White Europe said: FmArena responsible for canceling FM25, poor SI needs to completely re-work match engine to actually have some use for 90% of attributes that didnt do anything in the game for last 3 editions. Expand
I think the majority of attributes have done nothing since the dawn of time
gallina said: Separately, can 'OK' tactics be workable in-game? Expand
This website is essentially just a maximisation platform for FM tactics, to try and find the absolute best possible solution and have some statistical data to back that up. As you can see from the tactics table, these have all been tested for several thousand matches using the same players in a very controlled environment.
When you play the game yourself, you are only likely to be playing fifty games in a season, using players who may specialise in one role over another, and you also have a huge risk of injury. That's not to mention the fact that the RNG over fifty games is significantly higher than it is over five thousand. Therefore, if the universe aligns, a possession-based 4-3-3 could come out on top against a meta gegenpress 4-2-4 despite the 4-2-4 being optimal for the way that this match engine works. Over five thousand games it won't be better, but you could well find yourself winning the league with an 'OK' tactic.
It is important that in pursuit of maximisation that you don't lose what the purpose of playing the game is, fun.
flob said: I don't know man. I am in a unemployed save in my 2nd season where I brought a J3 team to J2 league and now at the start of the season, almost every player is unhappy with the training (I am doing L5 from the spreadsheet). Majority of the players, aswell the U21 squad, are unhappy with the training. It's like a virus. Expand
Well what I found since posting that is that you will be fine as long as you make sure that training responsibility is set to the assistant. If not, they will get upset almost immediately, all of mine were upset about the standard of strength coaching
azsumnasko said: I'm not sure how to prove it, but there is negative relation between all these trainings and the match performance. Example is week defense with good fast and tall defenders and lame attack with good fast forwards. Do you think this observation is correct? Maybe adding attack shadow play and defense equivalent?! Expand
I don't think the training that they do has an impact on match performance other than some notable things left out of the training that DO improve results (things like happiness and team cohesion)
It seems it does though, hence me wondering what I can do about it. Expand
What I'm testing atm is whether leaving it to your AM and then pressing "do it once" on general team training and "take control of this player" in individual training actually stops them being quite so upset. If you're always winning then it isn't a huge problem, but one bad result can shift things quickly because they all already have a gripe with you.
We have to find a balance between good development, match results, and injury mitigation
mmigueis said: I have a question for the top tactic makers: why are you convinced that having goalkeepers with PI “tackle harder” is best?? 🤣
Given the amount of uncertainty (“RNG”) in tactic testing in FM and the few occasions in game that a keeper would perform a tackle, i highly doubt that any tests would be able to accurately detect whether GK with tackle harder is better than without it…. I suspect everyone just uses it because top tactics use tackle harder in all other positions, so why not for gks too. But maybe there is more to it that I haven’t seen 😄 Expand
@dzek had some numbers on this and indeed it does seem to have no effect. I do it because it's fun to think about goalkeepers two-footing people
Yarema said: Maybe we could take a step back and not conclude everything based on like 0,1 differences. These tests are great but there is still plenty of error in them due to sample size.
Compare for example H7 and S7. Huge difference for no apparent reason and even in the opposite direction as most others would suggest. Expand
Poses an interesting question about RNG which we already know is present in the match engine itself BUT we don't know if RNG affects training in the same way (unless we do, feel free to correct me)
I think a lot of us have become blinded by searching for this mythical meta. At the end of the day, the point of the game is to have fun. Go use the 3-4-2-1 or any other system that you ordinarily think won't work because in the end it might
dzek said: I don't believe that will ever occur because it would essentially negate the concept of roles. In my view, the most crucial aspect is to ensure complete transparency regarding the impact of every instruction in the game, allowing us full control over our team. Currently, we are left to speculate about the effects of each instruction and the roles or positions they influence.
It reminds me of Call of Duty, where each weapon is accompanied by a detailed box with bars. This could be implemented for each role, with the ability to adjust each instruction by altering the Player Instructions or Team Instructions.
For instance, if I use two winger roles with a predefined "Dribble More" instruction and I apply a "Dribble Less" team instruction that affects them, it should be reflected by a corresponding decrease in the Dribbling attribute bar. Expand
That's a shout. I was thinking of a PES-style tactic creator (I don't know if they still use it now). This would get rid of the point of roles but I think roles have very limited usage in real life too, terms have been coined (i.e. inverted full-back) but nobody ever looks at people like Haaland and say "oh he's clearly an advanced forward". Somehow, FM is both too complicated and not complicated enough.
But yeah there absolutely needs to be more transparency about how the game actually works, people should not have to look online to gain a basic idea of how to play it.
dzek said: I always wondered about the double instructions and I always wanted to try them but one day I forgot and the next day I didn't have time. Now I think things are a bit clearer.
I too hope that in FM25 a lot of things will change and indeed the meta instructions have been a bit monotonous in recent years. But I'm keeping a low profile until we see them in action. Expand
The dream is to have an OOP system and an in-possession system with all of the roles fully customisable. We sort of have that now but there are many limitations to it
I have run more tests on this experiment (8.800 matches in total) and it seems to repeat with very little variation in values. I didn't see any changes from what I wrote in my previous post.
Here is the final table: (including the one above along with it)
FINAL DATA TABLEDribbles Made / 90
I have made also some experiments with "Get Stuck In" and "Tackle Harder". Here are the results:
DATA TABLETackles Made / 90
As you can see from these tests I come to the following conclusions:
1. Both instructions do nothing to GKs. 2. "Get Stuck In" maximizes tackling for defensive positions (DC/DRL/DM), so there is no need for "Tackle Harder". 3. Having both instructions enabled you will not see significant changes at any position except AMRL positions and even at those positions there will be little to no difference. Expand
Fascinating. I'm somewhat glad that tackle harder doesn't appear to make much of a difference as it has always appeared to me that TIs and PIs that do the same thing should not be able to add together and 'maximise' things like aggression and dribbling.
Hopefully for FM25 we see plenty of tactical changes as I am a little bored of the way that the meta has been working for the last few years, but I'm not holding out a lot of hope on that front.
I was curious to know what positions Run At Defense and Dribble More affects, as well as if you enable both instructions together, if they have any more effect on players.
Here are the results I found.
DATA TABLEDribbles Made / 90
As you can see from these tests Ι come to the following conclusions:
1. Both instructions do work. 2. 'Run At Defence' does not affect the DC position and the DM position very little. 3. Only 'Dribble More' works on DC position. 4. DM position is not significantly affected by the two instructions. 5. Having both instructions enabled will not see significant changes in any position except AMRL positions.
dzek said: The more you "influence" an environment, the more you will get certain results that sometimes lead down the wrong paths.
This sentence was for those who create databases to test tactics. So I mean that the results of the FM Arena testing league and others (including mine) do not replicate a regular season at all. It is simply the "score" of a tactic within the test league and nothing more. I'm with you on this one. Expand
dzek said: What were your results with this tactic?
The reason I'm asking is because, I think this year there are more combinations you can do to get results and/or trophies. And another reason is that I'm not so sure about FM Arena results compared to a regular season game. The more you "influence" an environment, the more you will get certain results that sometimes lead down the wrong paths. To be fair and not to be misunderstood this applies to my database and many others.
I also find a practice that TFF (for those who know him) used to do in his own databases, he tested his tactics, at max in 500 matches and his tactics were 95% equivalent to the normal game environment.
In recent years, I can't say the same thing has been happening. As well as not forgetting that since FM23, the game developers have changed/upgraded the AI within the ME and it's more reactive. Expand
It did okay as it still included various things such as hard tackling and high pressing and got 105 points with Celtic in the Scottish Prem. However, it did struggle when I tested it with Manchester City. I do totally agree that results improve when you affect it yourself (i.e. adding a team bonding session every week increases team cohesion which improves results) but I think it is likely that the top tactics here would perform best in the game too, taking RNG into account.
I've practically spent the last week playing a Xabi Alonso 3-4-3 type of tactic without all of the added instructions such as dribble more or focusing the ball down the flanks. While it is fun on a single-player save, it would never be as effective as the highest 4-2-3-1 and 4-2-4 tactics on here. I do love FM, but the tactical rigidity of the "meta" irritates me
I think the majority of attributes have done nothing since the dawn of time
This website is essentially just a maximisation platform for FM tactics, to try and find the absolute best possible solution and have some statistical data to back that up. As you can see from the tactics table, these have all been tested for several thousand matches using the same players in a very controlled environment.
When you play the game yourself, you are only likely to be playing fifty games in a season, using players who may specialise in one role over another, and you also have a huge risk of injury. That's not to mention the fact that the RNG over fifty games is significantly higher than it is over five thousand. Therefore, if the universe aligns, a possession-based 4-3-3 could come out on top against a meta gegenpress 4-2-4 despite the 4-2-4 being optimal for the way that this match engine works. Over five thousand games it won't be better, but you could well find yourself winning the league with an 'OK' tactic.
It is important that in pursuit of maximisation that you don't lose what the purpose of playing the game is, fun.
Well what I found since posting that is that you will be fine as long as you make sure that training responsibility is set to the assistant. If not, they will get upset almost immediately, all of mine were upset about the standard of strength coaching
Do you think this observation is correct?
Maybe adding attack shadow play and defense equivalent?!
I don't think the training that they do has an impact on match performance other than some notable things left out of the training that DO improve results (things like happiness and team cohesion)
It seems it does though, hence me wondering what I can do about it.
What I'm testing atm is whether leaving it to your AM and then pressing "do it once" on general team training and "take control of this player" in individual training actually stops them being quite so upset. If you're always winning then it isn't a huge problem, but one bad result can shift things quickly because they all already have a gripe with you.
We have to find a balance between good development, match results, and injury mitigation
Given the amount of uncertainty (“RNG”) in tactic testing in FM and the few occasions in game that a keeper would perform a tackle, i highly doubt that any tests would be able to accurately detect whether GK with tackle harder is better than without it…. I suspect everyone just uses it because top tactics use tackle harder in all other positions, so why not for gks too. But maybe there is more to it that I haven’t seen 😄
@dzek had some numbers on this and indeed it does seem to have no effect. I do it because it's fun to think about goalkeepers two-footing people
Compare for example H7 and S7. Huge difference for no apparent reason and even in the opposite direction as most others would suggest.
Poses an interesting question about RNG which we already know is present in the match engine itself BUT we don't know if RNG affects training in the same way (unless we do, feel free to correct me)
It reminds me of Call of Duty, where each weapon is accompanied by a detailed box with bars. This could be implemented for each role, with the ability to adjust each instruction by altering the Player Instructions or Team Instructions.
For instance, if I use two winger roles with a predefined "Dribble More" instruction and I apply a "Dribble Less" team instruction that affects them, it should be reflected by a corresponding decrease in the Dribbling attribute bar.
That's a shout. I was thinking of a PES-style tactic creator (I don't know if they still use it now). This would get rid of the point of roles but I think roles have very limited usage in real life too, terms have been coined (i.e. inverted full-back) but nobody ever looks at people like Haaland and say "oh he's clearly an advanced forward". Somehow, FM is both too complicated and not complicated enough.
But yeah there absolutely needs to be more transparency about how the game actually works, people should not have to look online to gain a basic idea of how to play it.
I too hope that in FM25 a lot of things will change and indeed the meta instructions have been a bit monotonous in recent years. But I'm keeping a low profile until we see them in action.
The dream is to have an OOP system and an in-possession system with all of the roles fully customisable. We sort of have that now but there are many limitations to it
I have run more tests on this experiment (8.800 matches in total) and it seems to repeat with very little variation in values. I didn't see any changes from what I wrote in my previous post.
Here is the final table: (including the one above along with it)
FINAL DATA TABLEDribbles Made / 90
I have made also some experiments with "Get Stuck In" and "Tackle Harder".
Here are the results:
DATA TABLETackles Made / 90
As you can see from these tests I come to the following conclusions:
1. Both instructions do nothing to GKs.
2. "Get Stuck In" maximizes tackling for defensive positions (DC/DRL/DM), so there is no need for "Tackle Harder".
3. Having both instructions enabled you will not see significant changes at any position except AMRL positions and even at those positions there will be little to no difference.
Fascinating. I'm somewhat glad that tackle harder doesn't appear to make much of a difference as it has always appeared to me that TIs and PIs that do the same thing should not be able to add together and 'maximise' things like aggression and dribbling.
Hopefully for FM25 we see plenty of tactical changes as I am a little bored of the way that the meta has been working for the last few years, but I'm not holding out a lot of hope on that front.
I was curious to know what positions Run At Defense and Dribble More affects, as well as if you enable both instructions together, if they have any more effect on players.
Here are the results I found.
DATA TABLEDribbles Made / 90
As you can see from these tests Ι come to the following conclusions:
1. Both instructions do work.
2. 'Run At Defence' does not affect the DC position and the DM position very little.
3. Only 'Dribble More' works on DC position.
4. DM position is not significantly affected by the two instructions.
5. Having both instructions enabled will not see significant changes in any position except AMRL positions.
Data of the table
Thoughts?
Only just seen this, fantastic work
This sentence was for those who create databases to test tactics. So I mean that the results of the FM Arena testing league and others (including mine) do not replicate a regular season at all. It is simply the "score" of a tactic within the test league and nothing more. I'm with you on this one.
Ahhhh I get you
The reason I'm asking is because, I think this year there are more combinations you can do to get results and/or trophies. And another reason is that I'm not so sure about FM Arena results compared to a regular season game. The more you "influence" an environment, the more you will get certain results that sometimes lead down the wrong paths. To be fair and not to be misunderstood this applies to my database and many others.
I also find a practice that TFF (for those who know him) used to do in his own databases, he tested his tactics, at max in 500 matches and his tactics were 95% equivalent to the normal game environment.
In recent years, I can't say the same thing has been happening. As well as not forgetting that since FM23, the game developers have changed/upgraded the AI within the ME and it's more reactive.
It did okay as it still included various things such as hard tackling and high pressing and got 105 points with Celtic in the Scottish Prem. However, it did struggle when I tested it with Manchester City. I do totally agree that results improve when you affect it yourself (i.e. adding a team bonding session every week increases team cohesion which improves results) but I think it is likely that the top tactics here would perform best in the game too, taking RNG into account.