2. Now, there are four Manchester City replicants Manchester City's players are already at the top of the game, Their ability is high , so It is impossible not high enough to support "high Flair " or "high Technique".
They use the same formation, Default 433 ball control Tactics, Team A's starting morale is a little bit higher, This is used to increase the degree of differentiation and avoid mixing the data together
Each person also has attributes corresponding to their duties according to their position, It didn't say that the attributes didn't match his duties, his position, his role
14.6/12.3≈120% There's always higher variance to higher Flair , Which means Flair's true meaning may be "unexpectedness." it causes both unexpected good and unexpected bad, But the average goal difference has gone down
As for Technique , since its Goal difference from the standard is negative, it is a negative attribute
3.Four teams of Manchester United replicants
Number of test match: 900 Also exactly replicating in-game player attributes. Standard Group : Goal Difference =4.1, variance =16.4 Flair +10: Goal difference =2.7, variance =19.3
19.3/16.4≈118% They use the same formation, Default 442 defend Tactics, The technique is also below Standard ,Under different conditions, higher Flair shows higher variance , higher Flair and higher technique show lower Goal Difference
So, I'm guessing now, Flair and Technique are negative attributes , Flair is "unexpectedness."
In my FM23 experiment Flair also had a little negative correlation with player rating while Technique was the lowest coefficient non-hidden attribute (except Flair).
The findings you have made regarding flair are not that surprising to me, but the findings regarding technique are genuinely bizarre.
- Flair does not have a weight in a player's current ability (CA); therefore, it makes sense that it may have positive effects in some cases and negative effects in others. From what I've understood of game manuals over the years, flair is supposed to make players try more unexpected and harder moves, which have higher likelihood of success the more technically proficient players are. It is perhaps a bit surprising that an increase in flair was not helpful with Man City players, but overall it makes sense that more flair does not increase performance in a clear-cut fashion.
- Technique does have a weight in a player's current ability, and all the descriptions I have ever seen of the attribute through the years would indicate that higher technique should have a clear net positive on player's performance. If your test results generalize, this would seem like a clear case of SI not properly testing how the attributes work in the game's match engine. The only rationalization I can come up with for the results you are finding is that players with higher technique try harder moves, but those moves are generally worse than trying simpler, less technical moves (e.g., moves based on pace and acceleration) even when players have high technique. This would suggest that perhaps not only higher technique is not giving sufficient boost but also that the "decisions" attribute may not be working properly (as most tests have found so far, decisions provides a unrealistic small boost to players (in some tests, higher decisions even seems associated with a reduction in performance!!)).
In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad.
bigloser said: In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad. Expand
——did you lower their technique no. + 10, if + 10 and result greater than 20, then = 20 (This means that the average actual increase is less than 10.)
I am busy now , I will try other conditions when I am free.
thank you and Your information Some community players speculate: Whether higher Technique/Flair in specific positions (e.g., midfield, or defenders) has led to a drop in overall goal difference, or whether it has an effect in all positions,
I did some tests with flair but didn't really notice a negative correlation. 4 team league, teams B, C, D have all players with 20 in all normal attributes. Team A players have everything on 20 except flair, which is on 1. All attributes are frozen (this means that the attributes stay on the same level all season). All players are proficient in all positions. The formation used was a 4-1-1-3-1, for all teams (FB (De), CD (De), DM (Su), CM (Su), W (At), AM (Su), AF (At)).
I ran 10 seasons (the first season 10 times, to be precise). Team A won the league twice, got third a couple of times and second a couple too. The goal difference wasn't really relevant either: 19, 15, -6,-1, -16, -10, -2, 15, -6, 12.
Didn't test combinations or other attributes, though.
Of course, 10 seasons amounts to a very small number of matches in a league with 4 teams (300 matches per team), so more testing is surely required.
caffeiner said: I did some tests with flair but didn't really notice a negative correlation. 4 team league, teams B, C, D have all players with 20 in all normal attributes. Team A players have everything on 20 except flair, which is on 1. All attributes are frozen (this means that the attributes stay on the same level all season). All players are proficient in all positions. The formation used was a 4-1-1-3-1, for all teams (FB (De), CD (De), DM (Su), CM (Su), W (At), AM (Su), AF (At)).
I ran 10 seasons (the first season 10 times, to be precise). Team A won the league twice, got third a couple of times and second a couple too. The goal difference wasn't really relevant either: 19, 15, -6,-1, -16, -10, -2, 15, -6, 12.
Didn't test combinations or other attributes, though.
Of course, 10 seasons amounts to a very small number of matches in a league with 4 teams (300 matches per team), so more testing is surely required. Expand
Thank you. If I have time I'll keep expanding the sample to see if it's random, if it's due to certain conditions, or if it's true
harvestgreen22 said: Thank you. If I have time I'll keep expanding the sample to see if it's random, if it's due to certain conditions, or if it's true Expand
I feel there is somepositive correlation using technique, but haven't been able to point exactly what is it.
I've updated my testing database. So 4 teams, all playing a 4-3-3 without any instructions (FB (Su), CD (De), DM (De), BBM (Su), CM (Su), W (Su), AF (At)). Each team plays 30 matches. Injuries are disabled. Attributes are frozen, and testing is done on holiday.
Teams 2, 3, 4 have technique on 1 for all line players, and all other attributes on 20. Team 1 has everything at 20.
The GK for all teams have everything at 10.
I'm still running the tests but right now the difference seems too big for technique to be negative or to not matter.
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner.
It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results.
Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them.
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them. Expand
Thanks for your work, I'm also slowly trying a new initial condition, Four teams use replicants, different 4-3-3, then test what the goal difference is without changing the attributes, then test what the goal difference is with -10 Technique 1,500 games tested now. I expect to increase it to about 5000.
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner. Expand
Thanks for your work, I'm also slowly trying a new initial condition, Four teams use replicants, different 4-3-3, then test what the goal difference is without changing the attributes, then test what the goal difference is with -10 Technique 1,500 games(match) tested now. I expect to increase it to about 5000.
mmigueis said: It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results. Expand
This is true. For the mean/std metrics I could've averaged everyone that isn't the Team 1. I'll keep this in mind for other analysis. But I think it's good to have everyone separated for the other tests.
I've done similar testing for Decisions. While I didn't have time to generate all graphs for it (yet), the results seem similar regarding position (this means, it won the league just as much as Technique), however the correlation between GD, Won and Pts is a bit lower (0.1 lower). Either this is just a result of a small dataset, or it might suggest that while Decisions do matter positively, it does have lower influence in this scenario.
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them. Expand
For sure, I think it's clear that the influence of the attributes varies per attribute and the combination of attributes.
That's one of the reasons why I think it's important to test for correlation instead of pure GD, and why I think it's useful to test extremes(1 vs 20). Extremes show us that the stats do have some influence, and that it's positive. Correlation shows us how much should we expect every attribute to matter relation to every stat in specific scenarios.
harvestgreen22 said: Thanks for your work, I'm also slowly trying a new initial condition, Four teams use replicants, different 4-3-3, then test what the goal difference is without changing the attributes, then test what the goal difference is with -10 Technique 1,500 games(match) tested now. I expect to increase it to about 5000. Expand
Thanks mate, I'll be keeping an eye out in your tests! I wish I had that patience to do 5000 matches, hahaha.
Also, sorry for stealing your post. I'll probably make a separate ones for my analysis.
caffeiner said: This is true. For the mean/std metrics I could've averaged everyone that isn't the Team 1. I'll keep this in mind for other analysis. But I think it's good to have everyone separated for the other tests.
I've done similar testing for Decisions. While I didn't have time to generate all graphs for it (yet), the results seem similar regarding position (this means, it won the league just as much as Technique), however the correlation between GD, Won and Pts is a bit lower (0.1 lower). Either this is just a result of a small dataset, or it might suggest that while Decisions do matter positively, it does have lower influence in this scenario.
For sure, I think it's clear that the influence of the attributes varies per attribute and the combination of attributes.
That's one of the reasons why I think it's important to test for correlation instead of pure GD, and why I think it's useful to test extremes(1 vs 20). Extremes show us that the stats do have some influence, and that it's positive. Correlation shows us how much should we expect every attribute to matter relation to every stat in specific scenarios.
Thanks mate, I'll be keeping an eye out in your tests! I wish I had that patience to do 5000 matches, hahaha.
Also, sorry for stealing your post. I'll probably make a separate ones for my analysis. Expand
(I only partially translated it because it wasn't finished)
1. In the first picture, I used four different 4-3-3 as teams A,B,C, and D. All four teams are Manchester city replicants,
There are 960 matches in the standard group, What is shown on the top left of the table is that I put 10 players "-10 Technique", 900 matches , and it shows a slightly higher number than the standard .
In other words, with less Technique, goal difference has increased
2. In the second picture, I used 1 4-2-4 (https://fm-arena.com/thread/12717-ef-424-if-hp-v2-p101-ac/), 2 different 4-3-3, and 1 5-2-3 as teams A,B,C, and D. All four teams are Manchester city replicants, I expect to test the -10 Technique and -10 Flair
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them. Expand
The latest test results , with 20,000+ match tested
1. The results contradicted my initial judgment, In some cases, Technique and Flair will be positive. It is not known which stage setting/instruction produced this effect. in the 11 tactic have been measured, Technique and Flair are negative in more cases ,and positive in few.
2. With fewer sample, the standard deviation of Preset 532 Catenaccio is decrease accordingly to Flair increase, which may be caused by insufficient test samples
3. Based on the average standard deviation of all tactic ,estimate, it is not convergent to speculate on the results of one or two seasons It take about 20-25 seasons , 600-750 matchs each situation to have a Reasonable Confidence Intervals
4. To reduce distractions, Three of the identical tested tactic were randomly replaced with different opponent tactic, The result is the same . Therefore, judging "positive or negative" does not depend on the opponent's formation
mmigueis said: It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results. Expand
bigloser said: In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad. Expand
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner. Expand
The latest test results , with 20,000+ match tested
1. The results contradicted my initial judgment, In some cases, Technique and Flair will be positive. It is not known which stage setting/instruction produced this effect. in the 11 tactic have been measured, Technique and Flair are negative in more cases ,and positive in few.
2. With fewer sample, the standard deviation of Preset 532 Catenaccio is decrease accordingly to Flair increase, which may be caused by insufficient test samples
3. Based on the average standard deviation of all tactic ,estimate, it is not convergent to speculate on the results of one or two seasons It take about 20-25 seasons , 600-750 matchs each situation to have a Reasonable Confidence Intervals
4. To reduce distractions, Three of the identical tested tactic were randomly replaced with different opponent tactic, The result is the same . Therefore, judging "positive or negative" does not depend on the opponent's formation
"Technique (relative-technical)
How good his technique and ball control is
(how comfortable he is on the ball).
Flair (absolute-mental)
How likely he is to try the unexpected."
I use two data points ,
Goal difference
and
variance (statistical language)
1.In the last test (last post),
all 10 attributes , Goal difference =20.1 , variance =17.8
Increase Flair to 20 , Goal difference =16.9 , variance =24.5
Increase technique to 20 , Goal difference =16.5 , variance =17.4
2.
Now, there are four Manchester City replicants
Manchester City's players are already at the top of the game,
Their ability is high , so It is impossible not high enough to support "high Flair " or "high Technique".
They use the same formation, Default 433 ball control Tactics,
Team A's starting morale is a little bit higher, This is used to increase the degree of differentiation and avoid mixing the data together
Each person also has attributes corresponding to their duties according to their position,
It didn't say that the attributes didn't match his duties, his position, his role
Number of test match: 3800
Standard Group : Goal Difference =4.3, variance =12.3
Flair +10: Goal difference =3.3, variance =14.6
Technique +10: Goal difference =-0.1, variance =9.7
14.6/12.3≈120%
There's always higher variance to higher Flair , Which means Flair's true meaning may be "unexpectedness."
it causes both unexpected good and unexpected bad,
But the average goal difference has gone down
As for Technique , since its Goal difference from the standard is negative, it is a negative attribute
3.Four teams of Manchester United replicants
Number of test match: 900
Also exactly replicating in-game player attributes.
Standard Group : Goal Difference =4.1, variance =16.4
Flair +10: Goal difference =2.7, variance =19.3
19.3/16.4≈118%
They use the same formation, Default 442 defend Tactics,
The technique is also below Standard ,Under different conditions, higher Flair shows higher variance , higher Flair and higher technique show lower Goal Difference
So, I'm guessing now, Flair and Technique are negative attributes , Flair is "unexpectedness."
I actually didn't believe this result at first, it's counter-logic,share if you have any views or test results
isn't this similar to the player attribution table? https://fm-arena.com/table/26-player-attributes-testing/
Slightly fewer matches but Techniques is slightly negative while flair is slightly positive
In my FM23 experiment Flair also had a little negative correlation with player rating while Technique was the lowest coefficient non-hidden attribute (except Flair).
The findings you have made regarding flair are not that surprising to me, but the findings regarding technique are genuinely bizarre.
- Flair does not have a weight in a player's current ability (CA); therefore, it makes sense that it may have positive effects in some cases and negative effects in others. From what I've understood of game manuals over the years, flair is supposed to make players try more unexpected and harder moves, which have higher likelihood of success the more technically proficient players are. It is perhaps a bit surprising that an increase in flair was not helpful with Man City players, but overall it makes sense that more flair does not increase performance in a clear-cut fashion.
- Technique does have a weight in a player's current ability, and all the descriptions I have ever seen of the attribute through the years would indicate that higher technique should have a clear net positive on player's performance. If your test results generalize, this would seem like a clear case of SI not properly testing how the attributes work in the game's match engine. The only rationalization I can come up with for the results you are finding is that players with higher technique try harder moves, but those moves are generally worse than trying simpler, less technical moves (e.g., moves based on pace and acceleration) even when players have high technique. This would suggest that perhaps not only higher technique is not giving sufficient boost but also that the "decisions" attribute may not be working properly (as most tests have found so far, decisions provides a unrealistic small boost to players (in some tests, higher decisions even seems associated with a reduction in performance!!)).
In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad.
bigloser said: In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad.
——did you lower their technique
no.
+ 10, if + 10 and result greater than 20, then = 20 (This means that the average actual increase is less than 10.)
I am busy now , I will try other conditions when I am free.
thank you and Your information
Some community players speculate:
Whether higher Technique/Flair in specific positions (e.g., midfield, or defenders) has led to a drop in overall goal difference, or whether it has an effect in all positions,
I did some tests with flair but didn't really notice a negative correlation.
4 team league, teams B, C, D have all players with 20 in all normal attributes. Team A players have everything on 20 except flair, which is on 1. All attributes are frozen (this means that the attributes stay on the same level all season). All players are proficient in all positions. The formation used was a 4-1-1-3-1, for all teams (FB (De), CD (De), DM (Su), CM (Su), W (At), AM (Su), AF (At)).
I ran 10 seasons (the first season 10 times, to be precise). Team A won the league twice, got third a couple of times and second a couple too. The goal difference wasn't really relevant either: 19, 15, -6,-1, -16, -10, -2, 15, -6, 12.
Didn't test combinations or other attributes, though.
Of course, 10 seasons amounts to a very small number of matches in a league with 4 teams (300 matches per team), so more testing is surely required.
caffeiner said: I did some tests with flair but didn't really notice a negative correlation.
4 team league, teams B, C, D have all players with 20 in all normal attributes. Team A players have everything on 20 except flair, which is on 1. All attributes are frozen (this means that the attributes stay on the same level all season). All players are proficient in all positions. The formation used was a 4-1-1-3-1, for all teams (FB (De), CD (De), DM (Su), CM (Su), W (At), AM (Su), AF (At)).
I ran 10 seasons (the first season 10 times, to be precise). Team A won the league twice, got third a couple of times and second a couple too. The goal difference wasn't really relevant either: 19, 15, -6,-1, -16, -10, -2, 15, -6, 12.
Didn't test combinations or other attributes, though.
Of course, 10 seasons amounts to a very small number of matches in a league with 4 teams (300 matches per team), so more testing is surely required.
Thank you. If I have time I'll keep expanding the sample to see if it's random, if it's due to certain conditions, or if it's true
harvestgreen22 said: Thank you. If I have time I'll keep expanding the sample to see if it's random, if it's due to certain conditions, or if it's true
I feel there is some positive correlation using technique, but haven't been able to point exactly what is it.
I've updated my testing database. So 4 teams, all playing a 4-3-3 without any instructions (FB (Su), CD (De), DM (De), BBM (Su), CM (Su), W (Su), AF (At)). Each team plays 30 matches. Injuries are disabled. Attributes are frozen, and testing is done on holiday.
Teams 2, 3, 4 have technique on 1 for all line players, and all other attributes on 20.
Team 1 has everything at 20.
The GK for all teams have everything at 10.
I'm still running the tests but right now the difference seems too big for technique to be negative or to not matter.
This latest result makes more sense to me...
Just a small update on my findings:
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner.
It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results.
Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them.
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them.
Thanks for your work, I'm also slowly trying a new initial condition,
Four teams use replicants, different 4-3-3,
then test what the goal difference is without changing the attributes,
then test what the goal difference is with -10 Technique
1,500 games tested now. I expect to increase it to about 5000.
caffeiner said: Just a small update on my findings:
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner.
Thanks for your work, I'm also slowly trying a new initial condition,
Four teams use replicants, different 4-3-3,
then test what the goal difference is without changing the attributes,
then test what the goal difference is with -10 Technique
1,500 games(match) tested now. I expect to increase it to about 5000.
mmigueis said: It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results.
This is true. For the mean/std metrics I could've averaged everyone that isn't the Team 1. I'll keep this in mind for other analysis. But I think it's good to have everyone separated for the other tests.
I've done similar testing for Decisions. While I didn't have time to generate all graphs for it (yet), the results seem similar regarding position (this means, it won the league just as much as Technique), however the correlation between GD, Won and Pts is a bit lower (0.1 lower). Either this is just a result of a small dataset, or it might suggest that while Decisions do matter positively, it does have lower influence in this scenario.
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them.
For sure, I think it's clear that the influence of the attributes varies per attribute and the combination of attributes.
That's one of the reasons why I think it's important to test for correlation instead of pure GD, and why I think it's useful to test extremes(1 vs 20). Extremes show us that the stats do have some influence, and that it's positive. Correlation shows us how much should we expect every attribute to matter relation to every stat in specific scenarios.
harvestgreen22 said: Thanks for your work, I'm also slowly trying a new initial condition,
Four teams use replicants, different 4-3-3,
then test what the goal difference is without changing the attributes,
then test what the goal difference is with -10 Technique
1,500 games(match) tested now. I expect to increase it to about 5000.
Thanks mate, I'll be keeping an eye out in your tests! I wish I had that patience to do 5000 matches, hahaha.
Also, sorry for stealing your post. I'll probably make a separate ones for my analysis.
caffeiner said: This is true. For the mean/std metrics I could've averaged everyone that isn't the Team 1. I'll keep this in mind for other analysis. But I think it's good to have everyone separated for the other tests.
I've done similar testing for Decisions. While I didn't have time to generate all graphs for it (yet), the results seem similar regarding position (this means, it won the league just as much as Technique), however the correlation between GD, Won and Pts is a bit lower (0.1 lower). Either this is just a result of a small dataset, or it might suggest that while Decisions do matter positively, it does have lower influence in this scenario.
For sure, I think it's clear that the influence of the attributes varies per attribute and the combination of attributes.
That's one of the reasons why I think it's important to test for correlation instead of pure GD, and why I think it's useful to test extremes(1 vs 20). Extremes show us that the stats do have some influence, and that it's positive. Correlation shows us how much should we expect every attribute to matter relation to every stat in specific scenarios.
Thanks mate, I'll be keeping an eye out in your tests! I wish I had that patience to do 5000 matches, hahaha.
Also, sorry for stealing your post. I'll probably make a separate ones for my analysis.
(I only partially translated it because it wasn't finished)
1.
In the first picture, I used four different 4-3-3 as teams A,B,C, and D. All four teams are Manchester city
replicants,
There are 960 matches in the standard group,
What is shown on the top left of the table is that I put 10 players "-10 Technique", 900 matches , and it shows a slightly higher number than the standard .
In other words, with less Technique, goal difference has increased
2.
In the second picture, I used 1 4-2-4 (https://fm-arena.com/thread/12717-ef-424-if-hp-v2-p101-ac/),
2 different 4-3-3, and 1 5-2-3 as teams A,B,C, and D. All four teams are Manchester city replicants,
I expect to test the -10 Technique and -10 Flair
This second is ready to begin
svonn said: Looking at the initial tests from @harvestgreen22, it is clear that not all attributes have a linear effect. Attributes like work rate seem to have a "saturation point", so anything above 10 has diminishing returns. There may be something similar at play here - testing with 1 technique might have a high effect, 10 might be some kind of sweet spot, while the drop above might be due to something like statistical noise or some other threshold that allows actions that aren't helpful, especially if the whole team is doing them.
The latest test results , with 20,000+ match tested
1.
The results contradicted my initial judgment,
In some cases, Technique and Flair will be positive.
It is not known which stage setting/instruction produced this effect.
in the 11 tactic have been measured, Technique and Flair are negative in more cases ,and positive in few.
2.
With fewer sample, the standard deviation of Preset 532 Catenaccio is decrease accordingly to Flair increase, which may be caused by insufficient test samples
3.
Based on the average standard deviation of all tactic ,estimate, it is not convergent to speculate on the results of one or two seasons
It take about 20-25 seasons , 600-750 matchs each situation to have a Reasonable Confidence Intervals
4.
To reduce distractions,
Three of the identical tested tactic were randomly replaced with different opponent tactic, The result is the same . Therefore, judging "positive or negative" does not depend on the opponent's formation
mmigueis said: It seems pretty conclusive, IMO! (Note that in comparing the "technique = 20" team to the "technique = 1" teams, you could average across all the "technique = 1" teams, as they are set up exactly the same).
Have you ever done similar tests for other attributes? It would be very interesting to see the results.
bigloser said: In the control group did you lower their technique? By default nearly all the Man City players have over 10 technique so you wouldn't be able to raise it by 10.
In the Machine Learning weights technique was only moderately weighted high on DM(RPM), AMC, ST. One theory is it's bad to have defenders with high technique/flair. The attribute table from 2022 is basically the same as today.
Or another is technique (decisions/vision/passing?) is related to the success of certain traits like tries killer passes/curls ball/round the keeper/shoots with power/places shots/etc. And you would need said traits to get any benefit. But that doesn't explain why it's negative as you would expect it to be neutral. In the Man City example Haaland only has "tries first time shots" related to technique. This is the only thing I can think of.
The most likely scenario is technique was good at some point in previous editions but other changes made in the engine over the years has unintentionally made it bad.
tolec said: isn't this similar to the player attribution table? https://fm-arena.com/table/26-player-attributes-testing/
Slightly fewer matches but Techniques is slightly negative while flair is slightly positive
tolec said: isn't this similar to the player attribution table? https://fm-arena.com/table/26-player-attributes-testing/
Slightly fewer matches but Techniques is slightly negative while flair is slightly positive
caffeiner said: Just a small update on my findings:
As before, I did 10 seasons (this means 300 matches). The configuration is the same as the other comment.
Databattle 1 has ALL attributes on 20. The other teams have ALL attributes on 20, EXCEPT Technique which is on 1.
Hypothesis: IF Technique doesn't matter, what we expect is to that results of the league will be random or close to random. IF Technique is NEGATIVE, what we expect is that Databattle 1 will be WORSE than the other teams overall.
Results: Databattle 1 overperformed all other teams in most seasons, having better GD, Position, winrate and more points most of the time. It won the league 7 times (7/10), and has higher mean points, GD, wins, and position. See images for more info (On the correlation matrix, the "Technique 20" data point is simply 1 if the team is Databattle 1, and 0 if it isn't. Also, remember that POSITION is "inversed", so having lower position is actually better). Even in the seasons it did not win, it came second by AT MOST 4 points.
Conclusion: When everything else is at 20, is seems to be better to have technique at 20 than it is to have it at 1. Eventually I'll add more data to this analysis, as I feel 10 tournaments/300 matches is a small amount, but for now I'll test other attributes in the same manner.
The latest test results , with 20,000+ match tested
1.
The results contradicted my initial judgment,
In some cases, Technique and Flair will be positive.
It is not known which stage setting/instruction produced this effect.
in the 11 tactic have been measured, Technique and Flair are negative in more cases ,and positive in few.
2.
With fewer sample, the standard deviation of Preset 532 Catenaccio is decrease accordingly to Flair increase, which may be caused by insufficient test samples
3.
Based on the average standard deviation of all tactic ,estimate, it is not convergent to speculate on the results of one or two seasons
It take about 20-25 seasons , 600-750 matchs each situation to have a Reasonable Confidence Intervals
4.
To reduce distractions,
Three of the identical tested tactic were randomly replaced with different opponent tactic, The result is the same . Therefore, judging "positive or negative" does not depend on the opponent's formation