what league you are playing, check average attributes values for the players in that league, calculated this new 'difference' attribute and then use coefficient to calculate player 'rating'.
To make it easier, and because we compare players that will play in the same league, we will just use original attribute values of players. This may introduce some additional error in the rating calculation, but since this makes calculation so much easier we accept this flaw of the method.
Btw how does this work with double attributes? Like Jum^. If the model uses adjusted values then squaring the full number seems odd and potentially overshadowing everything else.
Example:
Related attributes: Positioning, Bounce, Statistic type: Key headers: stat_typeÈ
So key_headers (headers that become goals?) the attributes used by the engine is Positioning and Jumping Reach only? I guess bounce I just a bad translation.
You guys are really cutting up the whole game 😅
Except not all key headers are a goal threat
I have been playing Football Manager for a few years now, but I am still struggling to master youth development in the game. While I have had some success with scouting and signing young talent, I often find that my youth academy players don’t develop as well as I hope.
I am currently managing a mid-tier team, and I’ve been investing in facilities and coaching staff, but I still feel like I am missing something. Here are a few specific questions I’m hoping the community can help me with:
- Training Schedules: What’s the best approach for creating personalized training schedules for youth players? Should I focus on specific attributes early on, or aim for balanced development?
- Mentoring Groups: How effective are mentoring groups, and how should I set them up to ensure my young players benefit the most?
- Game Time: How do you balance giving youth players enough game time without negatively impacting your team’s performance? Is it better to loan them out or keep them in the squad?
- Hidden Gems: Are there any tips or tools you use to identify high-potential youngsters in your youth intake?
I’d love to hear any strategies, experiences, or tips you have on this topic. Whether it’s about specific tactics, team instructions, or hidden features in FM, I am all ears!
Thanks in advance for your insights!!!
There are some things you can do with development, but the most important part is signing the right kind of players. That means generally a good/great starting point for their age, a good personality and obviously high potential.
You can find some training schedules on this forum. Many threads started by harvestgreen22 contain tons of different options and analysis. I'd warn that most of those schedules produce weird players, so perhaps some general principles.
You don't actually need that many sessions per week to reach quite close to the cap that the game seems to allow. 3-4 sessions per week on double intensity get you almost there. In the end you can even leave the schedule to your assistant - it should be good enough.
One key part of training seems to be the intensity and I'd never leave that to assistant/on default. Either use settings from harvestgreen22 or maybe ZaZ https://fm-arena.com/thread/12971-training-and-resting/ (those under "current test"!). That way you keep your players fresh, with less injuries while still maximising growth.
Mentoring is kind of useless. It's very hard to fix a bad personality. Much better strategy is to simply sign good personalities, when you can afford to.
After players reach about 19 years of age they desperately need game time in order to progress so at that point you either need to give them solid minutes in your team or find a good loan for them.
And the more reputations the team and the league have, the more ability point are earned.
I must admit I don't understand the hype here.
Kind of, yes. But the extent you can funnel these "earned points" is new. Creating 20/20 players on demand was unheard of.
B teams not in loadable leagues(like German Regional) usually do not develop much after 18 even if you micro manage the roster to make sure guys get phantom games. This may be CA dependent because the especially bad players that I move to the 2nd team to prevent them taking time from better prospects in the youth team tend get up arrows for a year or so before stopping.
Getting consistent 20-30 minutes of game time in blowouts for the first team gets more development than youth teams.
This may be selection bias since I'm more likely to force game time for high PA players.
Like I said maybe in another thread. Issue with such B teams is that they are semipro. So players don't train full time and they will hardly develop.
How much value do the u18 and u21 games have in comparison to first team football, in terms of generating player growth? I seem to remember first team football being more valuable. Am i wrong?
Youth games aren't much better than friendlies in terms of development
The issue with B teams is that they are usually semipro
Alternatively it might be possible to keep the number of runs down if we can find other metrics to measure.
Additionally some attributes like jumping reach are relative. If the whole team has 8 JR, having a 17 JR winger might have a huge effect. Whereas in a team full of 17 JR players there might be no difference between a 8 or 17 JR winger
We did that because we were concerned that increasing Jumping Reach to 20 would create a very unrealistic environment, like gnomes play against giants, which might distress the result of the test by giving extra undeserved importance to the Jumping Reach attribute.
Could you make it a bit more visible in the first post, it's there but unless you read super carefully it's very easy to miss.
Some would argue you can put an entire week of rest when you have 2 games in the same week. I like to structure around it and still fit in 2-4 quality sessions and then fill the remaining slots with rest or recovery. I only put rest when I can do 3x rest in a single day preferably the day after the match, otherwise I put recovery.
Key thing is to keep your players fresh enough so they can actually play in both matches, no use destroying them with training to gain 1-2 more CA per season at the cost of results.
Honestly I find footedness is very individual. Some players perform better on same side as their dominant foot and some on the other. Haven't figured out a way to know other than simply playing them and noticing where they perform better.
What I'm comfortable saying is that it's certainly not the end of the world if you play a right footed player as IF on right. I feel it was much more of an issue in previous FMs.
Just put the most impactful sessions as far after the match as you can, at least 2 days away, maybe 1 if you can super rest. And then fill the other slots with recovery. It can vary a bit whether you are home or away for either of the two matches. The day after the match should always be either full rest if home or recovery if you are travelling.
For example you have matches on tuesday and saturday, and you value quickness session the most in your schedule, then put it on friday - that way you have 2 days of rest before and your whole team should be able to participate. And you can put for example attacking on thursday, some of the team might be too tired to participate (and that's OK) but majority should be fine.