Delicious said: And how is gonna works this kind of process? I mean the queque,next days you gonna give additional test or will be like something like : 3 test and 1 hall of fame 60+(Example)? Just to understand a bit more Expand
Additional testing up to 3,840 matches has the highest priority, which means we won't be testing anything else except 60+ pts tactics in HoF until they are tested at least for 3,840 matches.
Additional testing up to 5,760 matches has the lowest priority, which means it'll be delivered only when the queue is free.
We've decided to give any tactic tactic that gets 60+ points in "Hall Of Fame" addition testing up to 3,840 matches and if it holds its score then it'll be further tested up to 5,760 matches.
So don't be surprised to see some tactics get additional testing up to 3,840 matches or 5,760 matches.
Delicious said: Is fine,i don't really care about that much to go inspect every game,isn't like i don't trust the results of our loved machine. Just felt weird to drop 4 points, it feel like we are playing FM-Lotto on this patch to be honest. Expand
Yup, we've been noticing that the RNG has increased in 23.3.0 patch. Probably, +/- 0.5-1 points to what we had before.
Unforutanlty, we can't test every tactic for 5,760 matches to guaranty +/- 1 point RNG.
I don't know we might try to start testing tactics for 5,760 matches instead of 1,920 matches but I'm afraid that would make things even worse for people because the wait time will be too long.
Honestly, I really don't think we need to be that extremely precious like +/- 1 pts or some like that because in the real game you hardly can "feel" the difference between 58pts tactics and 66pts tactics.
I don't know maybe we should remove the points and leave only the rating? What do you think?
JustinCredible said: Are you sure? I checked save 1 against screenshot 1 and save 6 against screenshot 6 (and both matched) before downloading 12 (which also matched screenshot 12). The odds of 3 of 12 screenshots randomly matching 3 of 12 downloads are incredibly unlikely! Expand Yes, I'm sure.
Some game saves and screenshot numbers might match and some might not so as I said to find some specific result you have to inspect all the game saves. Don't doubt, you'll find it.
JustinCredible said: Save 12 final points from your download don't match the screenshots Expand Unforutanlty, the save numbers and the screenshot numbers might not match, which means you have to inspect every game save to find a game save that matches the result of the screenshot number #12
JustinCredible said: No beef as we are all human and all make mistakes Expand I don't upload anything. Our tactic testing is fully automated. The testing server handles everything.
Check them and if you find anything suspicious then let me know
I just want to add that +/- 10-12 RNG between single runs isn't a rare thing, it really happens sometimes.
Also, 4 points correction between 1,920 matches and 5,760 matches is quite rare thing and most of time it's 1-2 points correction but as you can see even 4 points correction might happen. What can I say? Welcome to FM world!
P.S. Unforutanlty, the save numbers and the screenshot numbers might not match, which means you have to inspect every game save to find a game save that matches the result of the screenshot number #12
BulldozerJokic said: If the "front 3" are AMs, there will be slower (13) players playing there. I don't understand why @Zippo doesn't see a problem here. Expand In the actual game natural Wingers are faster than natural Attacking Midfielders and our tactic testing league keeps that balance.
When you play the actual game and you use a strikerless tactic then you put your strikers/wingers to play at the AMs and your strikers/wingers won't have "Natural" rating for the AM position but if you follow the link below then you can check how the playing position rating affects the result - https://fm-arena.com/table/19-playing-position-testing/
BulldozerJokic said: It can be solved by adding AM natural position to the strikers. But that will require re-testing everything Expand We always want to be sure in the reliability of our testing league and of course, we've already done what you suggested.
We added the strikers "natural" playing position rating and re-tested some highest scoring strikerless tactics.
TommyToxic said: Well explained as always, brings a lot of clarity to the topic.
Do you think the CBs in the test save are suitable to play in a 3ATB system? I know CBs in general have low dribbling, but I think when FM players play 3ATB they prioritize attributes differently on wide CBs than in a traditional 4ATB tactic.
Wide center backs go more forward and also have to cover space out wide so having slightly higher dribbling and pace is normal, and maybe a bit less strength/jumping reach to compensate. Something like Ben White or Tomiyasu.
Is it sensible to include a second type of CB (wide CB) in the test save teams, or does that make testing more difficult without necessarily making it better for testing?
Same goes for AMs as they are very well rounded players in the test save, but in normal FM games they are usually much more attacking players similar to inside forwards or false nines. Expand Once more, the players in our testing league are very well balanced and the testing environment represents the actual game very well.
Most importantly, the testing environment in our tactic testing is much more "fair" for different tactical approaches than if we tested with real teams from the actual game, testing tactics with real teams would be the easiest way to test but also it would be much more less "fair".
Also, another advantage of our tactic testing league is that it's free of occasionally terrible choices that the Assistant Manager can make when he picks the team to test different tactics so it ensures that some tactics won't be tested with more inferior players than other tactics.
kjordafen said: What are the pace and accelation for attaching midfields in these tests? If these are below 15 I think the test is somewhat unrealistic. Expand
In the actual game an average natural Central Midfielder has 13 Acceleration and 13 Pace.
In the actual game an average natural Winger/Striker has 15 Acceleration and 15 Pace.
And that balance we have in our testing league.
Also, here are some interesting facts there're only 2 natural Central Midfielders who have Acceleration and Pace higher than "17" in the actual game.
But there're about 100 Wingers who have Acceleration and Pace higher than "17" in the actual game.
kjordafen said: I agree with you that this is how it normally is, but many people can develop and find fast attaching midfields for example. Expand
Training a new playing position from zero level might take from 3 to 6 seasons and that in case if a player's Versality attribute is high enough to learn a new position at all. So we just can't consider training a new position from zero level as a viable option, also taking into consideration that there're are many people who prefer playing only 1-3 seasons keeping mostly the default team or who change teams after 2-3 seasons.
kjordafen said: In most fm-saves you could easily buy/find/use players with 15 pace and accelaration. If you are in any top 5 league you can find it straight away, and in most second tier leagues you will easily be Expand
Of course, in the actual game you can find/buy/use a natural Central Midfielder with 15 Acceleration and 15 Pace but the same time there're Wingers/Strikers who have 20 Acceleration and 20 Pace in the actual game. So for the actual game 15 Acceleration and 15 Pace aren't extreme values but in our tactic testing league 15 Acceleration and 15 Pace are the highest value for Acceleration and Pace attributes which means when a player having 13 Acceleration and 13 Pace in our tactic testing league then his speed is almost at the top level so the midfielders in our tactic testing league are very well balanced and that balance represent very well what we have in the actual game.
Guys, I forgot to mention that if there're 3 central defenders in your tactic then Inverted Wing-Backs are also hard-coded to play as regular Wing-backs with "Cuts Inside" and "Stay Narrower" PIs
dzek said: @Zippo, did you change anything on FM-Arena Testing League since you gave us that savegame? If yes, can we have a newly savegame file? It will be very helpful Expand
chalengr said: I found in fm-arena‘s testbase striker’s CA is much higher than AMC（acceleration and pace are higher too），I want to know if the difference is from this@Zippo （my English is not good） Expand
The Match Engine don't use CA to calculate the result of matches so CA is irrelevant in that terms.
The Match Engine uses the actual attributes to calculate the result of matches.
For example,
The striker above has 140CA
140CA Player
But if we add to him additional playing positions such as AMCL/AMCR/AMC/ML/MR/MC then his CA will increase from 140CA to 164CA
164CA Player
As you can see the player still has the same attributes but his CA has significantly increased from 140CA to 164CA but in matches his efficiency hasn't changed because his attributes haven't changed. In other words, even his CA has increased from 140CA to 164CA but he won't play better in matches because his attributes haven't changed.
So only by looking at the CA you can't say that one player is better than other, also, the attributes have a different CA cost for different positions. For example, in general the attributes cost much less CA for central defenders than for any other positions so you can't compare players only by looking at their CA. In other words, in the actual game in general Central Defenders tend to have lower CA comparing to Striker but it doesn't make them less efficient in matches than Strikers. it's just that for a Central Defender the attributes are "cheaper" in terms of CA than for any other position.
BulldozerJokic said: Well, looks like it's not only the strickers that are faster, but also the wingers. Meaning that any narrow-shaped and some 3 atb tactics might be underscoring points like crazy Expand
Our tactic testing leagues is designed to represent the actual game and not some "fantasy" custom football world.
Here're some facts from the actual game:
- in the actual game Central Defenders tend to have a very low "Finishing" attribute, it isn't higher than "6-7" so if in your tactic testing league you set the Finishing attribute of your Central Defenders higher than that, for example, you give them "16" Finishing then your tactic testing league will be represent the actual game very poorly.
- in the actual game Central Defenders tend to have a very low "Dribbling" attribute, it isn't higher than "8-9" so if in your tactic testing league you set the Dribbling attribute of your Central Defenders higher than that, for example, you give them "16" Dribbling attribute then your tactic testing league will be represent the actual game very poorly.
- Winger/Inside Forward/Full Backs/Strikers tend to be faster than other positions in the actual game so that's why in our tactic testing league Winger/Inside Forward/Strikers are faster than other positions.
TommyToxic said: I use the same setup just with IWB instead of FB. Is the consensus that FB with play narrow and run inside just works better than IWB? Expand
In this setup(where both DM slots are occupied) FBs that have the same PIs as IWBs would work 99% the same.
Delicious said: @Zippo question if i have a WBR/L (position) and i have it on IWBS how works the code? They move inside if there is no CM role? Just to understand if they keep the same "logic" even on WBx position Expand
In case you put IWBs into WBR/L slots then it still only checks the state of DMs slots.
The sate of CMs slots is irrelevant in both cases.
Yarema said: Surely on the side of lets say volante or RPM that space at least occasionally clears up to move into? Expand
Delicious said: You can make wbs sitting narrow as well but one cut inside and the other cut wide with the ball. About iwbs i was trying to test if the would over-load as well. Expand
The role and duty of the adjacent DM position doesn't matter, it doesn't matter whether it's Vol(A) or RMP(S) or any other role, it's hard coded that if there's no free DM slot near the Inverted Wing-Back then he act as a regular Wing-Back with "Cut Inside" and "Sit Narrow" PIs so the unique Inverted Wing-Back behavior is lost in this case.
Chewbacca said: sitting narrow is the only what distinguishes IWBs from WBs? Expand
Mostly, yes.
Look at the screenshot below. You see how narrow Shaw ( Inverted Wing-Back ) sits.
When "in possession" he plays like a central midfielder or central defensive midfielder and it's an unique feature of Inverted Wing-Back role which can't be recreated in any other full back's role.
And that unique behavior of Inverted Wing-Back can only be activated if a specific condition is met and that specific condition is a free DM slot near the Inverted Wing-Back.
Just to understand a bit more
Additional testing up to 3,840 matches has the highest priority, which means we won't be testing anything else except 60+ pts tactics in HoF until they are tested at least for 3,840 matches.
Additional testing up to 5,760 matches has the lowest priority, which means it'll be delivered only when the queue is free.
We've decided to give any tactic tactic that gets 60+ points in "Hall Of Fame" addition testing up to 3,840 matches and if it holds its score then it'll be further tested up to 5,760 matches.
So don't be surprised to see some tactics get additional testing up to 3,840 matches or 5,760 matches.
Cheers.
Yup, we've been noticing that the RNG has increased in 23.3.0 patch. Probably, +/- 0.5-1 points to what we had before.
Unforutanlty, we can't test every tactic for 5,760 matches to guaranty +/- 1 point RNG.
I don't know we might try to start testing tactics for 5,760 matches instead of 1,920 matches but I'm afraid that would make things even worse for people because the wait time will be too long.
Honestly, I really don't think we need to be that extremely precious like +/- 1 pts or some like that because in the real game you hardly can "feel" the difference between 58pts tactics and 66pts tactics.
I don't know maybe we should remove the points and leave only the rating? What do you think?
Yes, I'm sure.
Some game saves and screenshot numbers might match and some might not so as I said to find some specific result you have to inspect all the game saves. Don't doubt, you'll find it.
Unforutanlty, the save numbers and the screenshot numbers might not match, which means you have to inspect every game save to find a game save that matches the result of the screenshot number #12
JustinCredible said: No beef as we are all human and all make mistakes
I don't upload anything. Our tactic testing is fully automated. The testing server handles everything.
Sadly, I haven't got free time for it right now but I can give you the end game saves
https://drive.google.com/file/d/1lgEn8vvMqsLysp3WHR7h_ef_wphT_JBF/view?usp=sharing
Check them and if you find anything suspicious then let me know
I just want to add that +/- 10-12 RNG between single runs isn't a rare thing, it really happens sometimes.
Also, 4 points correction between 1,920 matches and 5,760 matches is quite rare thing and most of time it's 1-2 points correction but as you can see even 4 points correction might happen. What can I say? Welcome to FM world!
P.S. Unforutanlty, the save numbers and the screenshot numbers might not match, which means you have to inspect every game save to find a game save that matches the result of the screenshot number #12
Ok. I'll put it in the queue for "tornado" test.
Hi,
I've just checked and found that the 3 latest tested tactics were accidently tested twice as many as they should be tested.
But I think it is not a bad thing, the results are just more accurate now.
No, the above only applies for Inverted Wing-Backs.
Here're end game saves - https://drive.google.com/file/d/1TouuZEnFGiwT7KhJbaJT90YA_5h1Zijr/view?usp=sharing of your strikerless tactic test - https://fm-arena.com/thread/4648-cowboyball2-strikerless/ you can get any information you like from it.
BulldozerJokic said: If the "front 3" are AMs, there will be slower (13) players playing there. I don't understand why @Zippo doesn't see a problem here.
In the actual game natural Wingers are faster than natural Attacking Midfielders and our tactic testing league keeps that balance.
When you play the actual game and you use a strikerless tactic then you put your strikers/wingers to play at the AMs and your strikers/wingers won't have "Natural" rating for the AM position but if you follow the link below then you can check how the playing position rating affects the result - https://fm-arena.com/table/19-playing-position-testing/
BulldozerJokic said: It can be solved by adding AM natural position to the strikers. But that will require re-testing everything
We always want to be sure in the reliability of our testing league and of course, we've already done what you suggested.
We added the strikers "natural" playing position rating and re-tested some highest scoring strikerless tactics.
For example, we re-tested this strikers tactic - https://fm-arena.com/thread/4644-diva-meet-extreme-strikerless-iii-4330/
And the difference was -/+ 1-2 points, which means the highest score was 58 pts and the lowest score 54 pts.
If you ask me then -/+ 1-2 points doesn't change anything for strikerless tactics.
Do you think the CBs in the test save are suitable to play in a 3ATB system? I know CBs in general have low dribbling, but I think when FM players play 3ATB they prioritize attributes differently on wide CBs than in a traditional 4ATB tactic.
Wide center backs go more forward and also have to cover space out wide so having slightly higher dribbling and pace is normal, and maybe a bit less strength/jumping reach to compensate. Something like Ben White or Tomiyasu.
Is it sensible to include a second type of CB (wide CB) in the test save teams, or does that make testing more difficult without necessarily making it better for testing?
Same goes for AMs as they are very well rounded players in the test save, but in normal FM games they are usually much more attacking players similar to inside forwards or false nines.
Once more, the players in our testing league are very well balanced and the testing environment represents the actual game very well.
Most importantly, the testing environment in our tactic testing is much more "fair" for different tactical approaches than if we tested with real teams from the actual game, testing tactics with real teams would be the easiest way to test but also it would be much more less "fair".
Also, another advantage of our tactic testing league is that it's free of occasionally terrible choices that the Assistant Manager can make when he picks the team to test different tactics so it ensures that some tactics won't be tested with more inferior players than other tactics.
In the actual game an average natural Central Midfielder has 13 Acceleration and 13 Pace.
In the actual game an average natural Winger/Striker has 15 Acceleration and 15 Pace.
And that balance we have in our testing league.
Also, here are some interesting facts there're only 2 natural Central Midfielders who have Acceleration and Pace higher than "17" in the actual game.
But there're about 100 Wingers who have Acceleration and Pace higher than "17" in the actual game.
kjordafen said: I agree with you that this is how it normally is, but many people can develop and find fast attaching midfields for example.
Training a new playing position from zero level might take from 3 to 6 seasons and that in case if a player's Versality attribute is high enough to learn a new position at all. So we just can't consider training a new position from zero level as a viable option, also taking into consideration that there're are many people who prefer playing only 1-3 seasons keeping mostly the default team or who change teams after 2-3 seasons.
kjordafen said: In most fm-saves you could easily buy/find/use players with 15 pace and accelaration. If you are in any top 5 league you can find it straight away, and in most second tier leagues you will easily be
Of course, in the actual game you can find/buy/use a natural Central Midfielder with 15 Acceleration and 15 Pace but the same time there're Wingers/Strikers who have 20 Acceleration and 20 Pace in the actual game. So for the actual game 15 Acceleration and 15 Pace aren't extreme values but in our tactic testing league 15 Acceleration and 15 Pace are the highest value for Acceleration and Pace attributes which means when a player having 13 Acceleration and 13 Pace in our tactic testing league then his speed is almost at the top level so the midfielders in our tactic testing league are very well balanced and that balance represent very well what we have in the actual game.
No, there haven't been any changes.
（my English is not good）
The Match Engine don't use CA to calculate the result of matches so CA is irrelevant in that terms.
The Match Engine uses the actual attributes to calculate the result of matches.
For example,
The striker above has 140CA
140CA Player
But if we add to him additional playing positions such as AMCL/AMCR/AMC/ML/MR/MC then his CA will increase from 140CA to 164CA
164CA Player
As you can see the player still has the same attributes but his CA has significantly increased from 140CA to 164CA but in matches his efficiency hasn't changed because his attributes haven't changed. In other words, even his CA has increased from 140CA to 164CA but he won't play better in matches because his attributes haven't changed.
So only by looking at the CA you can't say that one player is better than other, also, the attributes have a different CA cost for different positions. For example, in general the attributes cost much less CA for central defenders than for any other positions so you can't compare players only by looking at their CA. In other words, in the actual game in general Central Defenders tend to have lower CA comparing to Striker but it doesn't make them less efficient in matches than Strikers. it's just that for a Central Defender the attributes are "cheaper" in terms of CA than for any other position.
BulldozerJokic said: Well, looks like it's not only the strickers that are faster, but also the wingers. Meaning that any narrow-shaped and some 3 atb tactics might be underscoring points like crazy
Our tactic testing leagues is designed to represent the actual game and not some "fantasy" custom football world.
Here're some facts from the actual game:
- in the actual game Central Defenders tend to have a very low "Finishing" attribute, it isn't higher than "6-7" so if in your tactic testing league you set the Finishing attribute of your Central Defenders higher than that, for example, you give them "16" Finishing then your tactic testing league will be represent the actual game very poorly.
- in the actual game Central Defenders tend to have a very low "Dribbling" attribute, it isn't higher than "8-9" so if in your tactic testing league you set the Dribbling attribute of your Central Defenders higher than that, for example, you give them "16" Dribbling attribute then your tactic testing league will be represent the actual game very poorly.
- Winger/Inside Forward/Full Backs/Strikers tend to be faster than other positions in the actual game so that's why in our tactic testing league Winger/Inside Forward/Strikers are faster than other positions.
and so on...
I hope this helps.
Cheers.
Can we expect to see Mr.Tornado visiting the top tactics?
Hi,
Probably, later today/tomorrow the top tactics will get an additional testing to find out whether the result was a fluke or not.
In this setup(where both DM slots are occupied) FBs that have the same PIs as IWBs would work 99% the same.
If the "Edit" button is available then feel free to edit whatever you want
In case you put IWBs into WBR/L slots then it still only checks the state of DMs slots.
The sate of CMs slots is irrelevant in both cases.
Delicious said: You can make wbs sitting narrow as well but one cut inside and the other cut wide with the ball.
About iwbs i was trying to test if the would over-load as well.
The role and duty of the adjacent DM position doesn't matter, it doesn't matter whether it's Vol(A) or RMP(S) or any other role, it's hard coded that if there's no free DM slot near the Inverted Wing-Back then he act as a regular Wing-Back with "Cut Inside" and "Sit Narrow" PIs so the unique Inverted Wing-Back behavior is lost in this case.
Chewbacca said: sitting narrow is the only what distinguishes IWBs from WBs?
Mostly, yes.
Look at the screenshot below. You see how narrow Shaw ( Inverted Wing-Back ) sits.
When "in possession" he plays like a central midfielder or central defensive midfielder and it's an unique feature of Inverted Wing-Back role which can't be recreated in any other full back's role.
And that unique behavior of Inverted Wing-Back can only be activated if a specific condition is met and that specific condition is a free DM slot near the Inverted Wing-Back.