Positional Familiarity is Wack

[Weston]

I have played FM for years, and trained players in new positions for seasons, and never once have I noticed the shade of green actually change in the circle no matter how hard I've tried. I've always been bothered by how hard it is to train players in new positions their attributes are perfect for.

Take the player below. He is most comfortable as a center back, but when he arrived he was ok at DM and even less green at CM. Despite this, I wanted him to be my ball winning midfielder, and planned to use him at either DM or CM. Because CM had the less convincing shade of green, I trained him there, as DM was already okay-ish. Finally, for the first time ever, I noticed the shade of green get better at CM! But... the DM shade of green got WORSE! How do you finally learn you can go from CB to CM and forget how to play DM in the process? It's even literally in between the two positions. So I've actually sabotaged myself, as I play him at DM the most.

Basically the two positions flipped familiarity. I get this must be a factor of him reaching his PA, so they had to take away something else to facilitate this, which is kind of ridiculous, but why not take away his ability to play LB, which is useless? And he's relatively old, so why is it finally happening for him but none of my younger players?

Ironically, if a player IRL moves position for one season (take Asamoah, was a CM, played LM/LB for a year) FM changes him to that position almost completely, yet there's been no increase to his familiarity to DM after being used there since I signed him.

[Tiger666]

What does "wack" mean?

[phnompenhandy]

^ "Yoof"-speak, I presume.

 

Not my experience. I always use a Segundo Volante, but I never get any DM ready for that role. I train and play them in the role relentlessly and over time the pis slices increase and colour changes. It takes a couple of years or more, so you  need patience.

 

Admittedly it works rapidly when the player moves clubs - I had a striker I was training for a certain role who went out on load; when he returned he was full-on green circle for that position.

 

At the end of the day it's mostly cosmetic - my players seem to perform just fine in the roles I use them for regardless of PF.

[Weston]

2 minutes ago, phnompenhandy said:

^ "Yoof"-speak, I presume.

 

Not my experience. I always use a Segundo Volante, but I never get any DM ready for that role. I train and play them in the role relentlessly and over time the pis slices increase and colour changes. It takes a couple of years or more, so you  need patience.

 

Admittedly it works rapidly when the player moves clubs - I had a striker I was training for a certain role who went out on load; when he returned he was full-on green circle for that position.

 

At the end of the day it's mostly cosmetic - my players seem to perform just fine in the roles I use them for regardless of PF.

I was religiously anti-anything-but-a-bright-green-circle at first and would only play players in their ideal positions, until someone told me that the only difference for out of position players is a hit to their decisions attribute (is that still true?), so then I experimented more. Sometimes it's okay, but sometimes it's not. I played RW Suso as a AMC and he was trash, despite having great attributes for the role, so I can't help but feel it was at least partially because of familiarity.

[HUNT3R]

3 hours ago, Weston said:

only difference for out of position players is a hit to their decisions attribute (is that still true?),

Yes, so it affects every decision the player makes. It's less of a hit the more familiar he is with the position.

[Weston]

9 hours ago, HUNT3R said:

Yes, so it affects every decision the player makes. It's less of a hit the more familiar he is with the position.

Which is why it's frustrating that my training my CB to better play CM he somehow forgot how to play DM in between despite playing there almost every match. Can you shed any light on the other issues in my initial post?

[Geordieboy52]

Positional familiarity doesn't dictate effectiveness, attributes do - these are just guidelines.

[phnompenhandy]

12 hours ago, Geordieboy52 said:

Positional familiarity doesn't dictate effectiveness, attributes do - these are just guidelines.

 

Apart from Decision, as Hunt3r says. I'd be interested in some insights from SI as to exactly how much Decisions is hit, because in my experience it's not that much. Having said that, I'm managing in non-league where 7 for Decisions is regarded as pretty good, so it may not be the same as in the Premier League.

[Viking]

Isn't versatility a factor here, too?

[pdstanbridge]

It certainly used to be versatality related. If you went into the editor and changed versatility from say 5 to 20 you would see the instant bright green appear. But I'm not sure it works like that now.

 

[Geordieboy52]

5 hours ago, phnompenhandy said:

 

Apart from Decision, as Hunt3r says. I'd be interested in some insights from SI as to exactly how much Decisions is hit, because in my experience it's not that much. Having said that, I'm managing in non-league where 7 for Decisions is regarded as pretty good, so it may not be the same as in the Premier League.

I'm fairly certain that's who I had the conversations with.  I remember retraining players in positions simply because their attributes suggested they should play there.  Didn't @Cleon famously retrain a left back to be a forward based on this principle?

[Tiger666]

29 minutes ago, Geordieboy52 said:

I'm fairly certain that's who I had the conversations with.  I remember retraining players in positions simply because their attributes suggested they should play there.  Didn't @Cleon famously retrain a left back to be a forward based on this principle?

I always retrain Alfred Duncan as a left back, even though he starts as a defensive midfielder. Performs superbly. Attributes is all that matters.

 

Edited January 23, 2018 by Tiger666

[The Enforcer]

@Tiger666

How long did that take to go from zero to natural - I notice the player is the same age?

[Weston]

5 hours ago, Tiger666 said:

I always retrain Alfred Duncan as a left back, even though he starts as a defensive midfielder. Performs superbly. Attributes is all that matters.

 

That's a brilliant move there, very cleverly done.

But how though? I literally train players with dark green circles for entire seasons and never see it flip to bright green. This was the first time I can recall a circle actually changing for me, and it was a 30 year old player switching to a position he plays less at the expense of the one he normally plays 

Edited January 23, 2018 by Weston

[HUNT3R]

Nothing wrong with them being Accomplished in a position.

[Tiger666]

1 hour ago, The Enforcer said:

@Tiger666

How long did that take to go from zero to natural - I notice the player is the same age?

Oh it usually takes a season or so, I just edited that for screenshot purposes.

[Weston]

1 minute ago, HUNT3R said:

Nothing wrong with them being Accomplished in a position.

But according to my best judge of current ability the player drops half to one whole star in a dark green position...

[Tiger666]

5 minutes ago, Weston said:

That's a brilliant move there, very cleverly done.

But how though? I literally train players with dark green circles for entire seasons and never see it flip to bright green. This was the first time I can recall a circle actually changing for me, and it was a 30 year old player switching to a position he plays less at the expense of the one he normally plays 

Depends on the player's versatility. Dark green is fine, you're not going to notice any difference between accomplished and natural.

[Weston]

I'm not loaning this lad out for the upcoming season just so I can keep him at my excellent training facilities and drill into him that he's becoming a striker now instead of a winger. We'll see if it ever actually happens.

[HUNT3R]

Just now, Weston said:

But according to my best judge of current ability the player drops half to one whole star in a dark green position...

It's just a star rating though.

[Weston]

2 minutes ago, HUNT3R said:

It's just a star rating though.

I feel like whenever I raise a problem about positions or stars or anything it always ends up boiling down to "it doesn't matter anyway" in the long run haha

So if training a position eats up so much of the max ability level set for the player but merely affects decisions, it's actually bad to train young players in new positions when they could take the boost to attributes instead and then just get put wherever you want them to go anyway?

I just wish I knew why Suso (RW) was such trash when I played him at CAM for a season.

[enigmatic]

To be fair, I think real life players do lose instincts of a position when deployed exclusively elsewhere. Though this is obviously more to do with CA-budgeting (I find the across-the-board drop in significant attributes when a player at his potential improves in a new position far more annoying though. Great news, your DM is now accomplished at playing centre back. He's just less good at jumping, heading, marking and positioning than before he started intensively training for a role which prioritises those characteristics)

-

Is "position only affects players via the Decisions attribute" the confirmed SI position of how the ME works, or just commonly accepted-but-not-actually-literally-true received wisdom based on statements to the effect that out of position players are less good at decision making?

Always assumed it was more like PPMs in terms of actual ME behaviour effects so a natural striker would try to make lots of curving runs off the last defender, hit channels and gamble at the near post (but might be terrible at timing them with poor Decisions) whereas, say Kroos with his excellent decision making played in an unfamiliar striker role wouldn't decide to beat the offside trap, run channels or gamble on the near post nearly as much as even a low-Decisions striker (but would be pretty damn smart at using the ball even compared with most good strikers)

This would make more sense than having a player with 18 decisions (and appropriate other attributes) looking more natural at the specialist requirements of a position they'd never even trained in before than most actual specialists.  Seems more consistent with my observations of players played out of position actually act too.

[HUNT3R]

Just now, enigmatic said:

Is "position only affects players via the Decisions attribute" the confirmed SI position of how the ME works, or just commonly accepted-but-not-actually-literally-true received wisdom based on statements to the effect that out of position players are less good at decision making?

It's true, although as said before - it affects Decisions, so every single decision made. Long-term, a player should really be re-trained to the position.

[enigmatic]

Have always suspected this is one of the misunderstood things like "Determination doesn't affect development"....

Bit disappointing if it isn't, tbh, since manuals have long hinted at subtleties around player familiarity with linked positions etc. An defensively-sound MR thrust into an unfamiliar DR position really ought to have different limitations to how they play than a technically-sound DC used there instead (much more willing to overlap, much less adept at the offside trap, holding all other attributes equal). Kroos' lack of training to make the runs of an average striker should be as abundantly obvious as his vastly superior linkup play, with the tradeoff representing a real tactical choice

Not to mention the issue that even huge differences in Decisions appears to have surprisingly moderate effects on in-game performance, before we've started scaling them based on a position familiarity scale and a hidden versatility stat.

Might revisit my "play every player in a position they have no familiarity in" experiment and see if I can get better results...

[Weston]

1 hour ago, enigmatic said:

Have always suspected this is one of the misunderstood things like "Determination doesn't affect development"....

Bit disappointing if it isn't, tbh, since manuals have long hinted at subtleties around player familiarity with linked positions etc. An defensively-sound MR thrust into an unfamiliar DR position really ought to have different limitations to how they play than a technically-sound DC used there instead (much more willing to overlap, much less adept at the offside trap, holding all other attributes equal). Kroos' lack of training to make the runs of an average striker should be as abundantly obvious as his vastly superior linkup play, with the tradeoff representing a real tactical choice

Not to mention the issue that even huge differences in Decisions appears to have surprisingly moderate effects on in-game performance, before we've started scaling them based on a position familiarity scale and a hidden versatility stat.

Might revisit my "play every player in a position they have no familiarity in" experiment and see if I can get better results...

What is misunderstood about the relationship between determination and development?

2 hours ago, enigmatic said:

To be fair, I think real life players do lose instincts of a position when deployed exclusively elsewhere. Though this is obviously more to do with CA-budgeting (I find the across-the-board drop in significant attributes when a player at his potential improves in a new position far more annoying though. Great news, your DM is now accomplished at playing centre back. He's just less good at jumping, heading, marking and positioning than before he started intensively training for a role which prioritises those characteristics)

This really bothers me. Surely there's a way to reconcile learning a reasonable new position without actually becoming a worse player, especially when the trade off is either losing attributes used in said new position, or losing the ability to play a position that is the logical bridge between the most natural and the newly learned. A big example being when I tried to strengthen that CB's midfield ability by playing him as a DM and preparing him for even further up the pitch as a CM in training, yet he ends up unlearning DM to do so...

[enigmatic]

12 minutes ago, Weston said:

What is misunderstood about the relationship between determination and development?

It was widely understood for years by all the FM experts, mods, long term players and even people like me that the Determination attribute only affected player performances and possibly morale, and that it was the hidden attributes Professionalism and Ambition that influenced the chances of them gaining or losing ability over time (alongside factors like age and match time). People that suggested Determination helped players develop their abilities faster were "corrected".

Then one of the people responsible for testing the player development model popped into a thread and pointed out that actually Professionalism, Ambition and Determination all had a direct and approximately equal effect on whether players gained or lost ability over time.

 

13 minutes ago, Weston said:

This really bothers me. Surely there's a way to reconcile learning a reasonable new position without actually becoming a worse player, especially when the trade off is either losing attributes used in said new position, or losing the ability to play a position that is the logical bridge between the most natural and the newly learned. A big example being when I tried to strengthen that CB's midfield ability by playing him as a DM and preparing him for even further up the pitch as a CM in training, yet he ends up unlearning DM to do so...

To be fair, the losing of attributes only a major issue when a player is at his CA (and not going to naturally lose pace anyway) and has attributes better weighted for their new position (or several other positions) which isn't that often. And the decline is only [a fraction of] 1 point in a few attributes so it's more cosmetic than radically performance-altering. But it is annoying when it happens.

I think the game causes players to slowly lose positional ability in secondary positions they didn't previously have and aren't playing or training in via a separate unrelated mechanism, but I could be wrong and that could be CA linked. I think if your player is losing positions quickly it probably is influenced by his hidden versatility attribute.

But if literally all the positional familiarity function does is effectively lower Decisions by a few points, I don't think the difference between "competent" and "natural" at DM is going to have any noticeable impact on a player with above average Decisions for the level anyway...

[The Enforcer]

3 hours ago, Weston said:

I just wish I knew why Suso (RW) was such trash when I played him at CAM for a season.

What are/were his attributes and how did his average rating compare at AMR with AMC?

[Weston]

48 minutes ago, enigmatic said:

To be fair, the losing of attributes only a major issue when a player is at his CA (and not going to naturally lose pace anyway) and has attributes better weighted for their new position (or several other positions) which isn't that often. And the decline is only [a fraction of] 1 point in a few attributes so it's more cosmetic than radically performance-altering. But it is annoying when it happens.

I think the game causes players to slowly lose positional ability in secondary positions they didn't previously have and aren't playing or training in via a separate unrelated mechanism, but I could be wrong and that could be CA linked. I think if your player is losing positions quickly it probably is influenced by his hidden versatility attribute.

But if literally all the positional familiarity function does is effectively lower Decisions by a few points, I don't think the difference between "competent" and "natural" at DM is going to have any noticeable impact on a player with above average Decisions for the level anyway...

The DM position lost one level of familiarity the same day the CM positioned gained one, so it was very clearly taking the ability from that one area and moving it to another. To add to the frustration he kept his small amount of familiarity at LB which he will never, and to my knowledge has never, played.

[Weston]

34 minutes ago, The Enforcer said:

What are/were his attributes and how did his average rating compare at AMR with AMC?

They were about the same - he had attributes to be a leading AMC trequartista. I could get a screenshot later when I get home if you're really curious. AMC was dark green, but he consistently turned in 6.4 type appearances with precious little productive output.

[The Enforcer]

@Weston

It seems like there are some low hidden attributes at work here if he under performs both as an AMR and AMC. What is his decisions attribute?

[Seb Wassell]

23 hours ago, Weston said:

I'm not loaning this lad out for the upcoming season just so I can keep him at my excellent training facilities and drill into him that he's becoming a striker now instead of a winger. We'll see if it ever actually happens.

After the age of 18 match experience at an appropriate level is more important than training facilities.

[Seb Wassell]

Training a player in one position and playing him another is never going to be the most efficient way of achieving a "Natural" in a new position. In almost all circumstances to reach "Natural" a player must in fact be playing in that position. That being said, "Accomplished" is perfectly acceptable and will not see a huge difference in performance from Natural.

Players have limits on how effective they can learn to become in new positions, this is based on factors such as age, attributes, current positions, Versatility, etc. Often these limits are around the Accomplished (15-17) point, especially for completely new positions. 

As alluded to above, if there is no room in the CA for improvement, either because CA=PA or the player has stopped developing, and a new position is to be learned sometimes another position will have to be forgotten to allow room for this in the CA. This is more likely with vastly different positions where attributes that were previously deemed unnecessary have now become vital, for example ST > DC: Tackling was previously weighted very low in the CA calculation but has now become of high importance and as such is going to shift the CA of the player. This is not always the case and will only occur as both the positions and CA are approaching max. An alternative to this that can be seen to occur in similar circumstances is a slight redistribution of attributes in order to allow for position retraining/CA balance.

[herne79]

20 hours ago, Weston said:

What is misunderstood about the relationship between determination and development?

 

20 hours ago, enigmatic said:

Then one of the people responsible for testing the player development model popped into a thread and pointed out that actually Professionalism, Ambition and Determination all had a direct and approximately equal effect on whether players gained or lost ability over time.

This thread .

[bar333]

A related question - does role suitability ever actually increase? Assuming the player is trained in that role and plays it. Also, if a player isn't natural in the select role, does it hit his decisions like an unnatural position or is there another penalty?

[HUNT3R]

3 hours ago, bar333 said:

A related question - does role suitability ever actually increase? Assuming the player is trained in that role and plays it. Also, if a player isn't natural in the select role, does it hit his decisions like an unnatural position or is there another penalty?

There's no penalty for playing a player in another role. Suitability is an opinion and one that can change if the attributes changes. Some roles also "need" a minimum CA to be deemed "suitable".

[Banditman]

5 hours ago, bar333 said:

A related question - does role suitability ever actually increase? Assuming the player is trained in that role and plays it. Also, if a player isn't natural in the select role, does it hit his decisions like an unnatural position or is there another penalty?

Role suitability increases based on . . . how suitable the player is for the role.  If you choose the role in the Training screen, you can see what attributes affect suitability for the role.  If you focus on training up those attributes, role suitability can and will increase.  Honestly, role suitability doesn't really matter . . . it's the attributes and fit in your system that matter.

If you've got a 5'4 Striker whose attributes say he's a Target Man, he's still not going to be very good in the air.  If your system requires your striker to win balls in the air, it doesn't matter if the game "thinks" that he can be a Target Man . . . he can't.   On the other hand, if the reason the game thinks your 5'4 striker is a TM is because he's got good Strength, Balance and Dribbling, then perhaps you can still use him as a TM by getting him the ball farther from goal and at his feet.

Role suitability is a general estimate . . . it's not a complete be-all end-all.

 

[The Enforcer]

5 hours ago, HUNT3R said:

Some roles also "need" a minimum CA to be deemed "suitable".

That would explain why the game thinks my star player is ineffectual in the Conference South as a wide playmaker even though he averages almost 11 in all of the recommended attributes for that role and is accomplished as a right midfielder.

Edited January 25, 2018 by The Enforcer

[Weston]

18 hours ago, The Enforcer said:

@Weston

It seems like there are some low hidden attributes at work here if he under performs both as an AMR and AMC. What is his decisions attribute?

Pulled up an old save to get screenshots. I never used him as an AMR because I played narrow formations. His average rating as an AMC for the season in all competitions was 6.74 but it felt even worse than that.

Blue / Gray attribute average for preferred role as AMR (Inside Forward): 15.67 / 14.63

Blue Gray attribute average for most commonly used role as AMC (Trequartista): 15.25 / 15.25

Now to be fair the whole squad underperformed that season and we fell short of expectations, but I feel like that was BECAUSE of players like Suso being in such poor form. And I don't know what could've caused it other than him being played out of position as that's the only visible factor that wasn't aligned in a seemingly logical way. Sold him to Porto for a paltry 8 million to get his wages off the books where he promptly turned in a 7.30 average in his first season.

[Weston]

13 hours ago, Seb Wassell said:

After the age of 18 match experience at an appropriate level is more important than training facilities.

But I want to train him to be a forward, and the previous season I loaned him out with the requirement he be used as a striker and the green dot didn't change at all the entire year...

[Weston]

13 hours ago, Seb Wassell said:

Training a player in one position and playing him another is never going to be the most efficient way of achieving a "Natural" in a new position. In almost all circumstances to reach "Natural" a player must in fact be playing in that position. That being said, "Accomplished" is perfectly acceptable and will not see a huge difference in performance from Natural.

Players have limits on how effective they can learn to become in new positions, this is based on factors such as age, attributes, current positions, Versatility, etc. Often these limits are around the Accomplished (15-17) point, especially for completely new positions. 

As alluded to above, if there is no room in the CA for improvement, either because CA=PA or the player has stopped developing, and a new position is to be learned sometimes another position will have to be forgotten to allow room for this in the CA. This is more likely with vastly different positions where attributes that were previously deemed unnecessary have now become vital, for example ST > DC: Tackling was previously weighted very low in the CA calculation but has now become of high importance and as such is going to shift the CA of the player. This is not always the case and will only occur as both the positions and CA are approaching max. An alternative to this that can be seen to occur in similar circumstances is a slight redistribution of attributes in order to allow for position retraining/CA balance.

Are you saying if you train a player to play a different position his attributes are better suited to his overall ability could actually decrease?

Also, how long on average does a reasonably young player need to play and/or be trained in a new position to finally see a difference? Because it feels like it just never happens for me.

Edited January 25, 2018 by Weston

[Weston]

4 hours ago, Banditman said:

Role suitability increases based on . . . how suitable the player is for the role.  If you choose the role in the Training screen, you can see what attributes affect suitability for the role.  If you focus on training up those attributes, role suitability can and will increase.  Honestly, role suitability doesn't really matter . . . it's the attributes and fit in your system that matter.

If you've got a 5'4 Striker whose attributes say he's a Target Man, he's still not going to be very good in the air.  If your system requires your striker to win balls in the air, it doesn't matter if the game "thinks" that he can be a Target Man . . . he can't.   On the other hand, if the reason the game thinks your 5'4 striker is a TM is because he's got good Strength, Balance and Dribbling, then perhaps you can still use him as a TM by getting him the ball farther from goal and at his feet.

Role suitability is a general estimate . . . it's not a complete be-all end-all.

 

But I've seen players listed more as comfortable playing roles their attributes are on average lower for, and shown as unsuitable for other roles their attributes are on average higher for. Just now on my team I have a player who has higher average blue attributes to play a role he has half yellow circle suitability for than the one he has bright green full circle suitability for.

If it's purely a mathematical average that decides the colors then am I being confused by the decimals skewing my math? Also why would the same role be different colors in different positions (BWM in DM vs CM for example) if the attributes don't change?

[enigmatic]

Suso's shown as more suitable for an inside forward AMR role despite having slightly lower figures for the theoretically important visible attributes because he's 20 out of 20 for AMR positional familiarity and probably about 17 for AMC.

I wouldn't pay too much attention to the role suitability suggestions anyway; they're purely cosmetic.

The CA algorithm that might actually force a few attributes down if a player's already hit his potential is separate and different.

And the real reasons for Suso's underperformance might be different again (hidden mentals, tactics, PPM that makes him unwilling to use his right foot for anything, playing a lot of games as sub where he doesn't have time to influence the game)

Edited January 25, 2018 by enigmatic

[Seb Wassell]

Role suitability is based on the attributes recommended for that role. It is purely a graphical representation of attributes vs. role and does not work in the same way as positional ability.

[macca72]

There’s 100% something not working with this though.

I posted ages ago about some strange things, players natural in a position and having no full circle roles, I’ve also regularly seen players who start the game with 3/4 full circles and weeks later with no change in atttibutes go to almost empty.

[Seb Wassell]

1 minute ago, macca72 said:

There’s 100% something not working with this though.

I posted ages ago about some strange things, players natural in a position and having no full circle roles, I’ve also regularly seen players who start the game with 3/4 full circles and weeks later with no change in atttibutes go to almost empty.

It's relative to squad/league. As mentioned above some roles also require a minimum CA. However, if you reckon you've uncovered a bug please do get it posed over in the bugs forum and we'll take a look  

[macca72]

I will Seb, I’ve seen it frequently with Kemar Roofe (I suspect you’ve got a Leeds save) he suffers with both issues.

Edited January 25, 2018 by macca72

[The Enforcer]

7 hours ago, enigmatic said:

You beat me to it. I was going to say that the PPM would definitely steer me towards playing him as a right-sided inside forward. By playing him as AMC, he starts in a position where instead he ideally wants to be arriving there at full speed with the ball at his feet.

As for team averages, it is one of my historical pet hates that a team is underperforming in a match, but then score a goal against the run of play and all the ratings get a boost. Why should a player who creates three clear cut chances, but the striker fluffs them all, rate lower than one who creates two clear cut chances that are converted into assists by the striker. Admittedly, I haven't researched this thoroughly for FM18 - yet!

Edited January 25, 2018 by The Enforcer

[Weston]

10 hours ago, Seb Wassell said:

Role suitability is based on the attributes recommended for that role. It is purely a graphical representation of attributes vs. role and does not work in the same way as positional ability.

But that's what I'm saying, sometimes players appear to have better attributes on average for roles they're supposedly less familiar with, and vice versa. Are there hidden factors to this as well?

[Weston]

Perhaps if you could answer this scenario it would help clear things up for me:

Say Player 1 has an average attribute rating of 17, and a coach rating of 4* in his preferred position. In a position he has a darker shade of green in but an equal attribute spread for the same coach rates him as only a 2*.

Player 2, meanwhile, has an average attribute rating of only 14, and a coach rating of 3*. His preferred position is the same position Player 1 is rated 2* in.

Now, is the coach correct in saying that Player 2 will perform 1* better than Player 2* because of positional familiarity despite having lower attributes, or is the coach incorrect and Player 1 only takes a hit to decisions and really would still be better overall?

I guess what frustrates me is how star ratings suddenly seem completely false when applied to secondary positions, if what you're saying is true. They can't both be accurate, because they appear to directly oppose each other, unless the game intentionally has coaches misunderstand the game mechanics to force us to think independently or something, which would be a bit awkward.

[enigmatic]

Coaches are wrong a lot, period, but they're definitely very focused on positions when recommending players for a particular position.

Of course there are other reasons why a player with lower average attributes might be a better player anyway (e.g. average technical abilities but high pace)