Just found something amazing!!! Must see!!!

[Ljuba82]

Haha. This isn't going to be "fixed" today or tomorrow. And the reason you're having trouble playing is because you've fiddled with it so much. Like what someone said early on in the thread, ignorance can be bliss! As many many people have noted within the thread, this position-driven attribute inflation doesn't hamstring core gameplay per se. It does however affect the relative attribute strength of players as we've noted. So if you want to play FM, you can either just work around it quite readily: Don't train strikers to be defenders, for one. It's all in the mind.

I just can't...every time I load FM...I remember this hahhaha...and just don't have passion for playing...

I don't think they can change code for FM 2009, even for 2nd patch, but if they don't change it for FM10, it's catastrophy!!!

@VMX

...your system would be ok if attributes would have PA. But they don't. So...if yuo give to player only one PA...in theory it will be possible that he train his tackling or marking from 1 to 16 for example in expense of his attacking attributes. Not realistic

...anyway...I think we agree that such system would be very complicated cause researchers will have to look at all of his positions, whenever they change something. Very time-consuming!!!

...one way of solving this problem is just that player reject to learn new position, but that wouldn't be realistuic too.

In RL there are lot of plkayers who can play in lot of positions.

P.S. Today I founded a player from FM09, Viola, plays in Spain, very young player...but he can play both as NATURAL ST and DC. His tackling, marking, positioning are very low, while his attacking att are not bad.

I don't know his CA, PA, because I don't have any editor, but it's not important. Important is that he maybe have 80 CA and that with little training (decreasing defending, increasing attacking att), he can play like he have 120CA. Very significant for lower leagues.

[RT--]

I actually think this process of spreading out attributes according to position-weightings is a much bigger problem than some people are appreciating. It doesn't need us to train players in different positions to cause problems.

Simply put, players who have a high number of positions have much better attributes, in general, than those who only have one, when you control for Current Ability.

More specifically, players who player in a number of positions - but whose 'important' attributes for some of those positions are low - have much higher attributes than those who can only play in one position, when you control for Current Ability.

A quick example, comparing two 'natural' DCs:

Now if you look at the attributes that are important for a DC (let's say: heading, marking, tackling, composure, decisions, anticipation, positioning, jumping, pace, strength), you can see there's very little difference between these two players.

Yet I will bet that Gamberini has a significantly higher Current Ability than Fazio (in the game, he's rated by my Real Madrid scout as four stars for CA, whereas Fazio is rated as two stars). I'm sure someone with an editor can check, to confirm this?

The major difference is that Fazio is also accomplished as a DMC and a MC. And yet he has quite low attributes for things like passing, technique, creativity, first touch, which are important for a MC. So his CA has been spread around, effectively giving him better DC attributes.

This is important, of course, because we know that the match engine only uses attributes, and not CA.

So Fazio will effectively be just as good at centre-back as Gamberini, despite having a lower CA.

Throughout the game, there will be these imbalances with players who have a high number of 'accomplished' positions set. They'll perform better in certain positions than they should, according to their CA, because the attributes important to their position will be relatively higher.

If we used the editor to do nothing other than give Gamberini 'accomplished' DMC and MC stats, I think we'd find that his DC attributes would suddenly jump up a bit.

In the meantime, the AI teams (and scouts/coaches) will rate Gamberini as a much better DC than Fazio, but yet they'll perform quite similarly because of their attributes. This, to me, gives us an unfair advantage over the AI, and is why I think this problem justifies having a closer look at.

[marcoavgousti]

Sort of like a carton of orange juice; the orange juice company can put it together, but what goes into it really doesn't concern me. I just like to enjoy it.

Lol nice philosophy. lol made me laugh.

[Mitja]

I actually think this process of spreading out attributes according to position-weightings is a much bigger problem than some people are appreciating. It doesn't need us to train players in different positions to cause problems.

Simply put, players who have a high number of positions have much better attributes, in general, than those who only have one, when you control for Current Ability.

More specifically, players who player in a number of positions - but whose 'important' attributes for some of those positions are low - have much higher attributes than those who can only play in one position, when you control for Current Ability.

A quick example, comparing two 'natural' DCs:

Now if you look at the attributes that are important for a DC (let's say: heading, marking, tackling, composure, decisions, anticipation, positioning, jumping, pace, strength), you can see there's very little difference between these two players.

Yet I will bet that Gamberini has a significantly higher Current Ability than Fazio (in the game, he's rated by my Real Madrid scout as four stars for CA, whereas Fazio is rated as two stars). I'm sure someone with an editor can check, to confirm this?

The major difference is that Fazio is also accomplished as a DMC and a MC. And yet he has quite low attributes for things like passing, technique, creativity, first touch, which are important for a MC. So his CA has been spread around, effectively giving him better DC attributes.

This is important, of course, because we know that the match engine only uses attributes, and not CA.

So Fazio will effectively be just as good at centre-back as Gamberini, despite having a lower CA.

Throughout the game, there will be these imbalances with players who have a high number of 'accomplished' positions set. They'll perform better in certain positions than they should, according to their CA, because the attributes important to their position will be relatively higher.

If we used the editor to do nothing other than give Gamberini 'accomplished' DMC and MC stats, I think we'd find that his DC attributes would suddenly jump up a bit.

In the meantime, the AI teams (and scouts/coaches) will rate Gamberini as a much better DC than Fazio, but yet they'll perform quite similarly because of their attributes. This, to me, gives us an unfair advantage over the AI, and is why I think this problem justifies having a closer look at.

you don't know for sure how these two players are set by researchers, but you're right gamberrini's CA is higher than fazio's. also fazio is right-footed only, which most probably increased his stats. there is also a posibility that their CA wasn't set acuratlly by researchers. also (maybe) the game added fazio's MC ability to accomplished, as it assumes that every DM should be able to play as MC also...

quite a buggy thing, isn't it.

[RT--]

you don't know for sure how these two players are set by researchers, but you're right gamberrini's CA is higher than fazio's. also fazio is right-footed only, which most probably increased his stats. there is also a posibility that their CA wasn't set acuratlly by researchers. also (maybe) the game added fazio's MC ability to accomplished, as it assumes that every DM should be able to play as MC also...

quite a buggy thing, isn't it.

Yeah, I noticed the footedness thing just after I posted, which complicates it a bit further.

I guess you'd need to do a proper test:

Let's say we take two players, Player 1 and Player 2, and give them both 150CA, exactly the same attributes (we'll make them good defenders but useless attacking players), and make them both 'natural' DCs.

In the game, they should appear almost exactly the same and will perform similarly in the match engine.

Now we make Player 2 'accomplished' as a MC, but change nothing else.

In the game, Player 2 will (I predict) now have better defensive attributes than Player 1, and so will perform better there in the match engine. Remember, we haven't changed his attributes, or his CA, we've just said that he's okay at playing in midfield.

I think the AI will rate both players as being equally good at DC. Yet Player 2 will perform better there, because his defensive stats will be better.

[RT--]

I have no problem with WFA affecting attributes, because WFA has an effect in the match-engine, so it balances itself out (though it makes it harder to compare players properly, and certainly makes it more difficult for the 'casual' gamer, who has no idea about WFA and how important that is).

But extra positions changing attributes (and I can see why it happens, because of the weighting system for each position) causes a problem. Because it will effectively increase the attributes relevant to a certain position, and have an effect in the match-engine, but because the CA is the same, the AI can't see this and so will rate a player lower than his attributes should justify.

Now I can sign 'versatile' players with lower CA (and therefore lower reputation, asking price) and get them to do a good job in a given position. And yet my coaches will be telling me that player's no good in that position, even when I know he is. The AI is hamstrung by the fact that it only 'sees' CA and position data when it looks at players.

[Mitja]

if that's true it's crazyness.

maybe we should just play the game less you know the better.

[RT--]

if that's true it's crazyness.

maybe we should just play the game less you know the better.

One option would be to code things so that the AI (opposition managers/coaches/scouts) judge players on attributes, rather than just CA/position. I'd particularly like to see this done so that AI managers would look for players, in a certain position, that have the attributes to match their tactics and squad. So an AI team with a big target man might look to go out and sign a winger who had particularly good crossing. Or an AI manager with two good crossing wingers might go out and look for a striker with really good jumping. Or managers who played man-marking would tend towards defenders with good marking. You get the idea...

At the moment, I'm pretty sure they don't do that. They identify that they need a striker, and then try and sign one purely based on CA/position (and other things, like their budget, who they can attract, obviously). But attributes don't really come in to it.

Overall, though, changing things that way would only goes so far to address this problem. CA is crucial, because it's a very simple research tool. It's just that it doesn't match up perfectly with position-based attributes at the moment, and SI need to try and think of a way to do something about that.

[matt_132]

its obviously not a problem, sure you can create a sick player but most people will feel they are cheating. It would only be a problem if the AI managers started using this tactic to create amazing players

[Ljuba82]

@RT--

Just checked your example...

...when I entered attributes of Gamberini...I get 171 CA (I enter that his weaker foot is 10, cause we don't know exactly the rating, but his CA is aproximately 171).

...when I entered attributes of Fazio...I get 160 CA (set weaker foot to 1 and I set DM and MC rating of 15).

...but...if you remove his DM, MC rating, his CA is now 179!!!

...so at DC his is significantly better player than Gamberini (which my experienced eye saw when I was comparing them on your screenshot )

P.S. REMEMBER that positional rating can make a player worse. For examle DC with low defensive attributes and with high MC attributes, but positional rating for MC awkward or so....

So...we just know that several positions change players overal ability on pitch, but that doesn't mean that he will be better if he has more positions, he may be worse!

[Ljuba82]

...one more thing RT--

...I don't belive that scouts always compare players CA...

When I made this test with "Ibrahimovic supermen"...he had 7 rating report, while some excelent ST in game with similar CA (Eto, Torres, Henry, Aguero) had only 4 or 5 stars.

[VMX]

Well, I don't know... but it'll surely be nice if someone from SI would give us a clue about how the system exactly works... (i.e.: if AI teams do take the attributes into consideration or not) and if they have any intention of changing this system in the future. If not, it's not very useful to spend our time trying to give them possible solutions when they're not planning to use any.

However, if they are actually listening and looking for possible alternatives, I'll be glad to keep trying to find a good idea that can replace/modify the current system.

But it would be crucial to know if SI need a system where the global ability of a player can be judged fully by his CA (this is: if the AI can only compare CA's and not attributes).

[G-Man11]

For those saying you don't get players who are natural DC's and ST's then I say look no further than Dion Dublin!

[Majkee]

Well, I don't know... but it'll surely be nice if someone from SI would give us a clue about how the system exactly works... (i.e.: if AI teams do take the attributes into consideration or not) and if they have any intention of changing this system in the future. If not, it's not very useful to spend our time trying to give them possible solutions when they're not planning to use any.

However, if they are actually listening and looking for possible alternatives, I'll be glad to keep trying to find a good idea that can replace/modify the current system.

But it would be crucial to know if SI need a system where the global ability of a player can be judged fully by his CA (this is: if the AI can only compare CA's and not attributes).

Yes, it is strange that nobody from SI has commented yet. I think that Ljuba and others here hit the nail on the head

Any comments from SI?

[RT--]

its obviously not a problem, sure you can create a sick player but most people will feel they are cheating. It would only be a problem if the AI managers started using this tactic to create amazing players

Matt - as the Fazio/Gamberini example indicates, this isn't just a problem if you decide to train players up in multiple positions. It's a problem with existing players who have multiple positions set, when you compare them with players who only have one position set.

[Majkee]

Matt - as the Fazio/Gamberini example indicates, this isn't just a problem if you decide to train players up in multiple positions. It's a problem with existing players who have multiple positions set, when you compare them with players who only have one position set.

But is this really true, that scouts are reporting back based on CA/positional ability? Are they really not reporting on attr/positional ability as it is in match engine?

[Ljuba82]

...yeah...it's strange nobody replies...

But is this really true, that scouts are reporting back based on CA/positional ability? Are they really not reporting on attr/positional ability as it is in match engine?

...as I said Majkee

...I don't belive that scouts always compare players CA...

When I made this test with "Ibrahimovic supermen"...he had 7 rating report, while some excelent ST in game with similar CA (Eto, Torres, Henry, Aguero) had only 4 or 5 stars.

...further explanation of SI how scouting works will help.

[Majkee]

...yeah...it's strange nobody replies...

...as I said Majkee

...I don't belive that scouts always compare players CA...

When I made this test with "Ibrahimovic supermen"...he had 7 rating report, while some excelent ST in game with similar CA (Eto, Torres, Henry, Aguero) had only 4 or 5 stars.

...further explanation of SI how scouting works will help.

SI should explain all the things you have discovered PaulC may be smiling secretly now

[Ljuba82]

SI should explain all the things you have discovered PaulC may be smiling secretly now

I don't believe they are smiling...it's more something like headache...

...but this stuff relly needs explanation...

[RT--]

...one more thing RT--

...I don't belive that scouts always compare players CA...

When I made this test with "Ibrahimovic supermen"...he had 7 rating report, while some excelent ST in game with similar CA (Eto, Torres, Henry, Aguero) had only 4 or 5 stars.

Perhaps you're right, but in the case of Fazio, multiple scouts (and my assistant, when I went to sign him) said they didn't recommend signing him as a DC, whereas they all thought Gambrini was a 'good player'.

And yet, as we've established, they'd both probably play equally well in that position.

[Hamenaglar]

Perhaps you're right, but in the case of Fazio, multiple scouts (and my assistant, when I went to sign him) said they didn't recommend signing him as a DC, whereas they all thought Gambrini was a 'good player'.

And yet, as we've established, they'd both probably play equally well in that position.

Scout ratings aren't perfectly reliable. Just as a player with 20 finishing won't score on every goal, the same way, all scouts will get it wrong at one point.

[RT--]

Scout ratings aren't perfectly reliable. Just as a player with 20 finishing won't score on every goal, the same way, all scouts will get it wrong at one point.

No, of course, although these were six or seven Real Madrid scouts (some of whom have 20/20 JA and JP) and my assistant, all saying exactly the same thing.

Obviously, on Friday, we can check this out with a little bit more accuracy, but I think we'll find that adding an extra position changes position-specific attributes, either for the better or worse, despite not changing ability.

[Majkee]

No, of course, although these were six or seven Real Madrid scouts (some of whom have 20/20 JA and JP) and my assistant, all saying exactly the same thing.

Obviously, on Friday, we can check this out with a little bit more accuracy, but I think we'll find that adding an extra position changes position-specific attributes, either for the better or worse, despite not changing ability.

Which is actually disserviceable.

Digress from current topic, what is better? To have one footed player or either footed?

(Back to what Ljuba wrote:

...all other attributes are same.

AML with right foot 20, left 19...........179 CA

AML with left foot 20, right 19............174 CA

AML with right foot 20, left 1..............145 CA

AML with left foot 20, right 1..............160 CA

AML with right foot 20, left 15.............172 CA

AML with left foot 20, right 15.............171 CA)

Because this is also illogical and I do not have impression, that we agreed on his topic, what is generally better.

[Gooner Steve]

Great find, if this was a tactic then people would be all over it and with lots of praise too.

Dion Dublin, Paul Warhurst, Marcus Gayle all ST who moved to DC in real life. Also I think Darren Purse (DC) started life as a ST.

Chelsea losing to Roma they send Terry upfront for the closing mins.

So im playing my game and I see these examples and think right im going to train Adebayor to a DC so he can help protect a lead. I really dont think this is that unrealistic to use, I always retrain players in my youth team!

Look at Gibbs last night, I remeber him being compared to Giggs when younger. Last night he was playing as LB. He still plays on the wing in the reserves but I guess this is his education to learn new skills and a new position.

[RT--]

It's probably not the best time to ask now, with them all working tirelessly ahead of Friday, but I think it would be really helpful if somebody from SI could clarify:

a) Whether they agree that it's possible that a player who has multiple positions could inadvertently end up with much better (or much worse) position-specific attributes for one of their positions than a player of comparable CA, who can only play in that one specific position. And if they do, whether they think this is a 'bug', or it's deliberate.

and

b) Whether the AI managers and in-game scouts/coaches use attributes to judge players, or whether they just use CA and position stats.

If a) is true, then I think the answer to b) is particularly important, because it dictates exactly how big a problem that it could cause in the game.

[iamjerome]

excellent thread. with regards to scout reports or general recommendation type things in the game however, i have a sneaky feeling they depend on the reputation of the player above anything else (as you may know there are different types of rep that would come in handy for a recommendation of a different sort of player eg. your own, in your league, in a different continent etc : world rep, local rep, current rep.

[zeusbheld]

I really hope that their there is no knee-jerk reaction from SI regarding this. As I and probably other use positional training, not to exploit but to improve players. Apart from the fact that you get players who cover several positions within your chosen formation.

agree. i think it's interesting to peek under the hood, but if you train your players as if they were real players then the boost just makes them more versatile. also i haven't played the demo that much yet but i'd imagine you can't just train a player at every position. in 2k8 players would 'forget' positions you train them in when you stopped training them at that position. presumably that still happens, and changes in ablity would fade as a result too.

i'm grateful that players can be made more versatile; it makes position-swapping tactics more practicala nd easy to cover several positions with a limited squad (especially if injuries persist as they have with the demo).

[Majkee]

in 2k8 players would 'forget' positions you train them in when you stopped training them at that position. presumably that still happens, and changes in ablity would fade as a result too.

I am not sure if this "fogetting to play on retrained positions" was relevant only to unaccomplished positions or for all retrained positions.

I think that only for retrained and unaccomplished positions imho - retrained and accomplished positions have not been forgotten if not trained consecutively.

[lemeures]

Good discussion, something I did with a keeper once thinking if he was a defender aswell he would be better at rushing out of his area. I think because i didn't use the editor to maximize these things the results weren't as potent as yours but were confusing all the same.

I disagree with the idea that too many positions have a negative result, with an example being someone that can play all positions on the right. If you look at most world class full backs they are incredibly balanced. How many people want a full back that can only defend and not make great runs down the flank, resulting in goals or crosses into the area? These are all attacking skills, much like many outside midfielders are also expected to work hard and defend when there full back is out of position.

The logical explanation for a right footed winger being better on the left would be the style of play that player has (see Ryan B at Liverpool), where the hold up of cutting back to cross is countered by the way he cuts inside to get to his stronger foot. I obviously haven't looked through all the players in the data base but believe it to be a commom rule that right footed left wingers play that way for that very reason.

The other problem with training players to other positions is they don't like it

[zeusbheld]

I am not sure if this "fogetting to play on retrained positions" was relevant only to unaccomplished positions or for all retrained positions.

I think that only for retrained and unaccomplished positions imho - retrained and accomplished positions have not been forgotten if not trained consecutively.

in 2008 i definitely had players 'forget' positioned when they had reached 'accomplished' status.

they don't 'forget' if they learned a position before they got to your club.

i haven't even completed a season in the Demo so i can't speak to whether it happens in 2009.

[Ben Jones]

Watch your backs. I've had a van parked outside my house for a week and everytime I use the phone I hear a clicking sound. We know to much!

[VMX]

Watch your backs. I've had a van parked outside my house for a week and everytime I use the phone I hear a clicking sound. We know to much!

hahahaha

[RT--]

I did a quick test with 08, earlier, looking at this position/attributes thing.

I gave 2 players 150CA, and exactly the same attributes (including hidden and WFA), and the same reputation. They basically have pretty good defensive attributes, but poor attributes in other areas. The only difference between the two players is that Player 1 is a DC only, Player 3 is a DC with 'accomplished' for MC. (If you're wondering what happened to Player 2, I made him exactly the same as Player 1, just to check they'd come out the same - and they did).

Here's what happened:

You'll see that Player 3 has actually come out with a 1 point deficit in CA (142/143), and has a slightly better WFA (7/6). This isn't a problem, because it's a tiny difference, and both things would actually make us expect that Player 3 would have slightly lower attributes than Player 1, if anything.

But you'll see that Player 3 has actually had a boost in his defensive attributes AND a boost in his other attributes as well. He'll now effectively play better at DC than Player 1, in the match-engine, despite them having equal CA and WFA. Because they have the same reputation, CA and WFA, I'm fairly sure the AI will view them as being exactly the same. But which one would you sign as a DC?

It's not a dramatic difference, but I'd still be interested to hear an explanation of why it's logical for this to happen.

[RT--]

Exactly the same thing (exactly the same change in all attributes, except 2 points change from teamwork > crossing) happens if you make Player 3 accomplished as a Striker instead of a MC, by the way.

[RT--]

And if we start to add multiple accomplished positions, you can see how it's possible to start getting a pretty good player with a 150CA:

[Ljuba82]

...to hear an explanation of why it's logical...

Well I just explain it on previous 2 pages

...it's because weight of defensing attributes for player 3 droped...and you can fill his CA with more attribute points.

Because he can play on MC, some attributes for MC position are now more important and are weighted more.

I'll try to explain as simple as I can...

Let's say that a player have only 2 attributes (not like in game 30+) and let's say that those attributes are TACKLING and FINISHING.

Let's say that weight for DC position for tackling is 10, and for finishing weight is only 1 (cause it's logic that tackling is more important for defenders).

...so...if you have a player with CA of 150, you can give him 14 in tackling attribute and 10 for finishing attribute.

14 (tackling attribute) * 10 (weight of tackling) + 10 (finishing attribute) * 1 (weight of finishing) = 150 CA points.

...but now...let's say that he can play as natural DC, but also as natural ST. Weight of his tackling attribute is no more 10, it's now 5, but weight of his finishing attribute is no longer 1, it's also 5 (logic is...if player can play equaly on both positions, both attributes have same importance.)

...and now with this new weightening...

14 (tackling att) * 5 (weight of tackling) + 10 (finishing att) * 5 (weight of finishing) = 120 CA pts

...so that means that you have 30 spare CA points to asign to him...so that means that you can give him aditional 6 pts for tackling. 6 *5 (weight) =30.

...and it's obvious that player with 20 tackling, 10 finishing is better than player with 14 tackling, 10 finishing...altough they have same CA.

Hope this helps.

...of course...in game is much more complex, but logic is same!!!

...I STILL WAIT FOR SI RESPONSE!!!

[RT--]

Well I just explain it on previous 2 pages

...it's because weight of defensing attributes for player 3 droped...and you can fill his CA with more attribute points.

Because he can play on MC, some attributes for MC position are now more important and are weighted more.

I'll try to explain as simple as I can...

Let's say that a player have only 2 attributes (not like in game 30+) and let's say that those attributes are TACKLING and FINISHING.

Let's say that weight for DC position for tackling is 10, and for finishing weight is only 1 (cause it's logic that tackling is more important for defenders).

...so...if you have a player with CA of 150, you can give him 14 in tackling attribute and 10 for finishing attribute.

14 (tackling attribute) * 10 (weight of tackling) + 10 (finishing attribute) * 1 (weight of finishing) = 150 CA points.

...but now...let's say that he can play as natural DC, but also as natural ST. Weight of his tackling attribute is no more 10, it's now 5, but weight of his finishing attribute is no longer 1, it's also 5 (logic is...if player can play equaly on both positions, both attributes have same importance.)

...and now with this new weightening...

14 (tackling att) * 5 (weight of tackling) + 10 (finishing att) * 5 (weight of finishing) = 120 CA pts

...so that means that you have 30 spare CA points to asign to him...so that means that you can give him aditional 6 pts for tackling. 6 *5 (weight) =30.

...and it's obvious that player with 20 tackling, 10 finishing is better than player with 14 tackling, 10 finishing...altough they have same CA.

Hope this helps.

...of course...in game is much more complex, but logic is same!!!

...I STILL WAIT FOR SI RESPONSE!!!

Thanks Ljuba - I should have made myself more clear. I realise how it happens (thanks to your explanations on this threads, and other thread I've read) in terms of the game, but (like you) I just don't get why it should happen like that. There are lots of players in the game with multiple positions who benefit from this, so it seems to be a fundamental problem with the attribute weighting, and I'm surprised SI haven't tried to come up with some way around it.

Maybe they'll come and explain that they think it's completely logical, and it'll make perfect sense to us all!

[RT--]

By the way, from an earlier discussion...

***CA SPOILERS***

I had a look at the editor yesterday for the full game (you can play around with it when you pre-download the game, without activating) and Fazio has a CA of about 150, Gamberini about 165.

If you go back and look at Fazio's attributes, that's quite an impressive achievement for a centre-back with CA 150. He is, as you might say, 'batting above his weight', there!

***END OF CA SPOILERS***

I guess it really helps to have some extra accomplished/competent positions

[voidex]

I really hope SI will address this fundamental issue.

[Majkee]

And if we start to add multiple accomplished positions, you can see how it's possible to start getting a pretty good player with a 150CA:

So it is possible to create monster...via retraining him. I am just wandering if this is possible IRL. May be it is possible but only with raising of CA too, not without it as it is in game. If you have better attr, you must have better CA IRL.

From my point of view, it is exploit and I want explanation from SI. What about you?

[RT--]

So it is possible to create monster...via retraining him. I am just wandering if this is possible IRL. May be it is possible but only with raising of CA too, not without it as it is in game. If you have better attr, you must have better CA IRL.

From my point of view, it is exploit and I want explanation from SI. What about you?

It's something which could certainly be exploited, I agree. I'm not sure how attributes would progress across a player's development, but if (as Ljuba said) you trained a DC to be a MC and ST (how easy that is, I'm not sure), and kept them on high defensive training, I think you'd effectively boost them as a DC, and make them better at that position than they could have been without the extra positions.

I'm more concerned, at the moment, about players who start out as naturally being good at various positions - I don't play FML or network games, so I won't come up against any training exploits. As you can see from Fazio, though, players can be really effective at one of their positions without having the CA to go with it. I still don't see any evidence that the AI would be able to spot this, either.

Fazio, in the editor, has been given 14 for tackling, but comes out with 16 in the game. Likewise, my imaginary 'Player 3' jumps from 14 for tackling to 17, in the game. I'm sure a lot of players have their stats adjusted from the editor, but the researchers can't be expected to anticipate a change of 2/3 points for key attributes.

I should say, that for some perspective, I don't think this is anywhere near a game-breaker, and it won't stop my enjoyment of the game. I think it's a quirk, though, and definitely worth SI having a look at.

[RT--]

I am just wandering if this is possible IRL. May be it is possible but only with raising of CA too, not without it as it is in game.

Well if, IRL, I took Nemanja Vidic, let's say, and started giving him some coaching on how to play as a striker - after 6 months, 1 year, when he'd become 'competent' at it - I'd be pretty amazed if his tackling and marking had suddenly become better. And not only better, but better than it could possibly have been if I'd never taught him to play as a striker.

By the same reasoning, I'd be surprised if my coaching staff couldn't see that - because he was now better at tackling and marking - he'd be better at playing centre-back than he was before.

[Majkee]

Well if, IRL, I took Nemanja Vidic, let's say, and started giving him some coaching on how to play as a striker - after 6 months, 1 year, when he'd become 'competent' at it - I'd be pretty amazed if his tackling and marking had suddenly become better. And not only better, but better than it could possibly have been if I'd never taught him to play as a striker.

By the same reasoning, I'd be surprised if my coaching staff couldn't see that - because he was now better at tackling and marking - he'd be better at playing centre-back than he was before.

Yes it is illogical, but that is the negative cause of attribute weighting.

[RT--]

Yes it is illogical, but that is the negative cause of attribute weighting.

Yep. The problem is, I can't think of a 'quick-fix' solution to that problem.

I think any solution that addresses this problem would have to make changes to the way that attribute weightings/CA/positions work, and that's not going to be easy.

My Player 3 has better or equal stats to Player 1 in every department, can play more positions and has an equally good weaker foot. Yet they both have the same ability and (I still suspect, all things considered) will be rated as equals by the AI.

[Majkee]

Yep. The problem is, I can't think of a 'quick-fix' solution to that problem.

I think any solution that addresses this problem would have to make changes to the way that attribute weightings/CA/positions work, and that's not going to be easy.

My Player 3 has better or equal stats to Player 1 in every department, can play more positions and has an equally good weaker foot. Yet they both have the same ability and (I still suspect, all things considered) will be rated as equals by the AI.

I am looking forward to the days after Patch 1 will be released, especially if someone from SI will spare some time with us on this topic or not.

[IzzetLeicester]

There should be an option in the editor to check what the player will look like when a new game starts up. It will allow the researchers and whoever likes to tinker around in the editor to see if the attributes match the CA before they finish their work. This will stop the database imbalance.

[PaulC]

It's probably not the best time to ask now, with them all working tirelessly ahead of Friday, but I think it would be really helpful if somebody from SI could clarify:

a) Whether they agree that it's possible that a player who has multiple positions could inadvertently end up with much better (or much worse) position-specific attributes for one of their positions than a player of comparable CA, who can only play in that one specific position. And if they do, whether they think this is a 'bug', or it's deliberate.

and

b) Whether the AI managers and in-game scouts/coaches use attributes to judge players, or whether they just use CA and position stats.

If a) is true, then I think the answer to b) is particularly important, because it dictates exactly how big a problem that it could cause in the game.

a) We are going to investigate further, but right now we are snowed under

b) They use an estimation of CA, but for some purposes also look at a select few attributes.

Paul

[matt_132]

i understand the problem now, and it suggests also that scout reports etc are pretty much useless.

However, could it not be that a player performs better if he has a higher CA. Maybe how good a player is at tackling for example, depends not only on his tackling attribute, but also on his CA.

This is just a guess and it is still pretty illogical but not more than the current system is.

Great find btw Ljuba, very interesting

[RT--]

a) We are going to investigate further, but right now we are snowed under

b) They use an estimation of CA, but for some purposes also look at a select few attributes.

Paul

Good man, Paul - thanks for the reply.

I can appreciate you're all busy right now, of course

[RT--]

However, could it not be that a player performs better if he has a higher CA. Maybe how good a player is at tackling for example, depends not only on his tackling attribute, but also on his CA.

This is just a guess and it is still pretty illogical but not more than the current system is.

I'm fairly sure that SI have confirmed in the past that the match-engine only uses attributes, and other things like morale - it doesn't use Current Ability.