Current Ability and Atrributes Research

[Leroy1883]

Originally posted by Mitja:

a) becouse if player A has passing 20 and weak (5) weaker foot, you would expect him to be able to pass with his natural foot 20, not 18. but he won't be able to use much of his weaker foot. it's common logic to me. it doesn't make guys like messi worse passers in terms of passing quality, when using natural foot.

b) which is even more important

feet theory shouldn't affect someone's mental or physical stats, right? I meen I met yet to see some player being braver or more agresive or better at jumping becouse he can use his weaker foot better.

If a right footed player has 20 in passing and passes the ball with his rightfoot(20) then he will suffer no penality in the match engine and still be a world class passer of the ball.

If he is forced to pass with his weaker foot then he will suffer a penalty. The player knows this and will try to use his stronger foot in situation, but opposition instruction such as show on weaker foot will make his life harder.

It just the same with finishing and technique e.g if you have a 20 finisher and a 5 technique he will not be as effective as a player with 15 technique.

Just think of the footedness as an attribute just like the rest, except that in some situations it will be 20 and in others 5.

[Leroy1883]

Okay here a little idea for you.

In game, your players may have an unassigned CA points allocation.

It is this pool that allows training to assign this points allocation to your attributes.

Once you have run out of unassigned points allocation you will not see training have an effect in terms of attribute improvement.

You can reclaim CA point allocation to a certain extent by not training a player in an area you don't want.

The relationship between CA and PA is different, it manifests itself naturally overtime and the rate can be increased through personally attributes, good coaches, good facilities, overall training workload and competitive match practice.

As CA increases, so does the unallocated points total, which is then fed into the training cycle.

I haven't got time to do a test now, but I might have time tonight.

The best time to perform a CA check for unallocated points is after a player returns from holiday or after a long injury layoff.

If anyone else wants to try it feel free.

I would recommend using FMM to do the analysis

[Hawshiels]

Originally posted by isuckatfm:

@Hawshiels

How long have you been looking at FM in this depth, and why haven't you posted on it before? Have I just missed your threads or have you been hanging onto these little nuggets to give you the edge in online FM network play gambling world

I've been looking at it for a few years now, but it changes quite often so it is sometimes hard to work out quickly what has changed after a new release.

I've never posted any of it before, mainly because:

1. There were a few issues in previous versions that I noticed and if I had posted them, it would only have caused un-necessary SI bashing. What we didn't know didn't hurt us and all that.

2. I didn't realise how interested people were in this sort of thing - mainly because I just use the info for buying players and for creating training schedules. I would occassionally post on the Good Player forum, but I would get responses like the one I got from my son ("He's rubbish") because of the low CA so I stopped doing it.

3. I didn't see that me having this info would be any real advantage or disadvantage to others, but when Law_Man started this thread I thought it might be good for everyone to be on a level playing field for the start of FM Live.

Glad everyone's enjoying it.

[Hawshiels]

Originally posted by Mitja:

I don't know for you but in my game I have a comlete mess with this issue. I'm realy ****ed. good player's stats go down bad player' go up, you realy can't judge a player by looking at him. for me this no realistic representation of anything. comlete mess!!

sorry for the rant.

Mitja: Please, please, please .. no more rants on this thread. You can make your point well without this.

And to clarify a couple of your points:

1. You can judge a player EXACTLY by looking at his attributes. They are still all comparable.

2. 'Stats' don't go down, but the value of certain 'attributes' will to maintain a realistic level of 'stats' within the game. If you don't understand this, I don't know how many more times I can explain it before bursting.

3. If you disagree with the attributes of Elano or anyone else, change them yourself. SI believe that they are representative, but they give you the editor to change them if you don't. But remember that if you (like the researchers) give him too many or too little attribute points based on his CA, the game will amend this for you.

p.s. I'm now asking you very nicely. Please.

[Hawshiels]

Originally posted by LSS:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Hawshiels:

If you tell me the original position of a single player and his new position, I can tell you how to work out exactly what he has lost in points. But unless you know what his potential actually is, you won't know whether it was just a small mistake ... or a big bad one!

If I stop the training and as the result the player loses the learned position, will he then get back the lost CA points? </div></BLOCKQUOTE>

LSS. Regarding your earlier point about these training regimes being 'cheat' regimes. I know what you mean about this, but here is how I think about it to give me a clear conscience. The manual for FM in my opinion is terrible. In fact, the best manual for the game is on these forums. In the past, I feel that I could have asked SI for many of the answers I have had to discover for myself, but over the past 3 years or so, the forums have deteriorated so SI have less involvement (which I can understand - but I still don't like it). I hope this changes soon, but in the meantime I'd rather get on with doing something positive myself than waiting to get answers from them. So, all I think I am doing is going into a little more detail than the manual does about how the game works. There you go. Clear consciences all round.

And now to your point above. If you stop training and the player 'loses' his knowledge of this position to below a score of 11 out of 20 (i.e. 1-10), you will indeed get the 'free' attributes for the retained position.

So, for example, if you have a MC/AMC (20/20) that you have trained to FC (14/20). As soon as the FC score reached 11, you would have lost the 'free' attributes for a midfielder (see much earlier in the thread for a list of these), so you would have to use some CA points to cover the totals within these. Note that any common 'free' attributes between MC/AMC and FC would be retained as 'free' attributes and would not cost anything.

When, the FC position is no longer trained and the player then drops from 11 out of 20 to 10 out of 20, the midfielder 'free' attributes would then all be returned to the player.

[Hawshiels]

Originally posted by Leroy1883:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mitja:

a) becouse if player A has passing 20 and weak (5) weaker foot, you would expect him to be able to pass with his natural foot 20, not 18. but he won't be able to use much of his weaker foot. it's common logic to me. it doesn't make guys like messi worse passers in terms of passing quality, when using natural foot.

b) which is even more important

feet theory shouldn't affect someone's mental or physical stats, right? I meen I met yet to see some player being braver or more agresive or better at jumping becouse he can use his weaker foot better.

If a right footed player has 20 in passing and passes the ball with his rightfoot(20) then he will suffer no penality in the match engine and still be a world class passer of the ball.

If he is forced to pass with his weaker foot then he will suffer a penalty. The player knows this and will try to use his stronger foot in situation, but opposition instruction such as show on weaker foot will make his life harder.

It just the same with finishing and technique e.g if you have a 20 finisher and a 5 technique he will not be as effective as a player with 15 technique.

Just think of the footedness as an attribute just like the rest, except that in some situations it will be 20 and in others 5. </div></BLOCKQUOTE>

Cheers Leroy for helping here. I'm glad most people are understanding what is being written. Sometimes I'm not sure if Mitja writes some things because he really doesn't understand, or if he is just mischief-making. I realise that English is not his first language though, so maybe he really doesn't understand what is being written.

[Hawshiels]

Originally posted by Leroy1883:

Okay here a little idea for you.

In game, your players may have an unassigned CA points allocation.

It is this pool that allows training to assign this points allocation to your attributes.

Once you have run out of unassigned points allocation you will not see training have an effect in terms of attribute improvement.

You can reclaim CA point allocation to a certain extent by not training a player in an area you don't want.

The relationship between CA and PA is different, it manifests itself naturally overtime and the rate can be increased through personally attributes, good coaches, good facilities, overall training workload and competitive match practice.

As CA increases, so does the unallocated points total, which is then fed into the training cycle.

I haven't got time to do a test now, but I might have time tonight.

The best time to perform a CA check for unallocated points is after a player returns from holiday or after a long injury layoff.

If anyone else wants to try it feel free.

I would recommend using FMM to do the analysis

The bolded sentence is true as far as 'chargeable' attributes, but you WOULD still see an improvement in the 'free' ones because they do not take up any of the CA points.

With regards to the fluiditiy of the CA and PA: The PA is a ceiling of points ('chargeable' attributes) for that player and it remains constant throughout. However, if you looked at the CA over the period of a career, you would indeed notice fluctuations. But I believe this is why the CA is not shown within the game. It would be very misleading to see the CA of a player go up and down for no 'apparent' reason. We now know the reasons, but I don't think it would help any casual gamer (or indeed the majority of them) to see a rating like this in the game.

[isuckatfm]

I would add to the argument about the realism of applying this to training by saying that in real life as a manager you would see a player day in day out. We would know our players, what thay are capable of and have an idea of whether or not training can improve them. As it stands in the FM world it is still very much guesswork as we don't fully understand the links between CA, attributes and training (until this thread shed a bit of light ).

As an example, in a previous save game I had a striker with great attributes to perform as the AM R in a 433 so I decided to retrain him. Yet after a few months game time despite the relevant training levels being at high levels, his attributes started dropping and I had no clue why. But now thanks to Hawshiels experiments and postings I have a better understanding of the actual cost of retraining a player.

[Cleon]

Originally posted by Mitja:

I don't know for you but in my game I have a comlete mess with this issue. I'm realy ****ed. good player's stats go down bad player' go up, you realy can't judge a player by looking at him. for me this no realistic representation of anything. comlete mess!!

sorry for the rant.

Please stop spoiling the thread, go and rant elswhere please. I won't ask you again

[Mitja]

Originally posted by Hawshiels:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Leroy1883:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mitja:

a) becouse if player A has passing 20 and weak (5) weaker foot, you would expect him to be able to pass with his natural foot 20, not 18. but he won't be able to use much of his weaker foot. it's common logic to me. it doesn't make guys like messi worse passers in terms of passing quality, when using natural foot.

b) which is even more important

feet theory shouldn't affect someone's mental or physical stats, right? I meen I met yet to see some player being braver or more agresive or better at jumping becouse he can use his weaker foot better.

If a right footed player has 20 in passing and passes the ball with his rightfoot(20) then he will suffer no penality in the match engine and still be a world class passer of the ball.

If he is forced to pass with his weaker foot then he will suffer a penalty. The player knows this and will try to use his stronger foot in situation, but opposition instruction such as show on weaker foot will make his life harder.

It just the same with finishing and technique e.g if you have a 20 finisher and a 5 technique he will not be as effective as a player with 15 technique.

Just think of the footedness as an attribute just like the rest, except that in some situations it will be 20 and in others 5. </div></BLOCKQUOTE>

Cheers Leroy for helping here. I'm glad most people are understanding what is being written. Sometimes I'm not sure if Mitja writes some things because he really doesn't understand, or if he is just mischief-making. I realise that English is not his first language though, so maybe he really doesn't understand what is being written. </div></BLOCKQUOTE>

hawshiels, leroy said exactly the same thing what I said...we are talking about 2 different things realy. your findings are amezing and I wonder if even guys from SI know so much about this

all I want to say is that

a) ME changes stats, adds up, takes off....all becouse of CA/PA system->

b) I believe that researchers stats should be those that are relevant, and same in game as they are in editor. researchers judge players by their attributes not by CA. this is a database issue and Paul said there is a some bug in this in-put of editor db to the game.

I won't disturb you anymore. I just hoped you could help me understand why this is happening. and you did partly. I will pay attention to this thread's findings. KUTGW

PS you think I don't understand things well and that I'm blind. well I feel the same for you . the things is you're exploring this aspect of the game. and I'm just moaning about it. it does make you better then me though

[Law_Man]

errr Mitja.... I personally wouldn't be at all suprised if the guys at SI knew as much about this as....well....they created it....!

I asked you before not to use this thread to rant, because you can do that (as you have done) on the GQ, and the last thing I think any of use wants in this thread is a repeat/re-hash of the points you make in the GQ forum. So although I can't stop you, I would ask that you didn't if it all possible

[Hawshiels]

[Law_Man: How's the legal eagle? Exams? Relaxed?]

Just to let you know that I have been playing the game all morning (I love working from home) and I've set up a couple of tests that I'll explain. The findings are just as expected in some regards, but by presenting them I think I'll be able to demonstrate why the game MUST make these changes based on feet, and also how we can learn from this to make the most of our players.

The first game I am playing:

1. I have selected a team to manage. But in it I have 6 created players. 2 x DCs, 2 x MCs, 1 x AML, 1 x AMR. The other players are 'largely' irrelevant at the moment. This is a 4-5-1 formation.

2. DC-1 is the left central defender and DC-2 is the right central defender. DC-1 is left-footed only with the highest available average stats for all 'chargeable' attributes (i.e. 18 and 19 scores). DC-2 is either-footed with the highest average available for each 'chargeable' attribute being (15). The same 'feet' test applies to the MC(left side is left-only), MC (right side is either-footed), AML (left foot only), AMR (either-footed).

3. No players have PPMs and all have the same mental stats and maxed out 'free' attributes. They are all the same age, height, weight, etc.

Now, what I aim to demonstrate with this test is this. I would expect the two-footed players to attempt more passes/crosses in a game, but I would expect the % of completed passes/crosses to be higher for the single-footed players since their passing ability is higher for their best foot. This seems to make sense and I hope to demonstrate this in the next post.

Following on from this, I then am going to run the same season, but this time max out the relevant attributes for each position (so, two-footed and single footed players have the same passing/tackling/crossing/dribbling scores. What I would then expect to see is that the two-footed players will still attempt more passes, but that there should be no real difference between the accuracy. This would mean that a two-footed player with 20 for passing (especially in midfield) is X% more effective than a one-footed player with 20 for passing. I will also check the effects on tackling - and eventually closing down (and showing onto weaker foot).

[Serdar]

Originally posted by Hawshiels:

[Law_Man: How's the legal eagle? Exams? Relaxed?]

Just to let you know that I have been playing the game all morning (I love working from home) and I've set up a couple of tests that I'll explain. The findings are just as expected in some regards, but by presenting them I think I'll be able to demonstrate why the game MUST make these changes based on feet, and also how we can learn from this to make the most of our players.

The first game I am playing:

1. I have selected a team to manage. But in it I have 6 created players. 2 x DCs, 2 x MCs, 1 x AML, 1 x AMR. The other players are 'largely' irrelevant at the moment. This is a 4-5-1 formation.

2. DC-1 is the left central defender and DC-2 is the right central defender. DC-1 is left-footed only with the highest available average stats for all 'chargeable' attributes (i.e. 18 and 19 scores). DC-2 is either-footed with the highest average available for each 'chargeable' attribute being (15). The same 'feet' test applies to the MC(left side is left-only), MC (right side is either-footed), AML (left foot only), AMR (either-footed).

3. No players have PPMs and all have the same mental stats and maxed out 'free' attributes. They are all the same age, height, weight, etc.

Now, what I aim to demonstrate with this test is this. I would expect the two-footed players to attempt more passes/crosses in a game, but I would expect the % of completed passes/crosses to be higher for the single-footed players since their passing ability is higher for their best foot. This seems to make sense and I hope to demonstrate this in the next post.

Following on from this, I then am going to run the same season, but this time max out the relevant attributes for each position (so, two-footed and single footed players have the same passing/tackling/crossing/dribbling scores. What I would then expect to see is that the two-footed players will still attempt more passes, but that there should be no real difference between the accuracy. This would mean that a two-footed player with 20 for passing (especially in midfield) is X% more effective than a one-footed player with 20 for passing. I will also check the effects on tackling - and eventually closing down (and showing onto weaker foot).

I'm really curious to hear results of this, but to be more accurate I strongly suggest you to run the same tests two or three times in a row before changing variables. Because as you know additional factors like morale and opposition also affects player stats. I think just one round for each test case may not be enough. Maybe 3 times half season rounds. Besides that it will be very interesting to know the conclusion.

[Joor]

Maybe I missed something, but do Attacking Midfielders Left and right get free crossing ?

[trebor1185]

From what Hawshiels has found and what i've seen myself AML/R don't get free crossing.

[Hawshiels]

Right then. I've got some really good stuff for you here, but first let me explain something that I'm not sure everyone realises beforehand. Most of you know this, but for those that don't.

It is important for me to explain the way the 'passing' attribute works within the game. A player with 10 passing and a player with 20 passing can both still make a perfect pass (or thereabouts) to a player standing 10 feet away. Equally, a player with passing 15 and a player with passing 20 can make an equally (or thereabouts) good pass to someone standing 50 feet away. But a player with passing 10 will never complete as many passes as a player with passing of 20 if they are attempting many longer passes. So, the passing attribute not only determines the accuracy of the pass, but it has a massive impact on the accuracy over a longer distance (i.e. 5 degrees of an error over 10 feet is nothing, but the same 5 degrees of error over 100 feet is alot). Make sense? Great. [This paragraph and its assumption also include assumtions that technique is the same/similar.]

Secondly, it is important to appreciate the importance within the engine of the 'decisions' attribute. The decisions attribute will help a player (especially given creative freedom) to decide how best to proceed - give any circumstance. "Should I cross it?", "Should I pass it back?", "Should I try my weaker foot?, "Should I shoot?". This is a key attribute so in all of my tests I ensure that this attribute is common.

Anyway, back to the findings.

The first season has been run 3 times to give me an average score (i.e. the same season was saved, run, stats noted, re-loaded, re-run, etc). Also, I have rationalised the numbers to take into account that one player may play less games than the others (there is not much of a difference here, but I did it anyway).

And the stats are like this. Lets just do passing first of all.

<pre class="ip-ubbcode-code-pre">

Defender 1 (Left only, Passing=19)

1671 Passes, 1428 Completed (85%)

Defender 2 (Either, Passing=14)

2084 Passes, 1685 Completed (80%)

Midfielder 1 (Left only, Passing =19)

3162 Passes, 2788 Completed (88%)

Midfielder 2 (Either, Passing=14)

2123 Passes, 1900 Completed (89%)

AML (left only, Passing=19)

1385 Passes, 1025 Completed (74%)

AMR (Either, Passing=14)

1299 Passes, 965 Completed (74%)

</pre>

Hopefully you'll find this interesting in itself, but I can use it now to explain WHY these numbers are the way they are. You need to know this before I post the next set of results which are wildly different for good reason.

Firstly, you will notice that two-footed defenders appear to be able to increase the volume of passes in a game (between 10% and 20%) and yet only suffer a 5% reduction in success rate. You will see in the following posts however, that I have set the passing range for the defenders to match the lowest passing rate. In other words, I have set both DCs to passing that is restricted to a distance that both can pass acccurately in. When I set the passing distance to be longer, the error rate for a defender where Passing=14 becomes much worse than a defender where Passing=18. This just makes sense.

For the midfielders, I need to tell you a bit more about the tactics I used to test this. I have my midfielders set to normal passing (i.e. in the middle of the slider). In addition, they have slightly less than normal creative freedom. What does this do? Well, it does two things. Firstly, it ensures that my midfielders use their decision making skills (which are set at nearly maximum, 19) more meaning they will try passes only if it 'makes sense' after going through their 'decision engine'. So, what we see in our results is that Midfielder 1 (left foot only) tried more passes than player 2 (either foot) but both players has a similar % success rate. Why is this? It's actually quite sensible and proves that the decision-making within the game is very important. What this shows is that:

- When Player 1 goes through the decision making process, there are more times when it makes sense for him to try a longer pass (because he will be more likely to pass over distance). So, he will attempt more passes.

- When Player 2 goes through the decision making process, there are fewer times when it makes sense for him to try the longer distance passes due to his reduced passing ability so he will choose to dribble more or run with the ball more until his decision making process tells him to pass/shoot/cross, etc.

However, in both instances, since the decision making skill is the same, you would expect their completion percentage to be the same. Note that the difference between midfielders and defenders in this example is due to the decision making. Defender 2 has to make more passes, because it makes more sense to try a longer pass (even at the risk of losing it), than trying to make forward runs or trying to dribble.

Now if we look at the AML/AMR players in this example. This is again completely consistent with the conclusions I made with the central midfielders. For the AML (left only), his decision making will allow him to pass more because he is likely to succeed more. AMR (either) is less likely to 'decide' to make as many passes because his passing is not as good so he will choose to run/dribble with the ball more.

Making sense so far?

[kolobok]

Maybe I missed something, but do Attacking Midfielders Left and right get free crossing ?

No, AMCs get free crossing.

[kolobok]

Making sense so far?

Perfect.

A few additional stats I would be interested though, if you don't mind. Assists, average ratings, value increase. Could you please post them even without comments.

Cheers.

[Grokaer]

Originally posted by Hawshiels: Making sense so far?

Yes, it made perfect sense when you explained it, although when I saw the figures initially before the explanation I was surprised about the 2nd midfielder's lower number of passes - then it became clear when you explained the importance of decisions, creative freedom and passing.

The more you have uncovered the more I am appreciating your work and also the work of SI in creating this game [well they get bashed too much in my opinion - sorry if that statement is slightly off topic].

[Cleon]

I know I've said it once already, but really well done on the thought and analysis put into this thread. A lot of work and effort as gone into this and I thought you deserved special praise again, so well done lads, excellent job

[Hawshiels]

Originally posted by kolobok:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Making sense so far?

Perfect.

A few additional stats I would be interested though, if you don't mind. Assists, average ratings, value increase. Could you please post them even without comments.

Cheers. </div></BLOCKQUOTE>

I was going to do this, but I couldn't work out why AML had so many more assists than AMR. Then I realised that it was due to him taking all the corner kicks. I will have to re-run these tests to make it more readable and usable. As soon as I've done that (I've got it running at the moment), I'll post it up. In the current version, I have been careful to make set pieces taken by the middle MC since he takes no part really in this study.

[Hawshiels]

Originally posted by Cleon:

I know I've said it once already, but really well done on the thought and analysis put into this thread. A lot of work and effort as gone into this and I thought you deserved special praise again, so well done lads, excellent job

Cheers Cleon Praise from yourself and Paul C on the same thread ... I'm sure I speak for the other guys also when I say we're greatly appreciative of your support.

[Grokaer]

Hawshiels - Have you noticed any changes between the original 8.0.2 patch and the updated version that was available last night yet with respect to your work [and with respect to everyone else's work on this thread] so far or is it too early to say?

I am guessing from the description on that post that your work will be largely if not completely unaffected...

[isuckatfm]

Great stuff again Hawshiels

You've confirmed something for me that I always assumed about how the match engine works, but couldn't be arsed testing as my crappy laptop would take forever.

[Hawshiels]

Originally posted by Grokaer:

Hawshiels - Have you noticed any changes between the original 8.0.2 patch and the updated version that was available last night yet with respect to your work [and with respect to everyone else's work on this thread] so far or is it too early to say?

I am guessing from the description on that post that your work will be largely if not completely unaffected...

I am doing all of the tests on 8.0.2. I didn't see any reason to upgrade for two reasons:

- Apparently the only fix is a data one for the Defoe issue (which I fixed using the editor anyway), and the Turkish finances (which again I used the editor to fix)

- I didn't want to invalidate (or give anyone any reason to invalidate) the results by changing versions. When I've run these tests, I am however going to download the update and try again just to be sure though. In the meantime, please note that until further notice, all results and findings are based on 8.0.2 - The Valentine's Day release.

[Hawshiels]

Now we'll look at something I think you'll really like - and in itself may just make reading this thread worthwhile.

TACKLING

This is an area within the game that has been vastly improved by SI - both in terms of its effectiveness and realism (especially in 8.0.2).

And with tackling, the decision process has much less of an effect on a player. It makes sense. When attacking, you have so many options to weigh up. Do you pass, shoot, cross, dribble, run with the ball, etc. When tackling, the player has only to decide whether to tackle or not And if not, it is down to the other tactics to tell him what to do (track back, close down, tight marking, etc).

So, now lets look at the results. These results come from the 3 seasons (i.e. the same season re-run 3 times) I used above for the passing comparison.

Firstly, the actual number of tackles is largely irrelevant because in this example, I was using top quality players all over the pitch (in the SPL) so the requirement to tackle is less than if I had a poor team that was always under pressure. The results would still be the same percentage wise, but the actual numbers wouldn't. And anyway, it is the percentages (in relative terms) that hold the interesting info here.

The statistics show that for each game (individual match), the one-footed players made the same (or similar) number of tackles as the player with two feet.

However, here is the good bit, despite the two-footed player having only a score of 14-15 for tackling (in comparison with the 18-19 score of the one-footed player), the either-footed player had a better tackling success rate. But if you consider this for a moment, this makes perfect sense. If the left footed player is having to make many tackling attempts with his right foot, you would expect these tackles to be poor. So, consider this:

For a LEFT-FOOTED central defender or midfielder, is it more likely that they will tackle with their right or left foot? I would suggest that the 'appropriate' foot to tackle with would be 50/50 (since it is in the middle of the park), but of course this same defender will favour tackling with his left, so the majority will be left footed tackles. So, this same defender will be mostly successful when tackling with his left foot (tackling score of 18-19), but will be mostly unsuccessful with his right (tackling of ?low score?).

Equally, a player that can use either foot with tackling of 14-15, will be in exactly the same positions. But rather than favouring one foot over the other, he will be able to tackle very successfully with either.

So, for every tackle the left-footed player makes with his left (tackling=19), he will be a little more sucessful than the either-footed player will be with his left (tackling=15). Equally though, the left footed player will be far less successful with his right-footed tackles in comparison with the right-footed tackles of the either-footed player.

Making sense again I hope. So, now for the numbers. Although they make the same number of tackles in a game, the either-footed player is likely to have a 10% higher success rate. So, if you looked at the stats screen for these two players, this is what I am looking at:

DC (left only) successful tackles % = 79%

DC (either) successful tackles % = 90%

The MCs are similar:

MC (left only) successful tackles % = 72%

MC (either) successful tackles % = 79%

The AMR/AML players however are different. What I found consistently was that the number of tackles were the same (similar) and their % of successful tackles were the same (similar). Here's why ......

When you have a player playing wide, on the side of the park his foot favours (i.e. left side for left footed player and vice versa), his tackling opportunities on his 'weaker' foot are going to be far fewer than if he was playing in the middle of the park. He will still have to make a certain percentage of his overall tackles on his weaker foot, but this percentage will be far less than a player playing the DC or MC role.

So, what I find is happening here with the AML and AMR players is this. Both players have similar numbers of tackles in the game. But since most of the tackles by the AML (left only) player will be on his left foot, his success rate is not going to be deteriorated as much as it would if he were a central defender and having to make more tackles with the weaker foot. So, the one-footed, 'great' tackling AML will be equal to the either-footed, 'good' tackling AMR overall.

I hope you will see how different tackling works compared with passing. When passing, the player can favour his strong foot. In tackling, the player will sometimes have to tackle with his weaker foot so it can make a massive difference to the effectiveness of central players - especially defenders where the requirement to make left and right footed tackles is important.

[De Siem]

Originally posted by Hawshiels:

I completely understand your point here and I realise it is perhaps a little frustrating.

However, we want and expect the game to be as realistic as possible and what they have done is made it just that by assessing the attributes in this way. It's just that we (as gamers) have not taken the 'foot' scores as seriously as we should have. You could argue though that SI could have been clearer about the importance of the 'weaker' foot though.

Perhaps a way of representing the attributes to make it easier for the user would be to have a passing, dribbling, etc rating for each foot. So, for Elano you would have passing scores of 19 and 3, whereas someone like Rosina would have scores of 19 and 15 (to reflect his two-footedness). There are maybe people that would argue that this would confuse gamers more.

Either way, WE know now (or at least the people reading this thread do) so lets get back to the game and enjoy it even more now.

Nice finding about the foot versus attributes thing, and thinking about it's so bleeding obvious isn't it? At least to those who play football themselves. I can give a cross, shoot or pass with my right foot fairly well, but trying the same cross, shot or pass with my left foot and the results are much worse!

[Hawshiels]

If we run the same season but this time comparing just two central defenders.

DC-1: Left-footed only. Tackling=20

DC-2: Either-footed. Tackling=20

All other attributes are the same (positioning, decisions, physical, etc).

Here's something you'll like.

DC-2 will have a 14% (on average) higher percentage of success in tackles than DC-1.

So, DC-1 is at 80% (a little higher than the previous example because I set the tackling to be 20 rather than 19), and DC-2 is at 94%.

Think you'll buy defenders the same way in the future?

I want/need to do more on this in this version, but I'm off to see what either-footed defenders I can get for my team 'Tenerife'.

[kolobok]

I was going to do this, but I couldn't work out why AML had so many more assists than AMR. Then I realised that it was due to him taking all the corner kicks. I will have to re-run these tests to make it more readable and usable. As soon as I've done that (I've got it running at the moment), I'll post it up. In the current version, I have been careful to make set pieces taken by the middle MC since he takes no part really in this study.

You don't need to re-run them, just subtract team's goals from corners from AML assists. It mignt be abiased a bit, but no to the extent when it matters. After all, if the difference is 1-3 assists, you can say it's just luck or... Do you have "either" foot ST? Because it also matters - if he is "right only" it's easier for him to pick cross from the left flank.

[Law_Man]

Originally posted by Hawshiels:

If we run the same season but this time comparing just two central defenders.

DC-1: Left-footed only. Tackling=20

DC-2: Either-footed. Tackling=20

All other attributes are the same (positioning, decisions, physical, etc).

Here's something you'll like.

DC-2 will have a 14% (on average) higher percentage of success in tackles than DC-1.

So, DC-1 is at 80% (a little higher than the previous example because I set the tackling to be 20 rather than 19), and DC-2 is at 94%.

Think you'll buy defenders the same way in the future?

I want/need to do more on this in this version, but I'm off to see what either-footed defenders I can get for my team 'Tenerife'.

This is absolutely fantastic Hawshiels, and with the passing too Since 07 I've always tried to buy central players with two feet (I do realise how potentially silly that sounds!)

Do you support Tenerife? At a guess I'd say there' likely to be more two footed DCs in Spain than England!

You know the test you've got to run now though don't you? Two footed v 1 footed strikers.... Now THAT would be interesting! Although I think creation of scoring opportunities is probably a lot more random than passing and tackles so you might want to try and holiday for a couple of seasons this time (depending on speed!) and then repeat, for better data. Even if I didn't have law exams I'm severely limited in tests like this because I've got a three and half year old laptop that only just manages to run FM and firefox!

Re law exams, just had my 9th hour of exams in four days, and got my last one tomorrow - a whopping 3 and a half hours of property law. Not fun.

[Serdar]

Originally posted by Law_Man:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Hawshiels:

If we run the same season but this time comparing just two central defenders.

DC-1: Left-footed only. Tackling=20

DC-2: Either-footed. Tackling=20

All other attributes are the same (positioning, decisions, physical, etc).

Here's something you'll like.

DC-2 will have a 14% (on average) higher percentage of success in tackles than DC-1.

So, DC-1 is at 80% (a little higher than the previous example because I set the tackling to be 20 rather than 19), and DC-2 is at 94%.

Think you'll buy defenders the same way in the future?

I want/need to do more on this in this version, but I'm off to see what either-footed defenders I can get for my team 'Tenerife'.

This is absolutely fantastic Hawshiels, and with the passing too Since 07 I've always tried to buy central players with two feet (I do realise how potentially silly that sounds!)

Do you support Tenerife? At a guess I'd say there' likely to be more two footed DCs in Spain than England!

You know the test you've got to run now though don't you? Two footed v 1 footed strikers.... Now THAT would be interesting! Although I think creation of scoring opportunities is probably a lot more random than passing and tackles so you might want to try and holiday for a couple of seasons this time (depending on speed!) and then repeat, for better data. Even if I didn't have law exams I'm severely limited in tests like this because I've got a three and half year old laptop that only just manages to run FM and firefox!

Re law exams, just had my 9th hour of exams in four days, and got my last one tomorrow - a whopping 3 and a half hours of property law. Not fun. </div></BLOCKQUOTE>

yeah it would be also very nice to see striker version of same test. To be honest it is what I'm most interested about. Maybe this can explain ridicilous misses coming from players with high finishing stats.

[Joor]

Originally posted by Hawshiels:

If we run the same season but this time comparing just two central defenders.

DC-1: Left-footed only. Tackling=20

DC-2: Either-footed. Tackling=20

All other attributes are the same (positioning, decisions, physical, etc).

Here's something you'll like.

DC-2 will have a 14% (on average) higher percentage of success in tackles than DC-1.

So, DC-1 is at 80% (a little higher than the previous example because I set the tackling to be 20 rather than 19), and DC-2 is at 94%.

Think you'll buy defenders the same way in the future?

I want/need to do more on this in this version, but I'm off to see what either-footed defenders I can get for my team 'Tenerife'.

Yep thats what I see ingame with my Milan team. Even tho Maldini got low psycical attributes hes still a great defender . 17 left foot 20 and right but "only" 15 tackling , he still make so meny key tackles each game

[Hawshiels]

Originally posted by Law_Man:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Hawshiels:

If we run the same season but this time comparing just two central defenders.

DC-1: Left-footed only. Tackling=20

DC-2: Either-footed. Tackling=20

All other attributes are the same (positioning, decisions, physical, etc).

Here's something you'll like.

DC-2 will have a 14% (on average) higher percentage of success in tackles than DC-1.

So, DC-1 is at 80% (a little higher than the previous example because I set the tackling to be 20 rather than 19), and DC-2 is at 94%.

Think you'll buy defenders the same way in the future?

I want/need to do more on this in this version, but I'm off to see what either-footed defenders I can get for my team 'Tenerife'.

This is absolutely fantastic Hawshiels, and with the passing too Since 07 I've always tried to buy central players with two feet (I do realise how potentially silly that sounds!)

Do you support Tenerife? At a guess I'd say there' likely to be more two footed DCs in Spain than England!

You know the test you've got to run now though don't you? Two footed v 1 footed strikers.... Now THAT would be interesting! Although I think creation of scoring opportunities is probably a lot more random than passing and tackles so you might want to try and holiday for a couple of seasons this time (depending on speed!) and then repeat, for better data. Even if I didn't have law exams I'm severely limited in tests like this because I've got a three and half year old laptop that only just manages to run FM and firefox!

Re law exams, just had my 9th hour of exams in four days, and got my last one tomorrow - a whopping 3 and a half hours of property law. Not fun. </div></BLOCKQUOTE>

I realised beforehand that either-footed (we should really stop calling them two-footed because you're right ... they all are), players would be better in the middle of the park. But I wasn't quite sure how much more effective they were. I'm delighted with this finding.

As for Tenerife, they are just one of a number of teams I follow. I like to watch alot of different teams from different countries, although Valencia is my favourite.

I have just run the test over again to identify the number of assists. You may have guessed the results in many ways.

I run the tests in a few different ways.

Test 1a

In the first of these new tests, I made the left-footed players 5 points higher in each of the attributes (passing and crossing) than the either-footed players.

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 14

The tactics I used had passing for all players set to the middle of the slider.

The left-only MC and AML contributed slightly fewer (around 8%) assists than the either-footed MC and AMR. So, this suggests that with a 'normal' passing length, the two-footed player has the edge.

Test 2a

I made the left-footed players 5 points higher in each of the attributes (passing and crossing) than the either-footed players.

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 14

The tactics I used had passing for all players set longer (the slider was set to the first notch of 'long').

The left-only MC and AML contributed slightly more (around 11%) assists than the either-footed MC and AMR. So, this suggests that with a 'longer' passing length, the two-footed player still needs the passing skills over the two-footedness.

Test 3a

I made the left-footed players 5 points higher in each of the attributes (passing and crossing) than the either-footed players.

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 14

The tactics I used had passing for all players set to the first notch of 'short' (coming from the normal position).

The left-only MC and AML contributed far fewer (around 17%) assists than the either-footed MC and AMR. So, this suggests that with a 'short' passing length, the two-footed player is more useful (even with poorer passing ability), than the single-footed player with better passing/crossing skills.

It is clear from these first 3 tests here that my tactics will have something to do with it. I have crossing/dribbling/running with the ball, etc all set to mixed to ensure that the decision-making of the player is used to determine how much of each to do. If a player had crossing set to often, it may be that the crossed assists would be higher and I would therefore expect the better skilled crosser (despite being only one-footed) to be more effective.

I run these same 3 tests again, but this time setting the passing/crossing skills to be exactly the same.

Test 1b

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 19

The tactics I used had passing for all players set to the middle of the slider.

The left-only MC and AML contributed far fewer (around 15%) assists than the either-footed MC and AMR.

Test 2b

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 19

The tactics I used had passing for all players set longer (the slider was set to the first notch of 'long').

The left-only MC and AML contributed far fewer again (around 15%) assists than the either-footed MC and AMR.

Test 3b

Passing/crossing for left-only was set to 19

Passing/crossing for either was set to 19

The tactics I used had passing for all players set to the first notch of 'short' (coming from the normal position).

The left-only MC and AML contributed far fewer (around 16%) assists than the either-footed MC and AMR.

We can see from this just how much more effective a stronger 'weaker' foot can be in the game. But this depends largely on the tactics involved. If you have a good either-footed player with a passing skill of 15, I would consider him to be as good as a one-footed player with a score of 20, PROVIDING we are playing a normal to short style of passing.

The striker position is something I already know the answer to from previous versions, but for the sake of consistency I'm going to check this properly in 8.0.2. If it is anything like previous versions, an either-footed striker is MUCH better - when other attributes are similar.

[Law_Man: I'm pleased that you are on your last exam. Let me help you out with this question if asked:

How should a lawyer charge a client for any work done?

A. A fixed price fee payable on completion

B. A fair rate based on the value to the client

C. By the hour (including lunch breaks), so take time over things and remember that if things are going well, create conflict with another lawyer just to draw things out a little longer and charge even more.

If asked, the answer they are looking for is 'C'. Good Luck!]

[Hawshiels]

For those people just tuning into this thread, let me give a quick re-cap of what we've discovered so far. This will help you to decide whether the rest of the thread beforehand is worth reading.

We now know exactly how the attributes relate to the CA (current ability) of each player.

We know that each position has a number of what we call 'free' attributes. We call them 'free' because they do not take up any of the points that count towards a player's CA and they can therefore be trained without taking up points that could be allocated to attributes elsewhere.

We know that training someone in certain other positions (in relation to the one they are currently natural in) causes them to lose these 'free' attributes. They only retain the 'free' attributes that are common to all of the positions they are comfortable playing in.

We know how and why the game increases/decreases certain attributes. The game increases and decreases attributes (that have been set using the database editor) in an attempt to stabilise the quality of a player based on his CA. If a player has not been allocated enough attribute points within the editor, the game will add more to the appropriate attributes. If a player has been allocated too many attribute points in the editor, the game will reduce these accordingly.

We know that the effect of this reduction/increase will be more severe in certain circumstances due to the skill (or not) a player has with his 'weaker' foot. Players that can use either feet within the game are considered more valuable in terms of their effectiveness within the match engine. In these circumstances, the game will normally re-align this player's attributes downwards - meaning that the attribute scores look less, but the effect within the engine is more balanced.

We now know how much more effective an either-footed player is in comparison with a single-footed player - with tests having been completed for central defenders, central midfielders and wingers. Striker tests are coming soon. We know the value of this effectiveness when considering their passing/crossing/tackling skills.

If you would like to read more of the detail behind this, you have the option of reading back through the rest of the thread, or waiting for a week or two until the thread can be consolidated into a 'report'.

Why do we continue with this thread, and what is its purpose?

1. To help understand the CA and attributes better with a view to building better training schedules.

2. To help identify what attributes (and other hidden gems) to look for in a player when buying one.

3. To help identify ways of ensuring that the players we have are trained to suit the positions and the tactics we deploy.

4. To help identify signs when a player is reaching his potential. This may be useful in determining when to buy/sell a player.

Remember that every part of the game is interlinked. A combination of what you learn here, what you learn from the various tactical discussions, and what you understand about the other aspects of the game will ultimately determine your success or not. This thread is just a cog in a very large piece of machinery. But an important cog nonetheless.

[Cleon]

If you wanted, you could create a new thread with your findings and put a link to this thread to show people the discussion before hand.

You might not want to do that, but I just thought id let you know its an option if you wish too

[Hawshiels]

Originally posted by Cleon:

If you wanted, you could create a new thread with your findings and put a link to this thread to show people the discussion before hand.

You might not want to do that, but I just thought id let you know its an option if you wish too

Good idea Cleon.

[Hawshiels]

Originally posted by Hawshiels:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Cleon:

If you wanted, you could create a new thread with your findings and put a link to this thread to show people the discussion before hand.

You might not want to do that, but I just thought id let you know its an option if you wish too

Good idea Cleon. </div></BLOCKQUOTE>

I've done this now. I made a link to this thread from the summary one, but the TEXT I used for it says "THIS THREAD ONLY". I'm just not sure of some people will know to click the link or if they'll just post in "THIS THREAD ONLY". Could you change the text to read "THE OTHER THREAD ONLY" or something you think will be clearer?

Cheers for your help.

[Cleon]

Now everything is clear, would you prefer to use this thread as reference only and the other to carry the debate onto? or?

[Joor]

How meny CA points did a player use to go from 1 to 20 in hes 2nd foot, was it 10 ?

I was just wondering whats the difference between having 11 in weak foot vs 14 .

I just know as soon someone got 16 or more in hes weak foot all technical attributes are lower when you begin a new game.

F.eks Maldini, who 20 in both foot in the database, when you start a game all hes technical attributes are down 2 pounts and some of hes mental too. Plus FM 08 lower one of hes feet to 17 !?

[Joor]

Maldini is a natural two footed player..so I think hes CA should have been 10 points higher 155-165 so he can keep hes abilites

[Law_Man]

Hawshiels: just eating a VERY big pizza and catching up with the brilliant thread and with all the hard work you've done!

With the test you did above with the left footed AML and MC, did you "control" all the other players too? I was just wondering about whether the FC(L) and FC® were of equal quality as that distort the figures a little.

Great work on starting a new thread that isn't to be used yet, and for producing a mini-summary KUTGW.

Your law chat made me chuckle. Better than that though, there have been a few cases where lawyers have been done for "over-charging their clients" - try more than 24 hrs in a day and more than 30/31 days in a month! (and that's not a joke).

If you want a juicy legal read (as opposed to boring corporate law): this important case from 1999 - though not safe work or your son really!

Cleon: thanks for being so helpful

[kolobok]

We can see from this just how much more effective a stronger 'weaker' foot can be in the game. But this depends largely on the tactics involved. If you have a good either-footed player with a passing skill of 15, I would consider him to be as good as a one-footed player with a score of 20, PROVIDING we are playing a normal to short style of passing.

Hawshiels,

Your findings are absolutely brilliant, but you are right - everything depends on the tactic. Passing slider position is not the unique thing that could affect outcome. E.g. if tempo is high either-footed player definitely looks better choice as he wouldn't need to spend time moving ball to the stronger foot. Slow tempo game would give one-foot player more time, hence you would expect more accurate passing from him.

What more important is that in your experiments you created players, identical in everything but attributes you have chosen. There are no such players in the game. So let's not give people wrong idea that there is any "attribute -foot usage" scale. Otherwise you will see a lot of complaints saying it does not work. There is a strong realtionship between attributes and foot usage ability, but the extent of it depends largely on the mix of players and their instructions. Even "first touch" attribute

Ok, a couple things I noticed when colecting data frome editor to do my part (it takes even longer than I expected as I have to add footage score, which is available on player by player basis only). There are some players with 0 score on one of the foot at least in 8.0.2 update (found in Arsenal squad) which is weird - I would expect an EPL level player to be able at least walk . Another thing that I noticed (though not sure as I only collected that attribute for 10-15 players) that the score appears at 5 points increment (i.e. 0-5-10-15-20). It probably tarnslates into "very weak - weak - ...- very strong" scale that we see in the game, but would be nice to confirm as I don't use editor or FMM, so know little to nothing about the scale. Finally, have you ever seen a player with score 10 for each foot? He would be "either-footed" I guess, but I wouldn't want such a player in my team. hence when you choose a player you would want at least one strong foot

[Law_Man]

In real life a player with a "10" for each foot wouldn't be classed as two footed, he'd be classed as equally average/bad or just as bad with his left as his right!

I'm not sure how FM would rank a player with 10 in each foot but hopefully Hawshiels will check that too as I've banned myself from FM until the weekend.

[isuckatfm]

I'm not sure how FM would rank a player with 10 in each foot but hopefully Hawshiels will check that too as I've banned myself from FM until the weekend.

Every player has to have 100/100 for one foot. As far as I remember the last time I messed with the editor it had a dialogue thing that stated that.

[Hawshiels]

Hi guys,

Firstly, there are no players in the game that don't have 20 for at least one foot. You make a good observation that many players are given the zero rating for their weaker foot. Under these circumstances (and I've still to properly quantify this), the game seems to apply a 'weaker' foot score. The game does not make up a weaker foot score if the database holds a real number from 1-20.

This also applies though to positions. An interesting thing I noticed and may have some impact on positions (and therefore losing free attributes) is how someone can get an 'automatic' position score depending on the positions they are good in. For example, if you have a DC and train him as a DL ... as soon as he reaches the 'green' stage for being a DL, he is automatically assigned a wingback position score. Equally, if he is trained as a DM, he will get a wingback score. So, I think it may be quite possible for a player to be trained as a DC and DMC when the game loads, but then he could be trained as a DL and the combination of the automatic wingback score from the DL could add to the automatic wingback score from the DM and cause him to go over the 11 mark without realising it. This is something that is very very complicated within the game at the moment and I'm trying to suss it out. But thought I'd give you some warning about this early.

Re, the weaker foot increments. The weaker foot scores in the database aren't really in increments so it may just have been the particular players you looked at. What interests me about this most though is the fact that they are not in these increments. I would have expected a researcher to have to choose from a list similar to the one you described (i.e. very weak, weak, strong, very strong). So either the researchers have the same simple 1-20 scale to choose from or the weaker foot score we see is based on the attributes and the simple score the researchers use. In other words, the database we see is an amended view of the data and not the one the researchers create in the first place. If the researchers do have the 1-20 scale though, I'm wondering if the researchers that use scores other than simple 1,6,11,16,20 scores understand the 'milestones' within the engine better and know the difference between a weaker foot score of 10 and 12 for example. I don't know for sure because I've not seen the software that the researchers use to collate their data.

Finally, it was a great idea of Cleon's to have a different thread we use for posting the findings - leaving this one free for the debate/discussion. The other thread will allow people to quickly work out if they are interested in this enough to read the detail in the meantime, or wait for the summaries and updates. The number of views and contributions to this thread is great considering the complexity of it. I thought it may be boring for most, but it appears to have created much interest. Nice idea Law_Man.

[Cleon]

I'll add the findings link to the first post of this thread too for you

[Andrewbanny]

Great thread guy's..lots of useful info in here.keep up the good work

[Law_Man]

Originally posted by Cleon:

I'll add the findings link to the first post of this thread too for you

What can I say? I'm not jus a pretty face

But in fairness its you that has provided most of these fantastic results (albeit with the thoughts of myself (and others) so you should take the credit for the time and effort you've put into this thread! Ok end of mutual back-slapping...

[Ellesthyan]

A huge round of applause for all you guys who contributed to this thread and especially to Hawshiels! You are a FM legend in the making: please keep up the excellent work!

[LSS]

From the results thread:

Originally posted by Hawshiels:

We know how and why the game increases/decreases certain attributes. The game increases and decreases attributes (that have been set using the database editor) in an attempt to stabilise the quality of a player based on his CA. If a player has not been allocated enough attribute points within the editor, the game will add more to the appropriate attributes. If a player has been allocated too many attribute points in the editor, the game will reduce these accordingly.

To me this sounds like a backward way of handling this attribute/CA thing. Isn't it easier for a researcher to assign specific attributes for a player than to evaluate how good he is in general (CA)?