It's so annoying that 10.3 requires a CB to be tall

[pauliowigs]

^^^There is a fourth band. I personally feel headers won is too much of an influence on match ratings, ahead of say interceptions/tackles for defenders. Its just the way match ratings are weighted that is at present slightly unbalanced, and that is the only point I care about in this thread. Same as dribbles for wingers, even though they might not be that effective. Same with headers for strikers, even if those headers didn't create anything.

That is why defenders with excellent jumping, but poorer attributes in other areas, get higher ratings than their ability/performance deserves.

Yes, but the result of what you were saying would mean it would be difficult to get centrebacks with low jumping to perform well when used ordinarily over a series of games.

I totally agree with what your saying regarding jumping/heading having too much influence on players. when i said that one group of people was:

"that it is too difficult to get good centrebacks with lower than ideal jumping to perform well over the course of a season. They appear to play ok in the match engine but they get low average ratings over the course of a season."

what your saying seems to agree reosonably well with this... I believe I have made the point that you're making(although perhaps not as well) in one of my earlier posts and i suggested that the solution may be to not punish centrebacks for lost headers that are not key headers or that do not lead to goals or tactical exposure. shutting the other team out of the game with key interceptions and close marking of players is as important to a defenders game as winning headers. Losing a couple of headers well outside the box when you have other defenders that can cover, for example, should not really be a major mark down.

This does not dispute what you are saying, you fall fine into that first category as far as I can tell (otherwise i don't either). You are offering up a very good reason as to WHY they may be getting too harsh a rating, and its a reason that I personally think probably hits the nail on the head.

Theres also been some discussion on here as to how you can sometimes get them to play well by making using certain specialist tactics that seem to shy the player in question away from headers, which seems to further back up what you are saying.

[SFraser]

Theres also been some discussion on here as to how you can sometimes get them to play well by making using certain specialist tactics that seem to shy the player in question away from headers, which seems to further back up what you are saying.

Supporting a Centreback that is weak in the air with someone strong in the air is just basic common sense, irrespective of any potential ratings bias.

Points like these only reinforce the perception that you want the game tweaked to make up for having absolutely no clue about football.

[x42bn6]

I don't think anyone here is suggesting that is not a possibility.

And that is why I'm putting evidence out to suggest it is.

This is where you become rediculous.

I'm really wondering if you read my posts.

I've argued about trends, Law of Large Numbers, more data and so on. All to justify a general trend. One or two counterexamples are meaningless to me because I'm aware the probability of finding an outlier for data that are not perfectly correlated is greater than zero.

[Lankylars]

I'm not quite sure why SFraser and x42bn6 are arguing, because your suppositions are not mutually exclusive. SFraser is right when he says that it is possible for a DC with a low jumping attribute to get good match ratings, and he has probably found a way to do it with his tactics. As a human player, this is probably true much of the time. x42bn6 says that looking at all the matches conducted under AI managers in the game, the is a correlation of .69 between winning headers and match rating. Firstly, these data come from the AI. Different than a human manager choosing things. Secondly, .69 is statistically significant, but by no means does it imply that winning headers is the only thing that contributes to match ratings, nor does x42bn6 say it is. You two are arguing about different things, clearly, or you have misunderstood the point that each other is trying to make.

Now, as far as evaluating the .69, I'm not sure that anyone here has the experience or ability to determine if that is too high a correlation. This is something that would involve a lot of real life stat tracking and match watching.

[SFraser]

I've argued about trends, Law of Large Numbers, more data and so on. All to justify a general trend. One or two counterexamples are meaningless to me because I'm aware the probability of finding an outlier for data that are not perfectly correlated is greater than zero.

Well so long as you don't rule out the possibility of achieving high ratings in Centrebacks with low Jumping, we have nothing to argue over. We can now discuss something relevent.

Have you found any actual evidence to support the theory that header win % is too highly weighted?

[x42bn6]

I'm not quite sure why SFraser and x42bn6 are arguing, because your suppositions are not mutually exclusive. SFraser is right when he says that it is possible for a DC with a low jumping attribute to get good match ratings, and he has probably found a way to do it with his tactics. As a human player, this is probably true much of the time. x42bn6 says that looking at all the matches conducted under AI managers in the game, the is a correlation of .69 between winning headers and match rating. Firstly, these data come from the AI. Different than a human manager choosing things. Secondly, .69 is statistically significant, but by no means does it imply that winning headers is the only thing that contributes to match ratings, nor does x42bn6 say it is. You two are arguing about different things, clearly, or you have misunderstood the point that each other is trying to make.

Now, as far as evaluating the .69, I'm not sure that anyone here has the experience or ability to determine if that is too high a correlation. This is something that would involve a lot of real life stat tracking and match watching.

Which is why I don't know why SFraser is belittling every one of my posts.

Well so long as you don't rule out the possibility of achieving high ratings in Centrebacks with low Jumping, we have nothing to argue over. We can now discuss something relevent.

Have you found any actual evidence to support the theory that header win % is too highly weighted?

And I really doubt you read my posts!

Correlation 0.69 suggests strong positive correlation. What "too highly weighted" means is up for debate. But to me 0.69 does suggest too much of a weighting simply because it means that a lot of the time poor Jumping will influence the ratings rather heavily.

[SFraser]

Correlation 0.69 suggests strong positive correlation. What "too highly weighted" means is up for debate. But to me 0.69 does suggest too much of a weighting simply because it means that a lot of the time poor Jumping will influence the ratings rather heavily.

And you have proven that Jumping is the only attribute involved in Header Win % calculations?

I wonder why you did not look at Header Win % rather than Jumping? Did you immediately jump to the same conclusion without evidence, arguement or reason as the OP?

[x42bn6]

And you have proven that Jumping is the only attribute involved in Header Win % calculations?

I wonder why you did not look at Header Win % rather than Jumping? Did you immediately jump to the same conclusion without evidence, arguement or reason as the OP?

I've not proven Jumping is the only attribute involved. I've proven nothing - proof is difficult with a black box. But the large amount of data suggests that at least some of the other factors have been averaged out and the pool of results is a mixture of lots of other cases which reduces the uncertainty (again, Law of Large Numbers).

I posted that data because you were posting lots of screenshots of "low" Jumping and "high" ratings.

The data suggests a strong correlation and that is my conclusion.

[SFraser]

I've not proven Jumping is the only attribute involved. I've proven nothing - proof is difficult with a black box. But the large amount of data suggests that at least some of the other factors have been averaged out and the pool of results is a mixture of lots of other cases which reduces the uncertainty (again, Law of Large Numbers).

Jumping was an assumption made by the OP. The claim relies upon Header Win %

It is up to you to do this properly.

The data suggests a strong correlation and that is my conclusion.

Have you tested for false positives, i.e Finishing?

Have you tested for equivelance with other Key Defender Attributes such as Positioning?

[x42bn6]

Jumping was an assumption made by the OP. The claim relies upon Header Win %

It is up to you to do this properly.

I've also given you data for % Header wins (0.31) which also suggests a positive correlation, although not as strong. It's a lot harder to gather this data as no such column exists in the player view and I don't think I have enough data for that.

Have you tested for false positives, i.e Finishing?

You can't test for false-positives because the "truth" is what we seek - you're begging the question.

Have you tested for equivelance with other Key Defender Attributes such as Positioning?

No, but I'll be doing that one day.

[SFraser]

I've also given you data for % Header wins (0.31) which also suggests a positive correlation, although not as strong. It's a lot harder to gather this data as no such column exists in the player view and I don't think I have enough data for that.

So the correlation between Header Win % and Ratings is lower than the correlation between Jumping and Ratings?

I can't say I am a correlation expert like yourself, but shouldn't it be higher if the initial claim is true? Indeed shouldn't it be higher if Jumping is the key attribute in Header Win %?

You can't test for false-positives because the "truth" is what we seek - you're begging the question.

Surely if you found a correlation between Finishing and Ratings in Centrebacks that was higher than Jumping and Ratings, then the entire premise would have to be seriously reconsidered?

No, but I'll be doing that one day.

I look forward to it.

I also have some other suggestions for you to do you stuff with, if you have the time. Determination should be a hugely significant attribute in Centreback Ratings.

[x42bn6]

So the correlation between Header Win % and Ratings is lower than the correlation between Jumping and Ratings?

I can't say I am a correlation expert like yourself, but shouldn't it be higher if the initial claim is true? Indeed shouldn't it be higher if Jumping is the key attribute in Header Win %?

Jumping may be the key attribute in Header Win % which means that there are other things which affect Header Win % (namely the opposition which you are facing of course). Which is why I'm not surprised they are different. No need to jump to conclusions about why it's higher.

Surely if you found a correlation between Finishing and Ratings in Centrebacks that was higher than Jumping and Ratings, then the entire premise would have to be seriously reconsidered?

If I find a correlation between Ratings and the position of the moon, I may have a problem in my survey.

The fact is that I simply don't believe Finishing is an issue. Yes Finishing is always useful when in front of goal but defenders never get in front of goal often enough to shoot with their feet, so I see no reason why you should examine why Finishing affects rating. Yes it does but I don't think that matters much.

If I have a reason to believe Finishing will unduly affect the results then I will look at it. Otherwise I'm not going to go round studying every single possible effect like Corners and game time because "there might be a problem".

If you are so insistent feel free to do it yourself.

I also have some other suggestions for you to do you stuff with, if you have the time. Determination should be a hugely significant attribute in Centreback Ratings.

Could you not do it yourself?

The Excel function you need is CORR, and in OpenOffice CORREL.

[SFraser]

Jumping is the key attribute in Header Win % which means that there are other things which affect Header Win % (namely the opposition which you are facing of course). Which is why I'm not surprised they are different. No need to jump to conclusions about why it's higher.

No need to jump to any conclusions at all.

Where is your evidence that Jumping is the key attribute in header win %?

[x42bn6]

No need to jump to any conclusions at all.

Where is your evidence that Jumping is the key attribute in header win %?

"is" should be "may be". This was in response to:

Indeed shouldn't it be higher if Jumping is the key attribute in Header Win %?

[chaddy101]

Can you two not agree to disagree? Jesus Christ. :confused:

[deejay10]

Supporting a Centreback that is weak in the air with someone strong in the air is just basic common sense, irrespective of any potential ratings bias.

Points like these only reinforce the perception that you want the game tweaked to make up for having absolutely no clue about football.

Suggesting someone has absolutely no clue about football, when you yourself only make your genius tactical observations with Manchester United is a pathetic way to respond.

Personally I consider myself very clued up tactically, and nowhere in this whole thread have I said I struggle to make defenders with low jumping play decently. I just feel the way match ratings is weighted at present is slightly unbalanced towards headers won, rather than interceptions/tackles.

I consider the same for wingers and aimless dribbles they make, and strikers with aimless headers and flick-ons. But that's for another discussion.

I also want to praise SI for making the huge improvements in the ratings system from FM09.

[Rupal]

I must apologise for re-entering this debate but I just want to make a point to SFraser over some of his contributions. I'm not trying to fan flames or cause trouble - exactly the reverse in fact.

Nothing is to be gained by denigrating other posters or posting in an aggressive manner, SFraser. Such remarks as 'a bit of a brick wall', misunderstanding in a 'ludicrous fashion', wanting the game 'tweaked to make up for having absolutely no clue about football', etc, do not help to convince anyone of the correctness of what you are saying. All that they do is to antagonise people, which achieves nothing. It gets in the way of rational discussion. It's not always the case that the person who shouts loudest is the most listened to.

Forum exchanges should be civilised, constructive debates. Even with strongly held opposing points of view being manifest there is no need for breaches of manners or unpleasantness on either side.

You have a great deal of knowledge of FM for which (you will probably be surprised to learn) I (amongst many others) have considerable respect. I don't want to quarrel with you because life's too short for that.

All that's required is a little tolerance, really. Couldn't it be that other people might be able to disagree with you without being excessively obtuse or wilfully confrontational? If you were a little less impatient with them it would do wonders.

Also I believe that deejay 10 has a point. Managing Man U and managing Hyde Utd are very different. Some of the solutions which are available to you as manager of the former simply aren't there for me managing the latter. I don't have the luxury of deciding which DC I play against which opponent - I'm lucky if the only DCs in my squad who aren't totally useless are both fit to play in my next match. So it's not much help if somebody says to me that I'm a fool for playing a DC with a jumping stat of only 12 and that my tactics must be utterly hopeless because what other choice do I have? I do the best I can with what I've got.

Please feel free to PM me about any of this. I'm anxious to mend fences here.

[pauliowigs]

Supporting a Centreback that is weak in the air with someone strong in the air is just basic common sense, irrespective of any potential ratings bias.

Points like these only reinforce the perception that you want the game tweaked to make up for having absolutely no clue about football.

oi oi now. Put that finger away and stop pointing it. Accusations of "having absolutely no clue about football" are obviously going to cause offence on a thread like this. I've not attacked you personally anywhere. You obviously know your stuff to some degree, but you yourself are claiming be open to the idea that the % of headers won could be an overweighted factor. That, in essence, is all I am really saying. I think most people here who are claiming a problem with "short" centrebacks can more than buy this as a likely explanation.

I don't disagree with what you are saying here regarding having a defender who is strong in the air support one who isn't. I've never suggested that two centreback midgets should get high average ratings as a back pair. OF course it makes sense to back up a centre back who is not a dominant force in the air with one who is! I'd previously gave the example of Aaron Hughes and Brede Hangeland in both FM and real life terms.

All I was getting at was that, at least in my opinion, you should be able to play two centrebacks, 1 who is short (well, actually whose stats suggest he may not be dominant in the air), in a simple symetrical 4-4-2 without having to do much more than possibly setting the short one to cover (and/or maybe the taller one to stopper). A series of further tweaks, or the neccasatie to play an arielly dominant strikeforce or a lopsided formation should not be neccasary.

Do not mistake this as criticism for your tactical advice. I do not doubt that your suggestions can maximise the player in questions rating and performance, and that your advice is sound and tactic is good. Even if player ratings and the weighting of % of headers won are fine as they are or are corrected in future then I am sure your advice would still help the low jumping centrebacks in question perform even better. But I am also sure you could offer similar advice for any sort of player in any area of the pitch, to try and help someone raise a player rating or the performance of a player.

Anyway, with the progression of this argument, hopefully now we can get some feedback that everyone can agree on as to whether the % of headers won has too much influence on a players match rating, or at least a centrebacks match rating in relation to other things such as tackles made and interceptions and suchlike.

[outtasync]

Anyway, with the progression of this argument, hopefully now we can get some feedback that everyone can agree on as to whether the % of headers won has too much influence on a players match rating, or at least a centrebacks match rating in relation to other things such as tackles made and interceptions and suchlike.

I've been having serious problems with getting Otamendi to perform as well as his value suggests he should.

For the past 5 seasons, I've been using a centre back pairing of Otamendi and Danny Wilson. Here are their respective season ratings from season 1 to 5:

Danny Wilson (worth £18.75m, jumping 17, heading 16): 7.31, 7.27, 7.41, 7.46, 7.23

Nicolas Otamendi (worth £31.5m, jumping 13, heading 15): 7.01, 6.86, 6.90, 7.05, 6.99

I've had both of them on identical settings (Central Defending/Defend) for the first 3 seasons, subsequently I've switched Otamendi to (Central Defender/Cover) under some advice from the Otamendi thread.

I cannot stand the fact that Wilson can outperform Otamendi every season since season 1 when he was only 17 years old. I also cannot stand the fact that even though Otamendi makes more key tackles and interceptions, he gets crap ratings just because he loses more headers.

I know some of you might say it's my tactics, so can anyone point me in the right direction before I ship him off to Real Madrid for 60m?

[SCIAG]

If Otamendi makes more key tackles and interceptions, then don't worry. If Wilson wins enough of your headers, he'll make up for the deficiencies in Otamendi's game. Otamendi is clearly performing well, even if his average rating doesn't reflect that. There's no problem

[outtasync]

Yes technically there shouldn't be a problem, but because the match engine fails to recognize his contributions, that is the problem in itself?

And his low ratings means that he will never be in contention for any awards, in fact Danny Wilson's "Awards" page is 3 times as long as Otamendi's.

[SCIAG]

It's a problem that he doesn't receive credit, but that doesn't mean you should sell him.

[outtasync]

Haha yes but I figured if I could find someone else who can do what he's been doing AND still be able to win headers, I could really do with an extra 60m.

[Draakon]

As initial question was about rating I just assume that ME don't give too generous ratings for CB who loses most of his headers.

As far as I understand most important stats when calculating ratings are:

1) Goals - mind that if you have one CB who you set to attack corners and he occasionally scores from those he's average rating is naturally much higher than his partner's)

2) Assists

3) KEY passes, KEY tackles, KEY headers. He can win just 4 headers out of 8 but if they're all key headers his rating will be still OK)

4) Runs, interventions (will matter if number get somewhere 10+)

In the same time of course negatively affecting stats are:

1) Missed clear cut chances

2) Lost key tackles/headers etc. (you can't see this number on your team stats though, look for opponent stats too)

Last, but not least - if team is winning or losing affects ratings of all players.

[Frame]

Cannavaro, Puyol are certainly not 190cm tall. They are not even 180cm tall.

Puyol is 1,80 if you include his hair

[r0x0r]

My centerback pairing at a top premiership / champions league side (so up against the best regularly) are 6'1 and 5'10, with jumping of 15 and 13 respectively. The latter one has 19 strength and 18 aggression, so despite being shorter than his partner, is the man I stick on any tall strikers I happen to come up against. The other has 17 strength, so still does okay, but doesn't have the aggression to really make it count.

Half way through the second season (their first as a partnership, one was my DM in the first year, the other my substitute defender), they have only conceded 8 in 20 games (by far the lowest in the league), and have 7.14 and 7.10 ratings respectively. The players either side of them are attacking fullbacks with 11 and 13 jumping, so while they can hold their own they are no giants. Our new DM is 6'2 with 15 jumping, so he also helps out a little, but is still "relatively poor" in the air.

Relatively speaking, their ratings are better than my midfielders, but worse than the strikers and wide players. So pretty much dead in the middle of the team's ratings range. Talent-wise? They are probably what I'd consider the weak link in the team, the midfielders they out-play are far better players.

So, lack of height clearly isn't holding them back.

[Draakon]

If I remember right, FM ME doesn't consider player's height as an attribute. Height already accounts to jumping rating and that's it. It doesn't matter how tall you are, jumping attributes includes your height, jumping ability and willingness to attack headers.

So there could be some players who are relatively short (especially for CB) but have good jumping because they're able to jump high and therefore compensates his shortness. Although there are very few (if any) tall players who's jumping is low because IRL tall players mostly try to capitalize on their height and you rarely see any tall CB who just standing with 2 feet on the ground.

[outtasync]

If I remember right, FM ME doesn't consider player's height as an attribute. Height already accounts to jumping rating and that's it. It doesn't matter how tall you are, jumping attributes includes your height, jumping ability and willingness to attack headers.

SI have said that height affects how high a player treats the ball as "in the air", after which the jumping stat takes over, so height is still very relevant here.