Hello , I have made a filter for positional ratings based on the tests done on attributes (so far). It doesn't include GK rating for obvious reasons. I thought I would post it here and maybe people can give feedback on potential changes that could be made. I have made an estimation on the stats not yet tested based on importance in previous years. The weighings are quite "harsh" but I also didnt want to give points to attributes that have no proven positive effect!
DreadPirateRoberts said: Hello , I have made a filter for positional ratings based on the tests done on attributes (so far). It doesn't include GK rating for obvious reasons. I thought I would post it here and maybe people can give feedback on potential changes that could be made. I have made an estimation on the stats not yet tested based on importance in previous years. The weighings are quite "harsh" but I also didnt want to give points to attributes that have no proven positive effect!
Forgot to check this today. Have been using my own ratings for squirrel python code and have found it to be excellent. I think it's far superior to using GS with yk but will check it against yours tomorrow
Since the meta attributes is pretty much the same as in last FM, i think ykykyk works just fine in this new FM, no need to use any other GS ratings so far.
Gracolas said: Since the meta attributes is pretty much the same as in last FM, i think ykykyk works just fine in this new FM, no need to use any other GS ratings so far. Expand
ykykyk is based on the machine learning findings more than the meta attributes.You will find that it gives high weightage to attributes such as finishing, position, technique, etc even though the FM attribute testing show them to be non-meta. The reasoning behind including non-meta attributes was that it helps find more well rounded players.
keithb said: My first observation is that Haaland isn't rated as the best striker?? He's one of the best FM players ever, let alone in this version.
Why have you rated jumping reach so poorly?????? It needs to be high for almost every outfield position and max for centre backs. Expand
I am assuming you checked the Fast Man rating? Check out the Target Man rating. It gives quite a high weightage to jumping reach. I use the TM rating because I noticed players with high TM rating generally do better, even when used by the AI. Generally, anything over 79 TM rating seem to do pretty well.
Floppyaams said: I am assuming you checked the Fast Man rating? Check out the Target Man rating. It gives quite a high weightage to jumping reach. I use the TM rating because I noticed players with high TM rating generally do better, even when used by the AI. Generally, anything over 79 TM rating seem to do pretty well. Expand
Jumping reach is set to 0 for many positions and only 65 for TM? This is all too low and a big mistake.
Haaland is the best FS and the best TM. I agree TM is better in FM because of how powerful jumping reach is. Actually his dribbling is probably a little low to be the best FS? Im not sure.
keithb said: Jumping reach is set to 0 for many positions and only 65 for TM? This is all too low and a big mistake.
Haaland is the best FS and the best TM. I agree TM is better in FM because of how powerful jumping reach is. Actually his dribbling is probably a little low to be the best FS? Im not sure. Expand
I agree, Jumping reach should have higher weightage for all central positions (CM, CDM, CB, and even AM). I also think crossing should be another key attribute for WBs because I notice players with high crossing (key example Borja Sosa. Cheap WB with 16-17 crossing) stats generally do very well, especially in terms of assists.
I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me
NandaldiaN said: I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me Expand
Hojlund has 17 jumping reach? What is the difference with balance and agility also?
NandaldiaN said: I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me Expand
All of them are pretty good, none are perfect. Plus there is more to performance than just these ratings. Hidden attributes, suitability to your tactic, form and so on all play a massive role.
Yarema said: All of them are pretty good, none are perfect. Plus there is more to performance than just these ratings. Hidden attributes, suitability to your tactic, form and so on all play a massive role. Expand
I dont see much diference between them in hidden atributes, the only one with more than +2 points in the diference is adaptability, which is far superior in Hojlund with 17 against 12 from konate
keithb said: Forgot to check this today. Have been using my own ratings for squirrel python code and have found it to be excellent. I think it's far superior to using GS with yk but will check it against yours tomorrow Expand @keithb would you share your modified squirrel ratings please ?
doru228 said: @keithb would you share your modified squirrel ratings please ? Expand
Hi I stopped playing the game but can roughly remember.
Speed for every position x5 FB and DM stamina x5 Dribbling x5 for wingers and am and maybe af Jumping reach x5 for cb and cf Anticipation and balance x3 for every position Concentrtion x3 I only used for cb and dm Agility for all i think x3 Then fb could have x3 dribbling and x1 elsewhere. Things like strength, first touch, finishing, flair, composure, work rate, flair etc you can put in the ones. Tbh speed needs to be x7 and this could probably be done in the code easily.
Then for the x1 up to you. Its important that all positions have around the same numbers allocated. ie when you do the calculation at the end - 1/37 - they all need to be roughly around the same imo.
Then test what you have done by looking at 20 best players for each position at the start of a new game to make sure the correct players are top. You dont want all your cb's being rated higher at dm or vice versa, it will likely mean you have the calculation wrong. But a cb can be a better dm and vice versa. Just not all of them.
Haaland(cf+af) and Mbappe(if+af) should be clear of everone, can be as high as 18. They will never be replicated in the game by newgens. If you look at Haaland his dribbling and work rate are not that high, 13 ish I think, so I wouldn't worry about those attributes for his role.
The scores 14+ is good enough for top flight 15+ is champions league 16+ is world class 17+(does happen) will be the best players in your save.
Hello , I have made a filter for positional ratings based on the tests done on attributes (so far). It doesn't include GK rating for obvious reasons. I thought I would post it here and maybe people can give feedback on potential changes that could be made. I have made an estimation on the stats not yet tested based on importance in previous years. The weighings are quite "harsh" but I also didnt want to give points to attributes that have no proven positive effect!
Link to ratings : https://mega.nz/file/KLhl1DiZ#goT2Kvmd_H6UC7YIxV4akOyKZfWe8Cm46k7E8-1GdEE
Nice!! probably te most accurate rating we got so far?
DreadPirateRoberts said: Hello , I have made a filter for positional ratings based on the tests done on attributes (so far). It doesn't include GK rating for obvious reasons. I thought I would post it here and maybe people can give feedback on potential changes that could be made. I have made an estimation on the stats not yet tested based on importance in previous years. The weighings are quite "harsh" but I also didnt want to give points to attributes that have no proven positive effect!
Link to ratings : https://mega.nz/file/KLhl1DiZ#goT2Kvmd_H6UC7YIxV4akOyKZfWe8Cm46k7E8-1GdEE
Thank you for your work! It will be interesting to compare the results of this filter to those of ykykyk balanced, ZaZ 50 and MD.
Forgot to check this today. Have been using my own ratings for squirrel python code and have found it to be excellent. I think it's far superior to using GS with yk but will check it against yours tomorrow
Since the meta attributes is pretty much the same as in last FM, i think ykykyk works just fine in this new FM, no need to use any other GS ratings so far.
ykyk doesn't really use the meta attributes all that much, to be honest, but it's better than the default CA and GS ratings
Gracolas said: Since the meta attributes is pretty much the same as in last FM, i think ykykyk works just fine in this new FM, no need to use any other GS ratings so far.
ykykyk is based on the machine learning findings more than the meta attributes.You will find that it gives high weightage to attributes such as finishing, position, technique, etc even though the FM attribute testing show them to be non-meta. The reasoning behind including non-meta attributes was that it helps find more well rounded players.
My first observation is that Haaland isn't rated as the best striker?? He's one of the best FM players ever, let alone in this version.
Why have you rated jumping reach so poorly?????? It needs to be high for almost every outfield position and max for centre backs.
keithb said: My first observation is that Haaland isn't rated as the best striker?? He's one of the best FM players ever, let alone in this version.
Why have you rated jumping reach so poorly?????? It needs to be high for almost every outfield position and max for centre backs.
I am assuming you checked the Fast Man rating? Check out the Target Man rating. It gives quite a high weightage to jumping reach. I use the TM rating because I noticed players with high TM rating generally do better, even when used by the AI. Generally, anything over 79 TM rating seem to do pretty well.
Floppyaams said: I am assuming you checked the Fast Man rating? Check out the Target Man rating. It gives quite a high weightage to jumping reach. I use the TM rating because I noticed players with high TM rating generally do better, even when used by the AI. Generally, anything over 79 TM rating seem to do pretty well.
Jumping reach is set to 0 for many positions and only 65 for TM? This is all too low and a big mistake.
Haaland is the best FS and the best TM. I agree TM is better in FM because of how powerful jumping reach is. Actually his dribbling is probably a little low to be the best FS? Im not sure.
keithb said: Jumping reach is set to 0 for many positions and only 65 for TM? This is all too low and a big mistake.
Haaland is the best FS and the best TM. I agree TM is better in FM because of how powerful jumping reach is. Actually his dribbling is probably a little low to be the best FS? Im not sure.
I agree, Jumping reach should have higher weightage for all central positions (CM, CDM, CB, and even AM). I also think crossing should be another key attribute for WBs because I notice players with high crossing (key example Borja Sosa. Cheap WB with 16-17 crossing) stats generally do very well, especially in terms of assists.
I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me
NandaldiaN said: I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me
Hojlund has 17 jumping reach? What is the difference with balance and agility also?
NandaldiaN said: I just dont know what ratings use right now... I bought konaté wich is by far better in ratings compared to hojlund but the last one is always performing better, scoring more, assisting more... and its not the first time it is happening to me with some players.. its weird to me
All of them are pretty good, none are perfect. Plus there is more to performance than just these ratings. Hidden attributes, suitability to your tactic, form and so on all play a massive role.
keithb said: Hojlund has 17 jumping reach? What is the difference with balance and agility also?
Konate | jumping 6 - 16 | Hojlund
Konate | balance 16 - 16 | Hojlund
Konate | agility 18 - 13 | Hojlund
Konate | dribbling 15 - 13 | Hojlund
Konate | pace 17 - 17 | Hojlund
Konate | accel 19 - 14 | Hojlund
Yarema said: All of them are pretty good, none are perfect. Plus there is more to performance than just these ratings. Hidden attributes, suitability to your tactic, form and so on all play a massive role.
I dont see much diference between them in hidden atributes, the only one with more than +2 points in the diference is adaptability, which is far superior in Hojlund with 17 against 12 from konate
keithb said: Forgot to check this today. Have been using my own ratings for squirrel python code and have found it to be excellent. I think it's far superior to using GS with yk but will check it against yours tomorrow
@keithb
would you share your modified squirrel ratings please ?
doru228 said: @keithb
would you share your modified squirrel ratings please ?
Hi I stopped playing the game but can roughly remember.
Speed for every position x5
FB and DM stamina x5
Dribbling x5 for wingers and am and maybe af
Jumping reach x5 for cb and cf
Anticipation and balance x3 for every position
Concentrtion x3 I only used for cb and dm
Agility for all i think x3
Then fb could have x3 dribbling and x1 elsewhere.
Things like strength, first touch, finishing, flair, composure, work rate, flair etc you can put in the ones. Tbh speed needs to be x7 and this could probably be done in the code easily.
Then for the x1 up to you. Its important that all positions have around the same numbers allocated. ie when you do the calculation at the end - 1/37 - they all need to be roughly around the same imo.
Then test what you have done by looking at 20 best players for each position at the start of a new game to make sure the correct players are top. You dont want all your cb's being rated higher at dm or vice versa, it will likely mean you have the calculation wrong. But a cb can be a better dm and vice versa. Just not all of them.
Haaland(cf+af) and Mbappe(if+af) should be clear of everone, can be as high as 18. They will never be replicated in the game by newgens. If you look at Haaland his dribbling and work rate are not that high, 13 ish I think, so I wouldn't worry about those attributes for his role.
The scores
14+ is good enough for top flight
15+ is champions league
16+ is world class
17+(does happen) will be the best players in your save.