Why do pass-related stats produce weird results and why is FM's game engine so weird?

[walker14]

hello. I play FM24 and did an experiment like this for a personal question. 

ㅁ Control variable 

- Tottenham vs Wolves. Wolves home. Both teams controlled by the user, no tactical or substitution changes. 

- Tactical skill maxed out. All morale perfect. Team Cohesion maxed. 

- All Stats 20 Assistant Coach Locker Room Conversations

- All positions except for the experimental group have a collective stat of 14

- Positive hidden to 20, Negative hidden to 1

- Minor injuries continue to progress. Reload for major injuries and ejections 

- Removed Player Traits

- Removed experimental players from dead ball kickers 

- All players except for experimental positions are two-footed

ㅁ Manipulation Variables

Multiple stats for each position

ㅁ Dependent Variable 

Team goals scored and goals against. XG. Goals, Assists, and Offense Points for Experimental Positions 

In these experiments, we took as few as 100 and as many as 300 stats in this environment, with results at the 0.01 to 0.05 level of error. Because the error in the experimental group is small, i assume it is controlled for variables.

It's no secret that very few of FM's stats have a significant impact on the game, including physicality. 

However, it's not just a matter of how much or how little it affects. Pass-related stats have some very strange effects The official FM website states that passing is affected by technique and vision. Crosses are listed in isolation. Let's talk about this in more detail.

Increasing the passing, technique, vision, and teamwork stats of the deep-lying forward from 14 to 17 (Table 1) actually decreases the scoring of the two inside forwards (Table 2), but increases their assists. 

In the inside forward attacking experiment, a player who is nearly perfect physically, mentally, and technically (Table 3) performs worse than a player with only physical ability and dribbling(Table 4) (Table 5).

The same was true in experiments with wingbacks. Fullbacks with passing-related stats (Table 6)(Table 7) performed worse than fullbacks with nerfed stats (Table 8)(Table 9) Here's a comparison between the two (Table 10)

The experiment with midfielders showed the same odd results as forwards and flanks: After analyzing 30 games of in-game footage, a pass-related stat of 20 was worse than a pass-related stat of 18 in a variety of areas, including missed passes and successful long passes (Tables 11-1,2,3).

When i experimented with pass-related stats from 14 to 20 for midfielders with stats based on 14 (Table 12), found that three midfielders - the advanced playmaker attack, box to box, and deep lying playmaker support - had fewer assists at 20 than at 18 (Table 13).

Why am I getting these strange results? I contacted the developer and was told that there are FM developers here and they should have an answer. 

Why are pass-related stats having this effect on the game? 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Resources.7z

Edited June 25 by walker14

[santy001]

There are some attributes we can't see but based on the model citizen description are likely to have been set to favourable levels. 

It's easy to forget that these hidden attributes are very important in their potential scope and that "normalising" them for tests isn't necessarily something you can control. For example 20/20 consistency doesn't guarantee consistency in every game. There's still a considerable chance that the player will be off form from that attribute alone without any way of knowing this. It's also not disclosed at what point or points this is assessed for a match or throughout a match. In theory a good number of players could be perpetually locked into a state with such a test that will always give out heavily flawed data that were you to have done the same for the very next match produced very different results.

The reason why you're getting weird results is quite simply weird inputs. The engine isn't designed to function in this manner and SI don't have to test the games functionality in this way. They have their own internal tools which can do a much better job. 

[NineCloudNine]

It’s hard to tell if your experiment has uncovered something interesting or just statistical noise:

* Your control group contains far too many weird inputs. Maxing team cohesion, morale and tactical familiarity is fine because this is achievable in-game but the player profiles are too extreme to act as a comparison.

* 100 games just isn’t enough to draw out significant structural features of the ME. Natural performance variation in football is so great and scoring so low that you need to simulate vast numbers of games to be sure you’ve found something that isn’t just a streak.

* Individual attribute differences of 16-18-20 are just not large enough to distinguish without very very large sample sizes of simulated games.

[walker14]

Sorry, but the answer is too abstract. Is it a strange input to change only pass-related stats, but not all other stats? The experimental team's tactics are universal, which is the sum total of their in-game results. If the stats work well in a better tool, why wouldn't they produce better results in 300 games? 

 

[wazzaflow10]

1 hour ago, NineCloudNine said:

It’s hard to tell if your experiment has uncovered something interesting or just statistical noise:

I've read this post 5 times and still can't figure out what they're trying to prove.

[Andrew Marines]

22 ore fa, wazzaflow10 ha scritto:

I've read this post 5 times and still can't figure out what they're trying to prove.

They're quite clear about what they're asking...

They're not trying to prove anything. Just trying to understand why improving passing related attributes actually decrease stats passing related and I'm actually curios why that is the case.

I do think too they need to do a bit more testing with various scenarios but saying they're not clear about what they are asking is a bit of an issue at your end...

[starbugg]

2 hours ago, Andrew Marines said:

They're quite clear about what they're asking...

 

Really?

When I read stuff like this, I am glad that I just play Football Manager FOR THE ENJOYMENT.

I will never learn to stop reading this stuff.

[Andrew Marines]

23 minuti fa, starbugg ha scritto:

Really?

When I read stuff like this, I am glad that I just play Football Manager FOR THE ENJOYMENT.

I will never learn to stop reading this stuff.

We're not talking about why we play the game.

What you enjoy doing is quite different from what I enjoy and  it is still different from what op enjoys.

Like IRL football, someone loves just to play it, someone else maybe just loves to analyze the data.

Those experiments are really welcome 

Edited Wednesday at 13:36 by Andrew Marines

[wazzaflow10]

3 hours ago, Andrew Marines said:

They're quite clear about what they're asking...

They're not trying to prove anything. Just trying to understand why improving passing related attributes actually decrease stats passing related and I'm actually curios why that is the case.

I do think too they need to do a bit more testing with various scenarios but saying they're not clear about what they are asking is a bit of an issue at your end...

I know what they're asking. I would have said asking if I didn't understand the question. Its a poorly conducted experiment hence why I said I don't know what they're trying to prove.

[walker14]

15 hours ago, wazzaflow10 said:

I know what they're asking. I would have said asking if I didn't understand the question. Its a poorly conducted experiment hence why I said I don't know what they're trying to prove.

Which part was experimented poorly? If you tell me your opinion in detail, i will reflect it. 

[wazzaflow10]

10 hours ago, walker14 said:

Which part was experimented poorly? If you tell me your opinion in detail, i will reflect it. 

A lot has been addressed already namely by santy and ninecloudnine. To reiterate some of their stance:

Your "control" variables aren't realistic. These inputs don't occur within the game naturally so it isn't something that would happen except in maybe very, very extreme cases. There certainly wouldn't be a team filled with this sort of distribution of attributes, let alone two. In short it's not an internally valid model. What I mean by that is by running an experiment that falls outside of something that has been trained you're inevitably going to find something "weird" because the model has never seen anything like this before.

As a concrete example suppose I'm growing tulips. I want to measure how tall they grow based on watering schedule, sunlight exposure, soil conditions etc. After a few trials I have an idea of what affects height. If I all of a sudden start planting petunias instead, none of my data for tulips is relevant. There may be general principles like all plants need water, sunlight, soil but you can't really claim to know anything about how a petunia would grow.

FM is the same way. The match engine has been trained and tuned on a particular set of data that exists. Just because you can change an input doesn't mean you should. That's not a control variable. FM experiments need to exist more or less naturally inside the game world. Putting in extreme values is going to yield an extreme result. You'd have to make a pretty compelling case as to why this initial state of inputs should exist.

Additionally, I have no idea what your tables are telling me in conjunction with the text. Is there some sort of regression? Is it just descriptive statistics - if so is there any association at the raw match level? Which group is which? Are the value statistically significantly different? There's probably more I could address but I really don't have an idea what you're trying to show to direct you further here. Its just very unclear other than you've shown me a table and claim something is weird.

[walker14]

6 hours ago, wazzaflow10 said:

A lot has been addressed already namely by santy and ninecloudnine. To reiterate some of their stance:

Your "control" variables aren't realistic. These inputs don't occur within the game naturally so it isn't something that would happen except in maybe very, very extreme cases. There certainly wouldn't be a team filled with this sort of distribution of attributes, let alone two. In short it's not an internally valid model. What I mean by that is by running an experiment that falls outside of something that has been trained you're inevitably going to find something "weird" because the model has never seen anything like this before.

As a concrete example suppose I'm growing tulips. I want to measure how tall they grow based on watering schedule, sunlight exposure, soil conditions etc. After a few trials I have an idea of what affects height. If I all of a sudden start planting petunias instead, none of my data for tulips is relevant. There may be general principles like all plants need water, sunlight, soil but you can't really claim to know anything about how a petunia would grow.

FM is the same way. The match engine has been trained and tuned on a particular set of data that exists. Just because you can change an input doesn't mean you should. That's not a control variable. FM experiments need to exist more or less naturally inside the game world. Putting in extreme values is going to yield an extreme result. You'd have to make a pretty compelling case as to why this initial state of inputs should exist.

Additionally, I have no idea what your tables are telling me in conjunction with the text. Is there some sort of regression? Is it just descriptive statistics - if so is there any association at the raw match level? Which group is which? Are the value statistically significantly different? There's probably more I could address but I really don't have an idea what you're trying to show to direct you further here. Its just very unclear other than you've shown me a table and claim something is weird.

Your comment confuses me. What specifically do you mean is unrealistic? All positive hidden values are 20 and negative hidden values are 1? Let's say i change the hidden value to level 10, which is a realistic number. The variables affecting the game (to what extent this will need to be confirmed through a separate experiment) will only increase further. The realistic figures you are talking about are the act of alternately planting tulips and petunias, checking how tall tulips have grown. Did I change a stat that had nothing to do with passing to measure a number related to passing? no. I kept other things fixed and only changed and measured the stats related to passing. I don't know what you specifically mean by the extreme values you're talking about.

 

My English is poor, so I didn't understand the latter part of your article correctly, but yes, if repeat it, will get similar results. and there will be errors between different groups. 

[wazzaflow10]

You've changed the inputs into the game to such a degree that you're bound to get weird results. I don't know how else to explain this more simply. Its not about 20's or 1's or 10's. Its that you've gone outside the realm of what the game has been trained on. The game has likely never been tested with the inputs you've given it. Its not to say that nothing can be changed but you've gone from 0 to 100 without documenting anything in between.

12 hours ago, walker14 said:

Did I change a stat that had nothing to do with passing to measure a number related to passing? no. I kept other things fixed and only changed and measured the stats related to passing

You have you just don't realize you did. Passing isn't just the passing attribute. Its a combination of attributes including footedness, flair, vision, composure, anticipation, decisions, technique.

Keeping things "fixed" doesn't make it a control group. Your control group would be where nothing is changed from the base game. You'd have to run many, many trials to figure out the expected range of results. There's a lot of RNG components that go into a single match you can't control for anyway as an end user. To overcome that it requires a significant amount of trials to find the limits of the game as they occur naturally. Its a lot of the reason these tests are somewhat pointless as an end user. You're trying to reverse engineer the game but have no mechanism to keep things in a truly stable state.

Anyway, once a natural limit is discovered, then you can go about experimenting the effects of changing a single attribute and measure their effects on results compared to the control groups. Eventually once you've documented incremental changes, then you can go about doing groups of changes to see how they interact and affect the engine. There's numerous ways to do it depending on how complex you wish to test the effects.

A weird result would be if a small change completely knocks the game off balance rather than just slightly nudges it in an expected way. If it takes getting to changing lots of inputs to something that wouldn't occur naturally then all that you've shown is the match engine hasn't been trained to handle those inputs and more or less broke the game.

Its fine that the engine has been broken because a certain set of inputs caused unexpected results but its not a valid result to say the entire engine is broken for all inputs.

[NineCloudNine]

14 hours ago, walker14 said:

Your comment confuses me. What specifically do you mean is unrealistic? All positive hidden values are 20 and negative hidden values are 1? Let's say i change the hidden value to level 10, which is a realistic number. The variables affecting the game (to what extent this will need to be confirmed through a separate experiment) will only increase further. The realistic figures you are talking about are the act of alternately planting tulips and petunias, checking how tall tulips have grown. Did I change a stat that had nothing to do with passing to measure a number related to passing? no. I kept other things fixed and only changed and measured the stats related to passing. I don't know what you specifically mean by the extreme values you're talking about.

 

My English is poor, so I didn't understand the latter part of your article correctly, but yes, if repeat it, will get similar results. and there will be errors between different groups. 

You believe you have eliminated other attributes from the test by standardising their values for all players. But this assumes that the effects you are measuring exist independently of the attributes you have standardised. They do not. Successful passing includes a very large number of factors, of which the passing attribute itself is only one.

You have created an entirely artificial and weird set of players. So there is no way of knowing if your results tell us anything about normal players, or just tell us that weird inputs produce weird outputs.

[walker14]

9 hours ago, wazzaflow10 said:

You've changed the inputs into the game to such a degree that you're bound to get weird results. I don't know how else to explain this more simply. Its not about 20's or 1's or 10's. Its that you've gone outside the realm of what the game has been trained on. The game has likely never been tested with the inputs you've given it. Its not to say that nothing can be changed but you've gone from 0 to 100 without documenting anything in between.

You have you just don't realize you did. Passing isn't just the passing attribute. Its a combination of attributes including footedness, flair, vision, composure, anticipation, decisions, technique.

Keeping things "fixed" doesn't make it a control group. Your control group would be where nothing is changed from the base game. You'd have to run many, many trials to figure out the expected range of results. There's a lot of RNG components that go into a single match you can't control for anyway as an end user. To overcome that it requires a significant amount of trials to find the limits of the game as they occur naturally. Its a lot of the reason these tests are somewhat pointless as an end user. You're trying to reverse engineer the game but have no mechanism to keep things in a truly stable state.

Anyway, once a natural limit is discovered, then you can go about experimenting the effects of changing a single attribute and measure their effects on results compared to the control groups. Eventually once you've documented incremental changes, then you can go about doing groups of changes to see how they interact and affect the engine. There's numerous ways to do it depending on how complex you wish to test the effects.

A weird result would be if a small change completely knocks the game off balance rather than just slightly nudges it in an expected way. If it takes getting to changing lots of inputs to something that wouldn't occur naturally then all that you've shown is the match engine hasn't been trained to handle those inputs and more or less broke the game.

Its fine that the engine has been broken because a certain set of inputs caused unexpected results but its not a valid result to say the entire engine is broken for all inputs.

So what specifically is incorrectly fixed? Your statement is still vague. 

https://image.fmkorea.com/files/attach/new4/20240629/7193570349_44021718_b8e5b8c515e2e501347a1937faa9b93c.png

Look at him. What's wrong with him? The official website says that passing is influenced by technique and vision. I didn't lower the rest of his stats any lower than necessary. 14 is about average for the top 10 teams in the first division. Also, what is the basis for your statement that footedness, flair, vision, composure, anticipation, and decisions have an impact?

 Everything I fixed is in-game. I didn't modify the game engine, so it's odd to say that modifying it to block a variable causes weird results. 

So what you're saying is that the things I modified to control the variable didn't control the variable, they skewed the results. What is the basis for that statement? 

Are you a game developer? If you're basing your statement on game engines, please be more specific.

[walker14]

8 hours ago, NineCloudNine said:

You believe you have eliminated other attributes from the test by standardising their values for all players. But this assumes that the effects you are measuring exist independently of the attributes you have standardised. They do not. Successful passing includes a very large number of factors, of which the passing attribute itself is only one.

You have created an entirely artificial and weird set of players. So there is no way of knowing if your results tell us anything about normal players, or just tell us that weird inputs produce weird outputs.

I fixed all other stats to 14 except for the experimental stats when doing this experiment. 14 is the average stat for the top 10 teams in the first division. Is 14 insufficient? Why is it insufficient? What is the basis for your opinions? If your basis is valid, I will experiment in the environment you describe.

Edited Saturday at 00:16 by walker14

[perpetua]

On 24/06/2024 at 13:05, walker14 said:

In the inside forward attacking experiment, a player who is nearly perfect physically, mentally, and technically (Table 3) performs worse than a player with only physical ability and dribbling(Table 4) (Table 5).

Interesting conclusion.

Any observations as to why this occurs?

I wonder if it's related to more rigorous defending of the more well rounded player by the opponent?

Or perhaps the single tool player (ie. dribble only), being more likely to use their only tool makes the player more effective or the defending less effective.

 

[walker14]

3 hours ago, perpetua said:

Interesting conclusion.

Any observations as to why this occurs?

I wonder if it's related to more rigorous defending of the more well rounded player by the opponent?

Or perhaps the single tool player (ie. dribble only), being more likely to use their only tool makes the player more effective or the defending less effective.

 

In a previous experiment, I collected 30 games of in-game results and found that higher passing stats were associated with more passes and lower success rates. More passes were also outed 

Conversely, lower pass-related stats resulted in fewer passes but a higher success rate. 

This was the result of experimenting with different Passing Technique Vision stats. My hypothesis here was that passing would determine the frequency of passes, technique would determine the quality of passes, and vision would determine the length of passes. 30 is a very small sample size, so i can't be sure. 

However, later experiments showed that high pass-related stats don't help with offensive point production. Perhaps, these stats can get in the way of a player's offensive behavior. High passing stats lead to frequent passing, which is not necessarily productive. In the fm game engine 

I don't really know why this is the case, so I'm waiting to hear from the developers to see if they can answer. 

[perpetua]

2 hours ago, walker14 said:

In a previous experiment, I collected 30 games of in-game results and found that higher passing stats were associated with more passes and lower success rates. More passes were also outed 

 

To an extent this is as expected.  If you are a better passer, you can attempt more difficult passes and a higher chance of having the pass intercepted.  So are you concluding that with no tactical changes or changes to the role that a player is given, a player with a higher passing rating, holding everything else constant, attempts more risky passes?  If so, I take this as a big win for the player AI and the ME.

[walker14]

52 minutes ago, perpetua said:

To an extent this is as expected.  If you are a better passer, you can attempt more difficult passes and a higher chance of having the pass intercepted.  So are you concluding that with no tactical changes or changes to the role that a player is given, a player with a higher passing rating, holding everything else constant, attempts more risky passes?  If so, I take this as a big win for the player AI and the ME.

It does have an impact, but I think it's a very small one. 

When I aggregated the number of long, medium, and short passes for a deep lying playmaker DM in 30 games and watched and recorded long pass highlights. The graph shows that the higher the passing stats, the more forward and lateral long passes, and the more outed passes. but the difference is very small. 

And I experimented with the passing stats of the three midfielders I mentioned earlier. The opponent was set up with a line up in a quick counterattack template-based tactic. In front of them were Holland and Salah, with the opposition defenders lagging behind the friendly attackers in the pace race. If the passers were supplying more risky passes, they should have made more assists, but the difference in assists between the passing-related stats was very small. As I wrote, there were even fewer assists when the pass-related stat was 20 than when it was 18. 

It's one thing for stats to not work at all, or to have a small impact, that's a rule of thumb for anyone who's been playing this game for a while, but passes have a really weird effect, and that's what I'm wondering. 

[herne79]

Some questions and observations:

1) What add on are you using to alter player attributes?  If it’s not an SI approved/supported product that “may” cause unexpected issues (that’s been shown before).

2) Regardless of 1), when you altered player attributes, did you also change their hidden CA value and also kept CA within the maximum 200?

3) If you changed their CA, did you also change the hidden PA value so that CA did not exceed PA?  If you haven’t changed CA and PA the game will change CA back to within PA limits.

4) You’ve noted and questioned comments above regarding unrealistic inputs.  Having looked at your included file in the opening post, there are perhaps some issues there.  Your use of “14” across the board is puzzling - strikers and inside forwards with 14 for tackling and marking for example?  14 may well be an average, but that’s applying an across the board average to everyone - I’d suggest it needs to be much more targeted than that.  Further, a 195cm striker with a Jumping Reach of 10 is another concern.  It’s noteworthy that Researchers have guidelines directly linking JR with Height - I don’t know the limits but 10 seems very low (too low?) for a 195cm striker.   That’s just a couple of examples of how things appear somewhat unrealistic.

5) To an extent you are correct - if we increase a player’s attributes we should expect to see an improvement in their performance accordingly and it can be puzzling why that may not happen.  It’s impossible to say for certain what’s happened here without examining the underlying code (which only SI can do) however from a layman’s point of view we normally expect players to grow organically, rather than just zapping through a whole bunch of changes using an editor as you have done.  So normally speaking we’d see players developing at different rates with attributes developing in a non-linear fashion over a period of time.  The game essentially stress tests this development so that players don’t just develop exponentially and within set limits.  When you step outside of these “rules” there can be unforeseen consequences, which is perhaps what others above are alluding to.

Personally if I wanted to test if player passing might improve I’d probably try using different players from the database rather than trying to make or alter my own.  Even then that may still throw out issues, such as a new player may not gel with his teammates, he may not be suited to the tactic, he may not form good partnerships with others and so on.  It’s a bit of a minefield tbh and the only thing I can say for certain is that it’s by no means as simple as changing player attributes and running some test matches.

[wazzaflow10]

2 hours ago, walker14 said:

If the passers were supplying more risky passes, they should have made more assists, but the difference in assists between the passing-related stats was very small. As I wrote, there were even fewer assists when the pass-related stat was 20 than when it was 18. 

Because you're missing key components of how attributes work together on top of having an incredibly small sample size. First the gap between 20 and 18 is relatively small in the game world. Secondly passing again isn't just passing, technique, and vision. That determines the players ability to spot and hit the pass on a technical level in a vacuum.

If a player has low composure they wont do well if they are being closed down or under pressure. If their decisions are low they might pick the wrong time or option for a high risk pass. If they are one footed and forced onto their weaker foot it won't be as successful. If they have low flair they might not try to curl a ball around a defender. If they have low anticipation they might not know a passing opportunity is occurring in a few ticks of game time.

Then we get to what their teammates are capable of and what the defenders are capable of. The game isn't as simple as I made the "best" technical passer in the world why isn't he creating 100 assists per game?

[walker14]

48 minutes ago, herne79 said:

Some questions and observations:

1) What add on are you using to alter player attributes?  If it’s not an SI approved/supported product that “may” cause unexpected issues (that’s been shown before).

2) Regardless of 1), when you altered player attributes, did you also change their hidden CA value and also kept CA within the maximum 200?

3) If you changed their CA, did you also change the hidden PA value so that CA did not exceed PA?  If you haven’t changed CA and PA the game will change CA back to within PA limits.

4) You’ve noted and questioned comments above regarding unrealistic inputs.  Having looked at your included file in the opening post, there are perhaps some issues there.  Your use of “14” across the board is puzzling - strikers and inside forwards with 14 for tackling and marking for example?  14 may well be an average, but that’s applying an across the board average to everyone - I’d suggest it needs to be much more targeted than that.  Further, a 195cm striker with a Jumping Reach of 10 is another concern.  It’s noteworthy that Researchers have guidelines directly linking JR with Height - I don’t know the limits but 10 seems very low (too low?) for a 195cm striker.   That’s just a couple of examples of how things appear somewhat unrealistic.

5) To an extent you are correct - if we increase a player’s attributes we should expect to see an improvement in their performance accordingly and it can be puzzling why that may not happen.  It’s impossible to say for certain what’s happened here without examining the underlying code (which only SI can do) however from a layman’s point of view we normally expect players to grow organically, rather than just zapping through a whole bunch of changes using an editor as you have done.  So normally speaking we’d see players developing at different rates with attributes developing in a non-linear fashion over a period of time.  The game essentially stress tests this development so that players don’t just develop exponentially and within set limits.  When you step outside of these “rules” there can be unforeseen consequences, which is perhaps what others above are alluding to.

Personally if I wanted to test if player passing might improve I’d probably try using different players from the database rather than trying to make or alter my own.  Even then that may still throw out issues, such as a new player may not gel with his teammates, he may not be suited to the tactic, he may not form good partnerships with others and so on.  It’s a bit of a minefield tbh and the only thing I can say for certain is that it’s by no means as simple as changing player attributes and running some test matches.

Question 1. I used the in-game editor 
Questions 2 and 3. Yes, of course
Question 4. In the game, height is related to the jumping reach growth limit. Jumping ability is determined entirely by jumping reach, height is irrelevant. 

Question 5. What are you basing your statement on? I experimented with the editor by switching players and got immediate results based on that. If you're saying that switching players causes something to go wrong, I don't understand. 

What's the point of experimenting without fixing the rest of the stats? It's a not good experiment where don't control and don't know what factors other than the experimental stats are affecting it. I appreciate your suggestion, but it doesn't seem relevant. 

[walker14]

26 minutes ago, wazzaflow10 said:

Because you're missing key components of how attributes work together on top of having an incredibly small sample size. First the gap between 20 and 18 is relatively small in the game world. Secondly passing again isn't just passing, technique, and vision. That determines the players ability to spot and hit the pass on a technical level in a vacuum.

If a player has low composure they wont do well if they are being closed down or under pressure. If their decisions are low they might pick the wrong time or option for a high risk pass. If they are one footed and forced onto their weaker foot it won't be as successful. If they have low flair they might not try to curl a ball around a defender. If they have low anticipation they might not know a passing opportunity is occurring in a few ticks of game time.

Then we get to what their teammates are capable of and what the defenders are capable of. The game isn't as simple as I made the "best" technical passer in the world why isn't he creating 100 assists per game?

First of all, 100 matches is not a large sample, but i experimented with three midfielders at the same time. Do you see the small difference per 100 match in the table? 

In the inside forward and wingback experiments, I also combined the results of two players in 200 games to get a sample size of 400. In my experience, goalkeepers have the smallest difference per game, and forwards have the largest difference per game. For forwards, the results are similar in 200 out of 300 games, and the other 100 games have a difference per 100 match. So I tried to get 300 samples. For the passing experiments, I gave everyone 20/20 footed.

So the key is this: what is the basis for your assertion? There is nothing but assertion in your words. There's no evidence. Judgment, anticipation 14 is of course not a low number, but this is the second issue. What are you basing what you're saying about how that affects the pass?

Edited Saturday at 17:12 by walker14

[wazzaflow10]

1 hour ago, walker14 said:

First of all, 100 matches is not a large sample, but i experimented with three midfielders at the same time. Do you see the small difference per 100 match in the table? 

In the inside forward and wingback experiments, I also combined the results of two players in 200 games to get a sample size of 400. In my experience, goalkeepers have the smallest difference per game, and forwards have the largest difference per game. For forwards, the results are similar in 200 out of 300 games, and the other 100 games have a difference per 100 match. So I tried to get 300 samples. For the passing experiments, I gave everyone 20/20 footed.

So the key is this: what is the basis for your assertion? There is nothing but assertion in your words. There's no evidence. Judgment, anticipation 14 is of course not a low number, but this is the second issue. What are you basing what you're saying about how that affects the pass?

I can't continue this conversation in good conscience. You have no idea what you're talking about. There is no amount of telling you that you're incorrect that you'll accept.

Edited Saturday at 18:22 by wazzaflow10

[walker14]

8 hours ago, wazzaflow10 said:

I can't continue this conversation in good conscience. You have no idea what you're talking about. There is no amount of telling you that you're incorrect that you'll accept.

This is a simple story, tell me the basis for what you say. If the conversation has no basis, then we're just talking about whose imagination is more vivid, and there's no conclusion and no point. 

If I'm wrong, you can let me know with a well-founded argument. If the basis of your words are just your thoughts, that's are meaningless.