Should the star system be revamped for FM25?

[jcafcwbb]

As we are having a new engine should the star system be revamped as it is a bit mis-leading. 

One example is the amount of times if I play a player in the AML position as a Winger - Attack he gets 3 stars. If I play him as a winger-attack in the ML position he can drop to 1.5 stars but I am effectively asking him to do the same job but defend deeper.

A specific example is Omar Bugiel. 

As a pressing forward or target man he gets 3 stars but as a shadow striker he gets 0.5 stars. This is despite his key attributes for the shadow striker role mostly being 10-12 which is decent for a L2 player. There is no reason for that drop.

How much is the star rating allied with the actual attributes rather than whether the player is seen as a natural or unconvincing in  that position.

I know you are supposed to look at this but for newbies isn't this mis-leading?

[kevhamster]

What needs to be realised is the star rating is based on your assistant's opinion of that player's suitability to that role.  I don't think it needs to be revamped as such, I just think it needs to be clearer that this is the case and to remember to use it as a guide first and foremost, but to trust what you're seeing with your own eyes in the match engine.

[jcafcwbb]

A second thought to this - in the era of big data should the star rating come from the Data Analysis team rather than the assistant manager?

[Brasilia88]

They'd have to run more and positioning would count as well due to dropping down the field a bit.

Meaning they'd have to start further away from goal and are, thus, required to be physically involved with an implications. This consequence the game doesn't count for, since every number for 'Positioning' (to number a major one) is a stat allocated to how well they are in that position, but not compared to themselves in other roles or positions. Surely, the system counts competence in said role, but that's trainable and, as far as I know, only affects their PA. Therefore, one could argue that older players could be totally incompetent in a slightly different role or position, but they still 'function' in a way.  

Besides, the star system compares itself with your own stock of players, so that is your to-go point. 

[Dotsworthy]

1 hour ago, jcafcwbb said:

As a pressing forward or target man he gets 3 stars but as a shadow striker he gets 0.5 stars. This is despite his key attributes for the shadow striker role mostly being 10-12 which is decent for a L2 player. There is no reason for that drop.

How much is the star rating allied with the actual attributes rather than whether the player is seen as a natural or unconvincing in  that position.

A player who is not natural in a given position will perform worse though, so it should be taken into account. While the player might have the perfect attributes they take a pretty significant hit to their decision making if they cannot play in that position.

The star ratings for positions seem much better this iteration. Prior to FM24 every full back in the game could only get to 2 star rating for IWB role.

I think the star rating tries to replicate the fluctuating nature of footballing talent which I quite like. In real life players can play like elite level premier league talent for two years but then regress back to their actual level, or be star players at the start of your save only to become overshadowed by better talent as your go on.

It could do with a bit of clarity for the new player.

[whatsupdoc]

2 hours ago, jcafcwbb said:

As we are having a new engine should the star system be revamped as it is a bit mis-leading. 

One example is the amount of times if I play a player in the AML position as a Winger - Attack he gets 3 stars. If I play him as a winger-attack in the ML position he can drop to 1.5 stars but I am effectively asking him to do the same job but defend deeper.

A specific example is Omar Bugiel. 

As a pressing forward or target man he gets 3 stars but as a shadow striker he gets 0.5 stars. This is despite his key attributes for the shadow striker role mostly being 10-12 which is decent for a L2 player. There is no reason for that drop.

How much is the star rating allied with the actual attributes rather than whether the player is seen as a natural or unconvincing in  that position.

I know you are supposed to look at this but for newbies isn't this mis-leading?

 

With the new unity engine we're talking about graphical improvements right? There's no indication that the match engine is going to receive any significant rewrite as far I can tell. 

[NineCloudNine]

9 hours ago, kevhamster said:

What needs to be realised is the star rating is based on your assistant's opinion of that player's suitability to that role.  I don't think it needs to be revamped as such, I just think it needs to be clearer that this is the case and to remember to use it as a guide first and foremost, but to trust what you're seeing with your own eyes in the match engine.

The role suitability stars on a player’s profile are a function of Current Ability, positional familiarity, the league you’re in and (as the game progresses) player reputation.

In the pre-game editor each player has a ‘role scores’ tab. This lists their role suitabilities in order. The top one always has a ‘score’ equal to Current Ability, with the rest having scores which decrement from that depending on the player’s attributes and positional familiarity. Only roles in positions for which a player has maximum positional familiarity will get the full CA-linked score.

In game this is turned into stars depending on the league you are in. For top leagues (Premier, La Liga, Serie A for sure) 100 CA = 1 star and it goes up half a star for every 10pts of CA to 5 stars at 180 CA.

You can experiment with this using the IGE if you have it. Change a player’s CA and their role stars will increase (though the order of them will not). Change their positional familiarity and their role stars for their ‘new’ position will increase.

The game is effectively saying that a player with 5 stars in a role (180 CA) is equivalent to a 160 CA player in a role for which they have 4 stars.

As your game progresses, a very high performing player may get a star rating in excess of the formula used at the start. So player reputation/performance does have a role, but not in the initial setup of the game.

Where I strongly agree with the OP is the silliness of saying that a 4-star AML winger will suddenly become and awkward 2-star if asked to be a winger in ML. I’m less convinced by the example of a Target Man being able to seamlessly transition to being a Shadow Striker. It’s hard to think of two more different attacking roles.

Edited January 19 by NineCloudNine

[jcafcwbb]

Omar Bugiel is a weird one in this regard.

In real life, before this season, he has been exclusively as forward and played in a straight 4-4-2 for Sutton. He joined us, AFC Wimbledon, in the summer and is being used in a different role/position this season. We have mostly played the 4-2-3-1 system with Omar dropping into the AM/SS role rather than the ST one.  The other formation we have tried is 4-3-1-2 where he has played as the one behind the front two. He has been one of our best players in this new role/position. His Fotmob rating, so far this season, is 7.04 and his Sofa Score one is 7.20. I have wanted to replicate this in the game but have scared off by his low rating. I may try just playing him there to see how the game reacts to it.

In the game, his only natural position is ST and nothing else. This must cause the game to downgrade his star rating. I understand the current game cannot accommodate for the real life player evolving his role or effectively changing his position but I do wonder if the updated data, in February, can reflect a player moving to a new role/position and if it can be added to a current save.

There might be a case for changing the way it is calculated - an example being -  possibly using the top 8 attributes to a role, adding the adaptability and the CA. Maybe removing the positional familiarity from the calculation can smooth out some of the quirks. I am not a computer programmer so would be interested to see if this was possible without breaking the game or if people think it was a good thing. 

[kevhamster]

2 hours ago, NineCloudNine said:

Where I strongly agree with the OP is the silliness of saying that a 4-star AML winger will suddenly become and awkward 2-star if asked to be a winger in ML. I’m less convinced by the example of a Target Man being able to seamlessly transition to being a Shadow Striker. It’s hard to think of two more different attacking roles.

And here's the key to use it as a guide - it's designed to help people set their team up to ensure suitable roles are being picked, but it is still just a guide.

[NineCloudNine]

1 hour ago, kevhamster said:

And here's the key to use it as a guide - it's designed to help people set their team up to ensure suitable roles are being picked, but it is still just a guide.

Yes…and no.

Your previous post described the role scores as an Assistant’s opinion. That’s not so. You can test this yourself by hiring a different Assistant - the role scores won’t change.

They won’t change if you bring in a new star player or three either, because how many stars a player with X CA gets depends on your league. You can test this too.

You are correct that players can be asked to perform - and may excel at - multiple positions and roles depending on your tactic and team needs. This happens all the time in real life.

However, the statistical basis for the stars can’t be ignored. The formula effectively gives a CA penalty to a player who is not playing their strongest role. In the top leagues this penalty is 10pts per half star. This may not matter at all, but multiplied across a team with a lot of players not in their strongest role it might heavily hamper your team. It depends on how much of a difference there is between the roles, a player’s own PPM’s, your TIs etc etc.

So yes, it is just a guide. But no, it’s also more than that because FM is a numbers based game.

I’ve done lots of playing around with this. A team where everyone is playing in their best role will be happier, more coherent and perform better than a team with the same tactic, the same collective/individual CAs but all playing in non-preferred roles.

Those are extremes of course - in reality every team will have a mix of players playing in preferred roles and players playing in non-preferred roles. How much effect this has depends on a billion factors, but it can’t be ignored.

Edited January 19 by NineCloudNine

[dunk105]

13 hours ago, kevhamster said:

What needs to be realised is the star rating is based on your assistant's opinion of that player's suitability to that role.  I don't think it needs to be revamped as such, I just think it needs to be clearer that this is the case and to remember to use it as a guide first and foremost, but to trust what you're seeing with your own eyes in the match engine.

Agree, and its also a comparison to your current squad.

Therefore a 5 star level player when you're squad is average will be a 3-4 star rating as your squad /  players in that position improve.

Its not something anyone should depend on but once you understand how it truly works then it is a useful guide.

[NineCloudNine]

23 minutes ago, dunk105 said:

Agree, and its also a comparison to your current squad.

Therefore a 5 star level player when you're squad is average will be a 3-4 star rating as your squad /  players in that position improve.

This isn’t accurate. Rather than retype what I wrote above, take a look.

The role star ratings depend on CA, positional familiarity and the division you are in. As the game progresses player reputation can also boost it.

I have experimented with adding new star players, removing star players, changing Assistants, changing Assistant attributes, changing player positions, CA, PA, attributes. Only changes to CA and position familiarity change the highest star a player gets. Changing player attributes alters the order of their preferred roles and how good they are at a role, but only up to the cap set by their CA and league.

If you have the IGE you can test this yourself. Give several players 200 CA. They will max at 5 stars for their preferred role (and perhaps others). The role stars of your other players will not change.

There may be some confusion here with the ‘assistants opinion of this player’s ability’, which does change as you describe. But role stars do not.

[Freakiie]

7 hours ago, jcafcwbb said:

Omar Bugiel is a weird one in this regard.

In real life, before this season, he has been exclusively as forward and played in a straight 4-4-2 for Sutton. He joined us, AFC Wimbledon, in the summer and is being used in a different role/position this season. We have mostly played the 4-2-3-1 system with Omar dropping into the AM/SS role rather than the ST one.  The other formation we have tried is 4-3-1-2 where he has played as the one behind the front two. He has been one of our best players in this new role/position. His Fotmob rating, so far this season, is 7.04 and his Sofa Score one is 7.20. I have wanted to replicate this in the game but have scared off by his low rating. I may try just playing him there to see how the game reacts to it.

In the game, his only natural position is ST and nothing else. This must cause the game to downgrade his star rating. I understand the current game cannot accommodate for the real life player evolving his role or effectively changing his position but I do wonder if the updated data, in February, can reflect a player moving to a new role/position and if it can be added to a current save.

There might be a case for changing the way it is calculated - an example being -  possibly using the top 8 attributes to a role, adding the adaptability and the CA. Maybe removing the positional familiarity from the calculation can smooth out some of the quirks. I am not a computer programmer so would be interested to see if this was possible without breaking the game or if people think it was a good thing. 

Nothing is stopping you from retraining him and playing him in a new position yourself though? Star ratings are just a suggestion and a player playing in a new position won't suddenly forget how to kick a ball.

[phd_angel]

19 hours ago, kevhamster said:

What needs to be realised is the star rating is based on your assistant's opinion of that player's suitability to that role.  I don't think it needs to be revamped as such, I just think it needs to be clearer that this is the case and to remember to use it as a guide first and foremost, but to trust what you're seeing with your own eyes in the match engine.

Ha! I was looking for this explanation. Yes, I'm aware that this what SI says, but the question could be: can this be improved?... Maybe, instead of just the sole opinion of the assistant manager, the average opinion of the "market" should also count.  Or, the ability of the assistant manager to judge players could be made more explicit or expanded with a more specific attribute, etc.

[Hovis Dexter]

The star rating is the opinion of the member of staff whom you ask to assess your players. Change the member of staff and the star ratings will almost certainly change. I think that this is realistic in that as the manager you ask your staff for their opinions as to the relative ratings of your players and depending on who you ask you will get a different opinion. You also ask your scouts to rate the players that they have scouted compared to your current team. Again, each scout may well give a different opinion.

The accuracy of the various star ratings does, of course, vary depending upon the judging ability of the person giving the opinion. A member of staff with judging ability of 20 will give a highly accurate star rating whereas one who has a judging ability of 1 is likely not to have a clue.

This all seems to be perfectly sensible and I don’t really see any reason to change the system. If people want more accurate rating they can always look at the CA  and PA ratings or use an external tool if that is their preference.

[NineCloudNine]

53 minutes ago, Hovis Dexter said:

The star rating is the opinion of the member of staff whom you ask to assess your players. Change the member of staff and the star ratings will almost certainly change. I think that this is realistic in that as the manager you ask your staff for their opinions as to the relative ratings of your players and depending on who you ask you will get a different opinion. You also ask your scouts to rate the players that they have scouted compared to your current team. Again, each scout may well give a different opinion.

The accuracy of the various star ratings does, of course, vary depending upon the judging ability of the person giving the opinion. A member of staff with judging ability of 20 will give a highly accurate star rating whereas one who has a judging ability of 1 is likely not to have a clue.

This all seems to be perfectly sensible and I don’t really see any reason to change the system. If people want more accurate rating they can always look at the CA  and PA ratings or use an external tool if that is their preference.

Again ...does anyone read the thread? ... ROLE stars are not the opinion of your staff and they do not change when staff do.

The stars under 'coach assessment' and in scouting do vary with staff skill. But the OP isn't about those

[NineCloudNine]

3 hours ago, phd_angel said:

Ha! I was looking for this explanation. Yes, I'm aware that this what SI says, but the question could be: can this be improved?... Maybe, instead of just the sole opinion of the assistant manager, the average opinion of the "market" should also count.  Or, the ability of the assistant manager to judge players could be made more explicit or expanded with a more specific attribute, etc.

The explanation you are pleased with is wrong, as regards role scores . The answers you are looking for are in my several long posts in thsi thread that no-one has read .

Edited January 19 by NineCloudNine

[Dadecane]

17 hours ago, NineCloudNine said:

Again ...does anyone read the thread? ... ROLE stars are not the opinion of your staff and they do not change when staff do.

The stars under 'coach assessment' and in scouting do vary with staff skill. But the OP isn't about those

In the squad page (maybe in squad planner i forget) where u can change “the opinion of” as u view ur players, changing assistants is 100% showing a change in stars for me man. Just not much, maybe a half star or 1 star variation on 2-3 pñayers out of my whole squad

[Hovis Dexter]

Go to Tactics and under the view insert the column entitle Role Ability. Hover over column title and it literally says "Assistant Manager's Opinion of this player's Current Ability in the Selected Role" ... change the person who gives you player reports or change your Assistant Manager and the stars may well change.

[NineCloudNine]

1 hour ago, Dadecane said:

In the squad page (maybe in squad planner i forget) where u can change “the opinion of” as u view ur players, changing assistants is 100% showing a change in stars for me man. Just not much, maybe a half star or 1 star variation on 2-3 pñayers out of my whole squad

Go to a player's page. Hover over the stars under 'role and duty'. No mention of anyone's opinion. Those stars are a result of CA, positional familiarity and the league you are in. You can see the way it works if you look at 'role scores' in the pre-game editor. As the save progresses, I have seen players' roles scores go up by half a point beyond the formula when they play very well. But they never vary from the forumla at the start.

Other stars in different places (eg 'coach summary' top right of player screen) clearly state that it's a staff member's opinion. Same with scouting. I believe that those scores are calculated the same way but then given a +/1 margin to make it look like it's uncertain. Like many areas of FM's interface, this is cluttered, unintuitive and often unclear,

Point is though, role stars are not "just an assistant's opinion". They have a statistical basis and need to be included in decisions about tactics, players and roles.

Rather than repeat myself again I'm going to leave this now. Good luck !

Edited January 20 by NineCloudNine

[Dadecane]

40 minutes ago, NineCloudNine said:

Go to a player's page. Hover over the stars under 'role and duty'. No mention of anyone's opinion. Those stars are a result of CA, positional familiarity and the league you are in. You can see the way it works if you look at 'role scores' in the pre-game editor. As the save progresses, I have seen players' roles scores go up by half a point beyond the formula when they play very well. But they never vary from the forumla at the start.

Other stars in different places (eg 'coach summary' top right of player screen) clearly state that it's a staff member's opinion. Same with scouting. I believe that those scores are calculated the same way but then given a +/1 margin to make it look like it's uncertain. Like many areas of FM's interface, this is cluttered, unintuitive and often unclear,

Point is though, role stars are not "just an assistant's opinion". They have a statistical basis and need to be included in decisions about tactics, players and roles.

Rather than repeat myself again I'm going to leave this now. Good luck !

Gotcha. Makes sense. I did read all your posts before and you seem to know your stuff. I rely heavily on the star system tbh. Its just I remember there was a page where changing the “opinion of..” scroll did incur changes to some star ratings, but it makes sense now.