Idea for FM 2010

[Deathcab for Cutie]

There is a thread for this, read the stickies, thank you come again.

in just about every post you make an idiotic comment and then dissapear. grow up, if you have nothing intelligent to say, don't say anything.

i blame global warming for people like you.

[bigdunk]

However, PPM's determine certain aspects of the things you mention and there is nothing available to govern the players propensity to follow tactics or attributes, unless there was an "instinct" attribute added.

Is that not the Teamwork attribute?

For what it's worth the way I would code something like this is to use probabilities. Basically, for every decision that a player makes, you assign probabilities to the possible outcomes and then pick randomly, biasing towards those probabilities. So the match engine might decide that given player A is running down the right wing, he has the following options and probabilities:

Cut Inside and have a shot (20%)

Beat his man and whip in a cross (70%)

Turn back and play a ball to his fullback (10%)

The better an option is for the player to take, the higher the probability. Obviously there would be more options (wait for his fullback to overlap, dink a ball into a strikers feet etc, but I'm keeping it simple here). What would then happen, is that a players attributes, PPMs and tactical instructions would tweak those probabilities. Lets say for example, the player has a PPM "cuts inside", so we change the probabilities:

Cut Inside and have a shot (30%)

Beat his man and whip in a cross (60%)

Turn back and play a ball to his fullback (10%)

But he has been given a high mentality, so he favours the attacking options:

Cut Inside and have a shot (33%)

Beat his man and whip in a cross (63%)

Turn back and play a ball to his fullback (4%)

And he also has been given a high width setting, so he favours the wider option:

Cut Inside and have a shot (28%)

Beat his man and whip in a cross (68%)

Turn back and play a ball to his fullback (4%)

And so on, for all the settings and for all the attributes/PPMs etc. We then arrive at a set of probabilities and pick a random one according to those probabilities. So given the exact same situation 100 times, 28 times he would have a shot, 68 times he would cross it and 4 times he would play it backwards. You would then follow a similar process for evaluating how successful he was at that choice (e.g. he has 75% chance of crossing it well, but he has a high crossing attribute, but a low technique attribute and so on).

[Nomis07]

Bigdunk, that's fine for PPM's, but it could never be that black and white when it comes to a player making the decision to follow a tactic or act on instinct. If it was, the player wouldn't actually be acting on instinct.

[bigdunk]

The player is never going to actually 'act on instinct', it's just a computer game

The whole point of doing it with probabilities is that it's not black and white. Some part of the player is telling him to follow the tactic, some part of him is telling him to do what he feels is right, these factors playoff against each other and you get a decision at the end of it. A player doesn't always act on instinct, and he won't always follow his instructions - different circumstances will give different results and doing it with probabilities covers all that. I would be amazed if FM is currently doing something that is massively different to this at the moment anyway.

[Nomis07]

That makes my previous question even harder to answer What determines the situation in which these settings take affect?

[bigdunk]

The probabilities always take effect for every decision. The weightings (i.e. how much each tactical instruction, PPM and attribute affect each probability) would have to be figured out for each situation, by watching the match engine play a game and just assessing each decision, looking at why the decision was made by the player and so on. Basically it's mammoth task, which probably explains why it's taken so long for the current match engine to get to the level it has. If you look at the list of changes for the match engine, you'll see comments like "reduced the effectiveness of X", "reduced occurences of Y", "increased amount of Z" and so on. I would suggest that this is PaulC tweaking the weightings that are being used for these decisions.

In terms of deciding what the current in-match situation is (and therefore what weightings to use), there are various ways you could do it. I guess I'd probably weigh up various factors (position on field, distance to goal, proximity of own players and opposing players, match score and so on) and use those to determine which weightings to use. Something along the lines of the closer you get to goal, the more likely you are to shoot and so on. Obviously it would be a lot more complicated than that though.

[Nomis07]

In terms of deciding what the current in-match situation is (and therefore what weightings to use), there are various ways you could do it. I guess I'd probably weigh up various factors (position on field, distance to goal, proximity of own players and opposing players, match score and so on) and use those to determine which weightings to use. Something along the lines of the closer you get to goal, the more likely you are to shoot and so on. Obviously it would be a lot more complicated than that though.

That's what I thought and tbh that's the problem I have with it, it seems overly contrived.

I don't want my players to take a shot based on my tactics, but conversely I don't want them to take a shot based on how many yards they are from goal!

[Daley]

That's what I thought and tbh that's the problem I have with it, it seems overly contrived.

I don't want my players to take a shot based on my tactics, but conversely I don't want them to take a shot based on how many yards they are from goal!

So then surely they are then playing purely on instinct which we know that they can't and I think you said you didn't want them to. (apologies if I am wrong about the last bit Nomis07 but didn't read back through all your comments).

[Nomis07]

So then surely they are then playing purely on instinct which we know that they can't and I think you said you didn't want them to. (apologies if I am wrong about the last bit Nomis07 but didn't read back through all your comments).

You're missing out

No, i'd love them to play with a mixture of both and some of the suggestions in this thread are ideal, but probably impossible. For me, the current tactical system and the reliance on sliders, although not ideal, is the best option available.

[Daley]

You're missing out

No, i'd love them to play with a mixture of both and some of the suggestions in this thread are ideal, but probably impossible. For me, the current tactical system and the reliance on sliders, although not ideal, is the best option available.

LOL! I had already read through the whole thread as it interested me but I have since had to actually do some work so lost exactly who had said what.

I totally agree that you need a mix of both otherwise the game will either be 'who's got the best tactics manager' or 'who can buy the bast players manager'.

As you said that sliders are far from ideal but like yourself I can't see another way around it!

[Mitja]

of course sliders could easily be replaced with concrete instructions. problem with sliders is that they don't say anything particulare. we should clearly see and understand the difference between 5th and 8th notch of mentality. real life manager says for example: 'i don't want you to play risky passes, keep them simple, you can pass in any direction but try to pass forward as much as possible. but i want you to bomb forward and overlap winger and put as many crosses as you can'.

now which slider or notch is that? instead of clicking a couple of simple and understandable tick-boxes you need to experiment with mentality, forward runs, creative freedom, width, tempo, crossing frequancy, crossing from...to achieve the same effect.

i'm going to write a thread about it.

[Nomis07]

The problem with concrete isntructions is it makes it too easy to create a super tactic. Fair enough, people should be able to create super tactics if they want, but when people say "concrete instructions", I think pre CM4 and those games were a bit too easy.

[bigdunk]

That's what I thought and tbh that's the problem I have with it, it seems overly contrived.

I don't want my players to take a shot based on my tactics, but conversely I don't want them to take a shot based on how many yards they are from goal!

They have to decide to take a shot based on something The important thing with the system I outlined above is that no one factor is the sole factor in deciding something. In terms of taking a shot at goal, distance from goal is clearly a relatively big factor (you are not going to shoot from 50 yards away, you are almost certain to shoot from 1 yard away). Tactics is probably a relatively small factor. If a player is told to play through balls, he'll slightly favour a pass instead of a shot. The same with crossing. Maybe mentality will slightly affect it to, but that should probably be about it. But because every factor is taken into account, none of them should really dominate.

If a player is 5 yards from goal, but a team mate has an easy tap in if he squares it then a player would still be capable of deciding to square it, even if he is very close to goal. His decision making, creativity, compsure and so on would all be factored in, as would proximity of team mates, proximity of opposing players etc.

[bigdunk]

of course sliders could easily be replaced with concrete instructions. problem with sliders is that they don't say anything particulare. we should clearly see and understand the difference between 5th and 8th notch of mentality. real life manager says for example: 'i don't want you to play risky passes, keep them simple, you can pass in any direction but try to pass forward as much as possible. but i want you to bomb forward and overlap winger and put as many crosses as you can'.

now which slider or notch is that? instead of clicking a couple of simple and understandable tick-boxes you need to experiment with mentality, forward runs, creative freedom, width, tempo, crossing frequancy, crossing from...to achieve the same effect.

i'm going to write a thread about it.

One big problem with having big lists of checkboxes like this, is that it's very easy to create contradictory instructions. The other problem I see is that lists of checkboxes might describe the end result that you want, but they don't always describe how that end result is going to happen. It's a little difficult to explain, but if you say want a midfielder to play simple passes, you can't just tell him to play simple passes, you need to ensure that he has other players close by to pass it simply to. With the sliders, you are telling the players the how, not the end result, so it becomes easier to create an overall tactic. Always assuming you can get your head around how the sliders actually work.

[Mitja]

why are there 20 notches of closing down? that's 20x5 meters??! imagine manager saying to a team: 'right lads today we're gonna start pressing them 20 meters from goal, no wait a minute, 25 meters!!' even 10 notches would be too many imo. plus we should be able to see concrete line where we start pressing. that would make sense.

what i really can't understand is why does the time wasting slider have 20 notches? it's either you're wasting time or don't. 3 notches would be more than enough.

why 20 for creative freedom? what's the diference between 12th and 15th notch? 'i want you to use 60% of your flair, no wait a minute, 75%?!?' imo it's either you want player to keep izt simple or make things happening, again 3 notches would be enough...

[Mitja]

The problem with concrete isntructions is it makes it too easy to create a super tactic. Fair enough, people should be able to create super tactics if they want, but when people say "concrete instructions", I think pre CM4 and those games were a bit too easy.

if it's too easy to create super tactics than that is a ME problem not manager's. i understand this is a game but there's no super tactic irl. if you set your player to bomb forward, dribble, press all over the pitch all at extremly high tempo style, than he should tire very soon. that's why there isn't such thing as super tactic irl.

[Mitja]

One big problem with having big lists of checkboxes like this, is that it's very easy to create contradictory instructions. The other problem I see is that lists of checkboxes might describe the end result that you want, but they don't always describe how that end result is going to happen. It's a little difficult to explain, but if you say want a midfielder to play simple passes, you can't just tell him to play simple passes, you need to ensure that he has other players close by to pass it simply to. With the sliders, you are telling the players the how, not the end result, so it becomes easier to create an overall tactic. Always assuming you can get your head around how the sliders actually work.

it might be very easy to create contradictory instructions but person would soon figure it out. you can create contradictory instructions with sliders even easier but it's much harder to see the problem. i think avoiding tick-boxes or any other kind of concrete instructions was just an excuse not to change things tactically. if people on these forums think it's ok not to change slider system then i can't see SI changing it, it's far easier for them.

fact is that RL managers use concrete instructions and can talk to players. they show them where to start pressing, when to pass, when to shoot, when to dribble, how to move, defend and all that. at least what we could have are simple and obvious instructions and their graphical/text representations.

only problem i can think is how it would implemented to AI.

[bigdunk]

Basically, we have to accept that FM will not have the same 'tactical interface' that real managers have for the forseeable future (i.e. we'll never be able to talk directly to our players). The major thing a real manager can do that we can't at the moment, is to give context to his instructions. He can see "go forward when the ball is on your side of the field", or "go forward if player Y is in position to cover your runs" and so on. We can't do this, and any kind of interface that would let us do this, is likely to be horrendously complex, both to program and to use. So we have the sliders instead. What this is doing, is still allowing us to have control over how often a player does something, but the context-giving ability is controlled automatically.

With time wasting I agree with you, to a certain extent. It probably doesn't need 20 notches, but I guess this is just for consistency with other sliders, and it will tie directly into one of the managers attributes, which are all rated out of 20 anyway.

The difference between the notches is directly related to the probablity-type system I have described earlier (always assuming FM is using a system like this). If a player has 15 creative freedom, he might be 20% like to do an action which is against his tactical instructions. If he has 18 creative freedom he might be 25% likely to do it. It is marginal decisions that will be affected, i.e. where a player is quite likely to do a couple of things, and a couple of notches difference will have a noticeable effect. If a player is either 95% or 99% likely to do something depending on his creative freedom, the slider is going to make no practical difference in this situation.

And putting concreate lines on the tactics screen is probably not going to work, because so many other things are going to effect whether a player decides to close down or not. It will just confuse the user because there are times when a player may be on the closing down side of the line but not closing down, and times when a player may be the other side but is closing down.

[bigdunk]

it might be very easy to create contradictory instructions but person would soon figure it out. you can create contradictory instructions with sliders even easier but it's much harder to see the problem. i think avoiding tick-boxes or any other kind of concrete instructions was just an excuse not to change things tactically. if people on these forums think it's ok not to change slider system then i can't see SI changing it, it's far easier for them.

It's not necesarily the user that is the problem, it's coding the match engine AI to cope with contradictory instructions that could be difficult.

And if people are happy with the slider system, then why should SI change it anyway?

[Mitja]

And putting concreate lines on the tactics screen is probably not going to work, because so many other things are going to effect whether a player decides to close down or not. It will just confuse the user because there are times when a player may be on the closing down side of the line but not closing down, and times when a player may be the other side but is closing down.

i'm quite sure it is how it works irl. it's a tendency not absolute instruction. but then it comes to player quality (decision stat) - is he sticking to instruction or will he be doing what he thinks is the best thing to do.

[Mitja]

It's not necesarily the user that is the problem, it's coding the match engine AI to cope with contradictory instructions that could be difficult.

And if people are happy with the slider system, then why should SI change it anyway?

that's what i said could be the biggest problem - AI. but we don't know that.

i'm not a slider hater, i just think it needs to change for this game to move forward. and SI said they will be changing it in the near future. for me it would be bigger step than 3d introduction, if it was made well. i feel many other FM players would agree, there was a pole a week ago which showed new tactical interface was one of the things people wish for FM10 the most (with better ME and 'bug free game' if i remember corectly).

[wrightinio]

Right, where to start...

One big problem with having big lists of checkboxes like this, is that it's very easy to create contradictory instructions. The other problem I see is that lists of checkboxes might describe the end result that you want, but they don't always describe how that end result is going to happen. It's a little difficult to explain, but if you say want a midfielder to play simple passes, you can't just tell him to play simple passes, you need to ensure that he has other players close by to pass it simply to. With the sliders, you are telling the players the how, not the end result, so it becomes easier to create an overall tactic. Always assuming you can get your head around how the sliders actually work.

But surely managers in real life don't really explain the how? They do give the end product however. How the end result happens should depend on the player, their attributes and the formation, but they should always be striving for the end product of the manager's tactics. For example, Ferguson might tell Ronaldo to "shoot on sight" (an option which could be implemented). This is the end product. He does not tell Ronaldo to take on 2 players by doing stepovers, and nutmeg a 3rd every time (this is the how). This depends on the situation and the attributes of the player. The end product is the same, but the attributes of the players have more influence (this would make buying the right players to fit a formation and training your players correctly becomes more important).

The problem with concrete isntructions is it makes it too easy to create a super tactic. Fair enough, people should be able to create super tactics if they want, but when people say "concrete instructions", I think pre CM4 and those games were a bit too easy.

I agree to an extent, but I don't think super tactics really exist anymore. The likes of 'Diablo' were around not long after the new 2D match engine was put in place and gradually, over time, the effect of these tactics has reduced immensely to the point where I don't believe any exist any longer (especially ones that are readily available to the community and work for everyone in much the same vein as Diablo did). I think the game, and the match engine itself, has now evolved to a point where perhaps a simplifying of tactics could even make the match engine better as I think there are too many variables currently, which is why we have seen so many discrepancies in the engine.

if it's too easy to create super tactics than that is a ME problem not manager's. i understand this is a game but there's no super tactic irl. if you set your player to bomb forward, dribble, press all over the pitch all at extremly high tempo style, than he should tire very soon. that's why there isn't such thing as super tactic irl.

Agreed, but like I said in my above post, I believe SI are now past the point of allowing people to create super tactics no matter what they do. I think the experience and know-how is now in place so that these won't sneak into the game again. There are no super tactics in real life, but there are brilliant players that use their abilities within a good, solid tactical framework in order to win competitions.

Basically, we have to accept that FM will not have the same 'tactical interface' that real managers have for the forseeable future (i.e. we'll never be able to talk directly to our players). The major thing a real manager can do that we can't at the moment, is to give context to his instructions. He can see "go forward when the ball is on your side of the field", or "go forward if player Y is in position to cover your runs" and so on. We can't do this, and any kind of interface that would let us do this, is likely to be horrendously complex, both to program and to use. So we have the sliders instead. What this is doing, is still allowing us to have control over how often a player does something, but the context-giving ability is controlled automatically.

With time wasting I agree with you, to a certain extent. It probably doesn't need 20 notches, but I guess this is just for consistency with other sliders, and it will tie directly into one of the managers attributes, which are all rated out of 20 anyway.

The difference between the notches is directly related to the probablity-type system I have described earlier (always assuming FM is using a system like this). If a player has 15 creative freedom, he might be 20% like to do an action which is against his tactical instructions. If he has 18 creative freedom he might be 25% likely to do it. It is marginal decisions that will be affected, i.e. where a player is quite likely to do a couple of things, and a couple of notches difference will have a noticeable effect. If a player is either 95% or 99% likely to do something depending on his creative freedom, the slider is going to make no practical difference in this situation.

And putting concreate lines on the tactics screen is probably not going to work, because so many other things are going to effect whether a player decides to close down or not. It will just confuse the user because there are times when a player may be on the closing down side of the line but not closing down, and times when a player may be the other side but is closing down.

You say a new system would be horrendously complex, but what do you think is in place now? SI obviously have an EXTREMELY complex engine and coding system in place, which is why it's taken them several years to get it up to speed and it still isn't 100%. What I was suggesting was a focus on simpler instructions, with attributes taking a slightly more prominent role. Everything that happens now with the match engine seems to interlock, so why couldn't they make it with a slightly easier system? Based on your probability theory, it would just mean attributes being factored in more than they are currently, and providing the players with less intricate, more decisive, 'end-product' tactical instructions.

It's not necesarily the user that is the problem, it's coding the match engine AI to cope with contradictory instructions that could be difficult.

And if people are happy with the slider system, then why should SI change it anyway?

I think most people aren't too happy with the slider system. I think some would agree that the current system is the best of a bad bunch, but that in no way qualifies it to be the best possible. SI shouldn't rest on their laurels with this tactics system and I believe have already said that they will be looking to make changes in the near future.

I also think that it's pretty easy to set up contradictory instructions now with the current system, even more so than the slightly simplified version I've suggested. In that respect I'd expect SI to be able to create something could cope with a simpler framework since they've done well with the current one but it's now become too intricate and borders more on a fantasy game than a realistic management game.

[The Vet]

The second part of that already happens, it's just a matter of needing a good tactic for those abilities to click and work properly.

As for too many slider, i'm not a fan, but i'm not a fan of the "Ferguson wouldn't have to..." argument either. What you are saying is, you want players to play to your tactics, but you want them to have a mind of their own, in certain situations, this raises two questions;

1. How do you programme a character to have cognitive ability that, in certain situations, bypass the tactical system?

2. What fun is there in picking a formation and telling them to attack and defend? If we're going down the reality route, then we aren't going to have many other options.

As I said earlier, i'm no fan of the sliders, but i'm yet to see anyone come up with a better idea.

Sorry for late reply.

0) That would be the first for me. It has always come down to a great tactic, which is often one that fools the AI. Once you have a great tactic, your players attributes are relative unimportant. I have crap crossers for wingers and great ones, and they both do they job pretty equally. To me it seems CA>attributes, and to me that is a mistake by SI.

1) Make the tactic of less importance that it already holds. Each player has a multitude of attributes, and I reckon they are there for a reason, and each attribute has already programmed action/reaction one would think. Furthermore, it is not so much as to bypass the tactics, since the players have been doing this for quite some releases now, as it is a footballer knows how to play his position, and he would also know what general approach his team would take coming into a game.

2) What fun is there fiddling with sliders that make no sense at all? It is nothing but a trial and error sliders, and doesn't play to your ability as manager. I am not saying it should be just "pick a tactic, play and turn off brain", but to me the force of CM/FM has always been that it was easy access, and you could concentrate on the task at hand; managing your players, buying/selling and win games. Unlike other footy managers. Therefore it should have a coherent system that requires know-how about the game(sport) to succeed. The slider system does nothing of the sort, especially considering many of them have several notches up/down which makes them nonsensical.

By making attributes more important than tactics, it is up to you to create the perfect blend of players to succeed. And to me, that is better than having a system where tactics means so much more than what players you actually have in your squad.

[kimoss_isc]

Seriously remove OI...it's useless and it can damage a tactic