Meta Attribute Weighted Training Sessions: Part 2(General Best)

by Han106, Jan 27, 2024

Sane said: What do you think about individual training? Are they worth using?

I have never tested, so this is just my guess. I think it is good to train Pace and Acceleration, even when they are too old to develop. That will allow you to funnel CA growth to those attributes, or at least make them lose other attributes instead of speed when they get old.

0

ZaZ said: It was just an example.

By the way, I am testing the schedule below right now. Keep in mind that it requires training with half intensity, else it will lead to a high number of injuries. I also discovered a glitch on training that allows you to "super rest" players (almost 40% more recovery of condition and fatigue than just full rest schedule), but I am still investigating it to understand exactly how it works.


can you elaborate that glitch please?

0

doru228 said: can you elaborate that glitch please?

In short, a day of full rest (three Rest sessions), plus going in holiday with the whole team during that day, recovers more physical condition and fatigue, than just having full rest without holiday (or having holiday without full rest). Apparently, the physical impact of going in holiday stack with the physical impact from training, so you can stack holiday with Rest to get a super recovery.

0

Lapidus said: No way that Stamina can be impactful as Acceleration and Pace for any positions.

In the FM-Arena test Stamina attribute was increased by 5 points for every positions including the DMs and WBs and that improved the result just by 2 points from 60 pts to 62 pts, which honestly can be just the RNG but when Acceleration or Pace attributes were increased by 5 then it improved the result by 21 points from 60 pts to 81 pts.

So no way Stamina attribute or any other attribute can be as impactful as Pace and Acceleration for any position.

Pace - https://fm-arena.com/tactic/7528-5-points-to-pace-for-all-positions/

Acceleration - https://fm-arena.com/tactic/7527-5-points-to-acceleration-for-all-positions/

Stamina - https://fm-arena.com/tactic/7557-5-points-to-stamina-for-all-positions/


Here's another test when Acceleration and Pace was tested, specifically, for Central Midfielder position - https://fm-arena.com/thread/5351-should-you-follow-the-highlighted-attributes-of-the-roles/


I don’t think your logic is correct.  When you test attributes one by one and for all positions it can only tell you if an attribute is important for the majority of positions.  The only way to know at a single position level is vary multiple attributes in the same experiment and vary by position as well (this is in effect what the machine learning data is reporting).  The best you can say about stamina is that it is not important for the majority of positions, and in a very controlled setting where players are always starting games at peak performance and have equally capable replacements available on the bench.

0

ZaZ said: In short, a day of full rest (three Rest sessions), plus going in holiday with the whole team during that day, recovers more physical condition and fatigue, than just having full rest without holiday (or having holiday without full rest). Apparently, the physical impact of going in holiday stack with the physical impact from training, so you can stack holiday with Rest to get a super recovery.

That’s just crazy 😂
Thanks for sharing, will test this out 👌

1

ZaZ said: It was just an example.

By the way, I am testing the schedule below right now. Keep in mind that it requires training with half intensity, else it will lead to a high number of injuries. I also discovered a glitch on training that allows you to "super rest" players (almost 40% more recovery of condition and fatigue than just full rest schedule), but I am still investigating it to understand exactly how it works.


"super rest" caught my attention. I hope you get results from your research :)

1

pixar said: "super rest" caught my attention. I hope you get results from your research :)

I thought "super rest" was a fitting name, because it is around 30-40% more effective than just resting, for both physical condition and fatigue. If you incorporate that in a training schedule, with three rest sessions + holiday once a week, after the match, then you will also have a nice reduction in injuries during the season, as well as allowing to repeat the main team more often, and completely avoiding fatigue. I am currently testing if it is better to do it once a week, or something like once every other week.

1

This "problem" has been present for several years

0

MeanOnSunday said: When you test attributes one by one and for all positions it can only tell you if an attribute is important for the majority of positions.

I agree with you.

MeanOnSunday said: The only way to know at a single position level is vary multiple attributes in the same experiment and vary by position as well (this is in effect what the machine learning data is reporting).

I don't agree. You have to vary one attribute for one position to know its value for this position (in a particular tactic). If you vary multiple attributes than you don't know which attribute is important for the performance increase.

Spoiler Why do scientists change only one variable in a controlled experiment?

In a controlled experiment, scientists change only one variable at a time in order to accurately determine the effect of that specific variable on the outcome. By isolating and manipulating one variable while keeping all others constant, scientists can establish a cause-and-effect relationship between the variable they are testing and the results they observe. This helps in drawing accurate conclusions and making reliable predictions based on the experiment's findings. If multiple variables were changed at the same time, it would be difficult to attribute any observed effects to a specific variable, leading to unreliable or inconclusive results.


https://www.quora.com/Why-do-scientists-change-only-one-variable-in-a-controlled-experiment

0

Meriten said: I don't agree. You have to vary one attribute for one position to know its value for this position (in a particular tactic). If you vary multiple attributes than you don't know which attribute is important for the performance increase.

Spoiler Why do scientists change only one variable in a controlled experiment?

In a controlled experiment, scientists change only one variable at a time in order to accurately determine the effect of that specific variable on the outcome. By isolating and manipulating one variable while keeping all others constant, scientists can establish a cause-and-effect relationship between the variable they are testing and the results they observe. This helps in drawing accurate conclusions and making reliable predictions based on the experiment's findings. If multiple variables were changed at the same time, it would be difficult to attribute any observed effects to a specific variable, leading to unreliable or inconclusive results.


https://www.quora.com/Why-do-scientists-change-only-one-variable-in-a-controlled-experiment


Neither you or the quote are correct in the context we are discussing, which is optimizing performance of a team with many positions and attributes.  If you think of performance as a multi-dimensional surface, changing one variable only tells you that performance is sloping upwards in that region of the surface.  But this can be towards only a local maximum, not the global maximum.  Think of climbing a mountain; if the steepest path from where I am standing is North then will going North always lead me to the summit?  No, if I looked around me I might see the summit is to the east and that what is North is just in a smaller peak below the main one.  Or suppose I test the strength of a bond by varying the amount of glue; what will happen is that strength will increase up to a certain point, but beyond that more glue will actually weaken the bond.  This is because I failed to consider another variable, the size of the surface to be bonded; this and the amount of glue interact and the optimum is actually a particular thickness of the the glue applied.  I could go on about chemotherapy’s that are effective when used on their own but are useless for practical purposes because they add nothing but side effects when used with other more effective treatments.  But you hopefully get the point.  There are simplistic problems where it’s possible to isolate a single variable, but FM is really not such a case.

0

MeanOnSunday said: There are simplistic problems where it’s possible to isolate a single variable, but FM is really not such a case.

We can isolate single variables in FM. Any attribute is a variable. Luckily, FM is just a game and there is no reason to assume that a higher attribute value leads to worse results (regarding your glue example). When I read hints of the developers I always read things like this: "A goalkeeper's ability to save penalties is initially governed by his Anticipation, Reflexes, and Concentration in reacting to the moment the ball is struck. His Acceleration will help his immediate chances of reaching the ball successfully, whilst his Agility, Reflexes and Handling will ultimately determine whether or not he pulls off the save." So a higher value is always better and maybe all these attributes are additively linked like e.g. "0.4*anticipation+0.8*reflexes+0.3*concentration+0,2*acceleration+0.4*agility+0,3*handling" when there are penalties. We shouldn't forget that this is a game designed for entertainment. But these are all theories and assumptions. Back to practice:

I can beat machine learning with our single variable approach and I can prove it:

Spoiler I will make a team which is worse than the opponent team according ykykyk rating. ykykyk is the result of machine learning and ykykyk is based on 40 million matches. I will even use the tactic which the machine used for the 40 million matches (ZaZ - Blue 3.0). My team has only one information source: A few thousand matches on fm-arena tables which lead to the conclusion that pace and acceleration are the most important variables. So it is me against the machine:

Both teams have this starting players, bench players and tactic:



My Team is "Arsenal 23/24" and my opponent is "Arsenal 23/24 (2)". ykykyk ratings are higher for all "Arsenal 23/24 (2)" counterparts on the field and on the bench (except goalkeepers):





So "Arsenal 23/24 (2)" should beat my team according machine learning. These are the results after 40 matches:


"Arsenal 23/24" : "Arsenal 23/24 (2)"



This is a clear result beyond RNG. It isn't surprising when you consider how strong pace and acceleration are on the fm-tables. 40 million matches aren't enough for a machine to see this. This is the big advantage when you just change one variable to see what are the really important variables. Machine learning is a great tool when you are Google and have the ressources to simulate billion of billions of matches. We don't have these ressources so our approach is better than machine learning with 40 million matches.

1

French said: This "problem" has been present for several years



About that advice, the Recovery session is doing more harm than good there. In short, Rest only has effect if you have three sessions a day. You can test it and check with in-game editor. Three Rest sessions will recover ~50% more player condition, and 100% more Fatigue, than any number of Recovery sessions, and it does not matter how many Recoveries you put in the same day, the gain of condition and fatigue are about the same. In fact, that happens with any other training session, which means Rest only works if you have three sessions in the same day. With Super Rest, you recover around 30-40% more of each, making Recovery sessions completely pointless.

Also, keep in mind that the effects of Recovery on condition and fatigue are negligible, recovering almost the same as you would recover in Overall sessions (in other words, it is natural recovery, not anything to do with the Session). There is no evidence that it contributes in any way to reduction of injuries, other than helping to keep match sharpness (negligible difference to training sessions other than Rest).

P.S.: I also think some people are putting too much credit on a study from three years ago, considering the engine has changed a lot since then. I am not saying they do not have merit, just that results might be outdated.

1

ZaZ said: About that advice, the Recovery session is doing more harm than good there. In short, Rest only has effect if you have three sessions a day. You can test it and check with in-game editor. Three Rest sessions will recover ~50% more player condition, and 100% more Fatigue, than any number of Recovery sessions, and it does not matter how many Recoveries you put in the same day, the gain of condition and fatigue are about the same. In fact, that happens with any other training session, which means Rest only works if you have three sessions in the same day. With Super Rest, you recover around 30-40% more of each, making Recovery sessions completely pointless.

Also, keep in mind that the effects of Recovery on condition and fatigue are negligible, recovering almost the same as you would recover in Overall sessions (in other words, it is natural recovery, not anything to do with the Session). There is no evidence that it contributes in any way to reduction of injuries, other than helping to keep match sharpness (negligible difference to training sessions other than Rest).

P.S.: I also think some people are putting too much credit on a study from three years ago, considering the engine has changed a lot since then. I am not saying they do not have merit, just that results might be outdated.


correct me please if i understood it wrong, having one session of rest a day is pointless? is better to put 3 schedules a day and then find another day in the week to clump up 3 rest sessions ?

1

doru228 said: correct me please if i understood it wrong, having one session of rest a day is pointless? is better to put 3 schedules a day and then find another day in the week to clump up 3 rest sessions ?

Yes, exactly that. If you will do three rests per week, then do all in the same day, or it will have no effect. That does not include Travel when you play away from home, so if you have Travel + Rest + Rest after a match, then the Rest sessions will be wasted, recovering almost the same as if you did three sessions of Overall training (but without the attribute growth). There will be a video about that later at EBFM, so soon you will be able to see the data.

2

ZaZ said: Yes, exactly that. If you will do three rests per week, then do all in the same day, or it will have no effect. That does not include Travel when you play away from home, so if you have Travel + Rest + Rest after a match, then the Rest sessions will be wasted, recovering almost the same as if you did three sessions of Overall training (but without the attribute growth). There will be a video about that later at EBFM, so soon you will be able to see the data.

Would the rest session be of any use for players who's condition was low enough to trigger the "no pitch or gym work" intensity scheduling? I'm not totally clear what sessions count as no pitch or gym work so I've been trying Rest/Match Review/Team Bonding recently

0

ZaZ said: Yes, exactly that. If you will do three rests per week, then do all in the same day, or it will have no effect. That does not include Travel when you play away from home, so if you have Travel + Rest + Rest after a match, then the Rest sessions will be wasted, recovering almost the same as if you did three sessions of Overall training (but without the attribute growth). There will be a video about that later at EBFM, so soon you will be able to see the data.

amazing find, i think this will tottaly change the training schedules.
Do you work with EBFM ?

0

RFC said: Would the rest session be of any use for players who's condition was low enough to trigger the "no pitch or gym work" intensity scheduling? I'm not totally clear what sessions count as no pitch or gym work so I've been trying Rest/Match Review/Team Bonding recently

As far as I know, the "No Pitch or Gym" just sends players home when they are below the threshold. They appear with the Rst symbol by the side, so I assume it is the same. I will test it just in case.

doru228 said: amazing find, i think this will tottaly change the training schedules.
Do you work with EBFM ?


Yes, we have been collaborating recently, he is a very nice guy.

1

Hi.
First of all, I would like to write an amazing piece of work.
Could you create a training schedule that reflects all of this?
Cheers

0

ZaZ said: I thought "super rest" was a fitting name, because it is around 30-40% more effective than just resting, for both physical condition and fatigue. If you incorporate that in a training schedule, with three rest sessions + holiday once a week, after the match, then you will also have a nice reduction in injuries during the season, as well as allowing to repeat the main team more often, and completely avoiding fatigue. I am currently testing if it is better to do it once a week, or something like once every other week.

Hey, this is almost certainly a dumb question, but how do you holiday the players for 1 day? Like if I try to send the boys on holiday they can go for 1 week minimum. Like when I right click the player, training, training intensity, vacation. Just curious what you mean here.

Quick edit: I think it's just right clicking and resting instead of an actual vacation.

0

Perfekt said: Hey, this is almost certainly a dumb question, but how do you holiday the players for 1 day? Like if I try to send the boys on holiday they can go for 1 week minimum. Like when I right click the player, training, training intensity, vacation. Just curious what you mean here.

Quick edit: I think it's just right clicking and resting instead of an actual vacation.


Just right click the player > Training > Send to rest.

0

So I've watched the EBFM video. Kind of interesting data on double intensity. But when I try it in my own save the fitness level of my players is just that tiny bit worse that they don't fully recover in time for next match in 2 matches per week scenario. Anyone else tried the same?

0
Create an account or log in to leave a comment