City is not rubbish, but Guardiola is

[richj24]

I say this with the full realization that I can only provide anecdotal evidence. However, I am a very, very heavy user, I played hundreds of hours actively on at least twenty saves of FM 2018, and I have played a lot of FM 2019 as well. Showing you my playing time is rather pointless because I always leave the game on at my home PC. Majority of that playing time was dedicated to playing as Manchester United, and I always create a new save after six or seven seasons because I do not enjoy playing with generated players. I think this experience allows me to provide some persuasive anecdotes regarding Manchester City's performance since I spent hundreds of hours playing as their direct rivals in the same league. What I have noticed from all those hours of playing, is that there is a problem with Manchester City and Pep Guardiola's performance, and that reveals a much greater problem in your game that has to be addressed.

As far as my experience goes, there was a very clear, distinctive and repetitive pattern with Manchester City in FM 2018. It starts with Pep Guardiola's team performing much poorer than expected, usually ranking at 6th or 7th in the league. He then gets sacked, another manager (usually Carlo Ancelotti) comes in, does much better, but by that time my Manchester United becomes so powerful for them to launch any credible title challenge. A similar pattern can be seen in FM 2019 as well. His replacement may be different from time to time, but one thing stays consistent: Pep Guardiola performs very poorly.

Therefore, the problem is not with the ability you have given MCFC players since they perform better with another manager, but with Pep Guardiola. Then with Guardiola, it is not his attributes/ability, as he is one of the most highly rated coaches in this game. I have very strong suspicions that this is largely due to your current match/tactical engine, that puts Guardiola's tactical style at a significant disadvantage. I attached two screenshots, of Guardiola recording a 55% win rate, and Zidane, his replacement recording a win rate of 88%. This is just one, of a vast number of trials that have yielded the same results. The same team, in the same season, performs much better under a manager with worse ability.

Some who like to philosophize about the tactical aspect of Football Manager may disagree, but Football Manager's tactics are more simplistic and formulaic than its developers claim. The game clearly favors certain approaches over others. I have played the Football Manager series when it was the Championship Manager in 2003, and every single series had an overpowered approach. There have been a lot of tactics that were close to being cheats. There are different ways to enjoy a game, but if your objective is to win, this game quickly becomes a very simple game of finding that favored approach and implementing it.

After repeating the same scenario over and over again, it has become very clear to me that Guardiola's approach of passing football is weak, and after numerous trials on this version, it seems like this is continuing. Some would retort by saying, it is "possible" for Guardiola to perform poorly in a game since a proper simulation is always able to present us with different possibilities. However, the number of times I have seen this happen is far too great for this to be a phenomenon presented by a properly functioning simulation engine. For one of the best managers in football, who has been given stellar attributes in your game, to perform that poorly, almost every single time, in tens of saves is very, very unlikely. Of course, I am not expecting this game to be perfectly simulated. I don't think artificial intelligence, and simulation has reached that point quite yet. However, for one of the best teams in the world, under one of the best coaches in the world, to perform that poorly over and over again because his approach is weak, seems like a fixable glitch that you can work on.

I have seen people get shot down on this forum for bringing up similar problems, and their testimonies have been called to be freak occurrences. I would like to say that it is not. I am not sure whether you find my testimony as credible. You have a vast amount of data, far greater than mine. However, as far as my experience goes. there is a very clear flaw in your engine that has been prevailing for a long time, which has not been addressed. I have brought this up because this suggests a problem far greater than just MCFC's performance in the game, but a certain simplicity and limitation in your game that weakens the excitement often associated with the game. This is a problem that has to be addressed, and I hope you do so soon.

[Mitja]

5 or 6 years ago i would had thought we'll have ME that will replicate real football way beyond what we have. 

why does every AI 4-3-3 need to be same, out of all roles there Pep plays with 2 wingers and 2 B2B mcs. 

[Vali184]

One picture is worth a thousand words, or so they say.

[sedge11]

To be fair this is the 3rd maybe 4th thread I've seen about man city and Pep, and you will get a lot of different opinions as, as proven from the threads he performs very differently in everyone's saves.

I think to improve on the point your trying to make, it would be worth posting what Zidane's win percentage is after the same amount of games.

Just to share though in my Pompey save Pep did really well, and in fact city have been poor in the 4th season after he left.

As you can see Pep won the league in the 1st and 2nd season and also won the champs league in the 2nd season. He left of his own accord and is now manager of Juventus. Benitez took over at city and they sit 4th with 8-9 games to go in the 4th season.

I think maybe its just the ME can't replicate his style as well as it could perhaps, and as this is the 3rd/4th thread on this then people do have issues with it, but on a lot of people's he also does perform so its difficult for SI to address in my opinion.

 

[bar333]

This isn't news. The match engine has heavily favored more direct, wing play based style of football over a patient, creative possession game since FM16 at least, and it has gone more and more in that direction every year. It's difficult not to see that, some people claim that's not the case at all and just go with the beaten old "it's your tactic" line, and I'm not here to rag on anybody - I'm absolutely willing to concede the possibility of some people having such a deep understanding of how tactics actually work in the ME that they don't see it unless they want to. Many others who design their tactics according to intuition and their understanding of how football works rather than how the ME works will see it, inevitably.

Like you I've ran a good number of saves in England in FM18 and also one in FM19 and Guardiola surviving into the new year in the first season was the exception rather than the rule. I remember it well because it always jumped out at me how awful his replacements always were, bringing back Pellegrini, Mancini etc.

Obviously FM is not realistic as a simulation but it'd be nice if more of it was at least believable, at least within the first couple years of a career. The fun in FM is doing something unbelievable in a simulation that is otherwise pretty believable. It's a bit less special when already within the first season you're getting a wacky and unrecognizable football world.

[Sticx]

To be fair it's just not Guardiola. Barcelona finished 4th in the league in my first season, Bayern 4th, PSG were knocked out of the CL by Inter. It's pretty much every big club doesn't play anywhere near what their real life counterparts do. I don't know how anyone can claim the ME doesn't favor more direct play over short patient possession. 

[richj24]

8 hours ago, bar333 said:

This isn't news. The match engine has heavily favored more direct, wing play based style of football over a patient, creative possession game since FM16 at least, and it has gone more and more in that direction every year.

1

 

4 hours ago, Sticx said:

To be fair it's just not Guardiola. Barcelona finished 4th in the league in my first season, Bayern 4th, PSG were knocked out of the CL by Inter. It's pretty much every big club doesn't play anywhere near what their real life counterparts do. I don't know how anyone can claim the ME doesn't favor more direct play over short patient possession. 

Exactly. For some years now the go-to tactic has been to position as many attackers as possible, throw fullbacks forward and have them cross the ball. It has been Fullback Manager for quite some time now. The balance was so off in FM 2018, that many users opted to go with 3 forwards without wingers because fullbacks perform so well in all phases of the game. I noticed that with FM 19 crosses have been slightly weakened, but the trend of blatantly favoring direct play way too much is continued. 

[Mitja]

but isn't the same fashion in top level football, there are hardly any teams that don't relly on directness of counter attacks? even the king of possessional football had to adjust to high tempo EPL to make his team successful. same can be said about his Barca they destroyed Real with counters in last El Classico. in modern football you simplay can't afford not to play on counters it's just the metter of how much space your opponents allow behind their defenisve line.

difference between counter attacking football and attacking football is (unlike in FM) marginal in real football, there is no Counter or Attacking Mentality there. did Dortmund play defensive football in yesterday's win over Bayern? they counter-attacked with numbers at any opportunity. no, they simply played deep and happily let Bayern to control possession. but you cannot play on counter and not be risky but in FM less risk is definition of more defensive football. also you can't control possession with shorter passing in final third and play with lots of risk, which is definition of more attacking football, on contrary your passing will need to be much more cautious. the risk team takes in such situation when it's them that need to attack constantly on packed defense is behind their back. 

such teams that play ''attacking'' or ''possessional'' football are usually top teams in their league who play against packed defense every week and need to have perfectly timed behaviour and movement in final third and organised attacks into small detail. such teams can either attack with high tempo (more direct) and sacrafice some possession or choose the harder way not to relly on conter attacks but such teams are very rare now. but once the ball is in final third against organised defense it's just a metter of how patient manager wants to play like Pep or Sarri or relly more on crossing. (we have these tools in FM too, Hit Early Crosse/Mixed/WTBIB)

i'd like to mention few words about passing directness and statistics. it's really interesting that in EPL most teams except City and Chelsea and few bottom teams have between 10 and 20 % of long balls compared to total passing attempts. i find this statistics really interesting since it shows that difference between teams that play short or direct style is almost marginal and is far more conected to team quality. also the % of long balls corelates a lot with passing % and possession stats to a degree.

one important thing that FM doesn't seem to simulate really well is the total number of passes per game and in real football the bottom teams have almost twice less passing attempts than better ones and almost three times less than City or Chelsea. in FM this differnce is almost marginal and this is surely one of the main reasons why managers that employ ''possessional'' style don't do really well, their teams don't dominate possession like they should.  in fact more Defensive Mentality generates possession far better and this contradiction has been major issue since introduction of 2D. 

in real life Mentility is just the sum of all other player and team instructions. now after finally having almost all the tools needed, even for transition phase maybe the time has come to get rid of this outdated but key instruction?

 

 

[andu1]

this ME favours direct style with lots of goals scored from crosses . thats why players like Messi and Aguero underperform badly and managers like Guardiola who are set up with short style dont have results. 

In every FM AI Mourinho overperforms in 90% of the cases and Guardiola underperforms. Is that so hard to sse by the dev team? a thing that happens for at least 3 

 

p.s people come with examples in which AI Guardiola performs better and the matches in the leagues are not even in full detail. please come with examples if you run the matches in full detail

[richj24]

1 hour ago, Mitja said:

difference between teams that play short or direct style is almost marginal and is far more conected to team quality. also the % of long balls corelates a lot with passing % and possession stats to a degree. 

one important thing that FM doesn't seem to simulate really well is the total number of passes per game and in real football the bottom teams have almost twice less passing attempts than better ones and almost three times less than City or Chelsea.

3

You made so many great points, but I want to emphasize this because it really illuminates the biggest weakness of FM's ME. I would even go as far as saying that EVERY problem in Football Manager's current ME stems from this inability to simulate the number of total passes, and the number of different types of passes.

Like you have said, the difference between a "short-pass oriented" team and a long-ball oriented team is much smaller than commonly understood. However, the difference in quality is actually what enables the good teams to make those marginal differences count. With good teams, their tactical objectives are very detailed, and they incorporate different approaches at different tactical phases. Good teams that play a "possession-based style football", are asked to make short and careful passes to build up and maintain possession, but very often they would make long and direct passes to exploit the opponent.

In here lies the dilemma for the developers of Football Manager. FM's ME is not sophisticated enough to properly reflect the small difference in style and the big difference in quality, but they are laden with the task to get the ME to produce realistic results. We cannot replicate, or even imitate how teams play with the tools available on Football Manager. They have to create sophisticated scenarios with simplistic tools, and the best way to do it is to add various arbitrary adjustments that force realistic results. Therefore the matches become very contrived as a result, and the simulation platform overall inevitably produces some strange characteristics. My suspicion is that this explains why possession based systems perform very poorly on FM. With the tools in FM's ME, Guardiola's MCFC would create mountains of overly dominant results. That is too unrealistic, so they made some adjustments to put such systems at a disadvantage.

However to play the devil's advocate, if the ME became "too realistic", most players would be very bored by the fine margins that they have to deal with. The game becomes too hard and complicated. Users, including myself, want to dramatically feel their influence on what is happening in the games, AND get realistic results at the same time. That being said, a completely realistic AI is not what I am asking for. All I expect are incremental improvements, of some obvious problems.

After all, this is a game and not real football. But I have to ask, has the game made meaningful improvements as the series progressed? I am not sure whether people realize how expensive Football Manager is. The top game studios produce a full priced game once every few years, and they make monumental improvements, but they are held to a much higher scrutiny than the Football Manager series, which is slightly cheaper but produces a new full release every year. Considering trend in ME improvements for the last couple of years, I am not sure whether progress is there. The various improvements made were largely peripheral, and not much has changed with the matches, and they are the core of this game.

[Svenc]

So the point is? That it is frustrating yes. That it never happens in real, no. All three goals from Angers after a sending off as well. If this wouldn't be in the game, it were a massively poor simulation of football. As a contest of who is really the superior side, one-off matches in football have never sufficed any, in particular as it's a) an ultra low scoring sports, and b) not for the lack of trying given that it is 90 minutes of kicking, and a significant amount of attempts per minute average. This is a silly sports in which silly men try to get a ball on a target using their feet of all things! That is, minus the keepers, who are allowed to use their hands, dirty cheats. It's a miracle that only North Americans seem to resist it in numbers really.

I would never argue that the game "perfectly" recreates football. I've already argued the opposite. Aside of ME issues, AI issues and more: Inherently there tends to be far less randomness in the game engine, overall -- same as there tends to be fewer injuries, as injuries cause massively player kerfuffles, at which point the game tends to be "blamed" for being "unfun", which is a fair point. Honestly, football management oft likely isn't very fun at all. It's an extremely high pressure job of thinnest of margins, with people who don't have an inkling about how competitive sports work from the go judging your work day in day out.)

The way the game tends to portray it is that it's neither as competitive as the real thing, nor as "random", so that you can feel like the uber special one. The Clubs Own Personal Jesus turning the fortunes around. Evidently in football teams can have actually better chances rather than just moar for weeks rather than one-offs and still drop points in sequence. Given that modern day analysis suggests that even the biggest chances may be roughly comparable to tossing a coin (which your average football scoreline and winning margin of +1/+2 goals at best has always somewhat hinted at ever since), that perhaps should be not much of a surprise.

[richj24]

15 minutes ago, Svenc said:

All three goals from Angers after a sending off as well.

My point is there for you to understand. I am not moaning about a handful of games. I am highlighting a very obvious phenomenon that consistently happens across hundreds of saves encompassing many full releases of the game. Every single idiosyncrasy or outliers you stated would be a massive problem if it happened as often as what I have described above. When it happens almost every single time, it is no longer an outlier.

Football Manager is a toy model of reality. That is why it is a game. They are however products, that have to improve.

[Svenc]

It's not a phenomen, it's conformation bias (see my paragraph about Guardiola "managing" in FM). I've never seen it otherwise once the save games were provided (oft even with terrrible tactics -- e.g some rubbish download from FMBase et all with no defenders outside the centre backs whatsoever, which the game allows but doesn't happen in football). If the game were as "random" as football, you would have seen Dortmund slumping to 18th place midseason at least once on any iteration. They had the better chances almost every single week -- and still could barely find a win. The real football world exists but once -- the FM football world exists as many times as any of of its hundreds of thousands of players starts a new save.

On FM, it's one-offs at best. Never much sequences. That's not strictly defending the game. The game tends to spring fewer surprises because of it. You could benefit of them as well. However, I think there is oft an issue in that AI tends to play extremely defensively far too oftenly.

edit: I had two tabs open. Both the last posts were meant to go to this thread if anybody could move them. @Ackter @herne79 @themadsheep2001Also sorry for possible confusion @rich. I agree that oft times, Guardiola tends to do quite rubbish indeed, he was a topic in that other thread as well, hence my moisplaced posts (which oft has some to do with the ME, but also primarily limitations of the AI, sadly). @richj24

https://community.sigames.com/topic/456352-love-the-game-but-times-like-this-are-quite-silly/

[richj24]

3 minutes ago, Svenc said:

It's not a phenomen, it's conformation bias

I am not sure you understand the extent of my observation. If you think I am lying about how often I have seen it, that is fine, since I didn't provide you with the evidence. But as far as I'm concerned, it is "believe it or not". There is no bias. Your suggestion that it is a confirmation bias, sounds like you still think that I am writing this after running a handful of saves. Like I have said, I have ran the same teams, in the same league, close to a hundred times in FM 2018, and I have been doing pretty much the same thing in FM 2019. The same thing happens over and over again with MCFC and Guardiola. Do you think that it is normal, for one person to see this happen this many times? Every single cliched, rehashed arguments are not really relevant. Unless you can provide compelling evidence that there is a block of data where MCFC performs decently, it is difficult for me to reverse my position, purely because of the compelling evidence that I have seen.

[richj24]

14 minutes ago, Svenc said:

Also sorry for possible confusion @rich. I agree that oft times, Guardiola tends to do quite rubbish indeed. @richj24

I am perfectly fine with people being skeptical about what I say, because I can't really produce compelling evidence. I erased a lot of my saves because of disk space, and I didn't really collect screenshots either... I understand that this is too anecdotal to really persuade anyone. 

[Svenc]

4 minutes ago, richj24 said:

I am perfectly fine with people being skeptical about what I say, because I can't really produce compelling evidence. I erased a lot of my saves because of disk space, and I didn't really collect screenshots either... I understand that this is too anecdotal to really persuade anyone. 

To make that clear: This was a reply I posted in the wrong thread by accident. I had two tabs and threads open, and in both of them Guardiola was part of the discussion. As I've argued in that other thread, Guardiola oft not doing well in-game isn't confirmation bias at all.  It's a real thing -- even back then with Bayern, which was a far more one sided league than the EPL will ever be. Sorry!

[richj24]

1 minute ago, Svenc said:

To make that clear: This was a reply I posted in the wrong thread. I had two tabs and threads open, and in both of them Guardiola was part of the discussion. As I've argued in that other thread, Guardiola oft not doing well in-game isn't confirmation bias at all.  It's a real thing. Sorry!

Oh so that was the confusion I was confusd about! haha no worries man

[Amarante]

People say the ME favors direct style football while i'm hear short passing my way to league titles. 

A short passing game is more difficult to recreate and create it also requires have the right roles and instructions set. 

[Mensell76]

33 minutes ago, Amarante said:

People say the ME favors direct style football while i'm hear short passing my way to league titles. 

A short passing game is more difficult to recreate and create it also requires have the right roles and instructions set. 

You are a human and can tweak the tactic to a sort of short passing possession based tactic that in this ME can 't look anything like Guardiola football. Strangling the opposition till the short through ball or final dribble can be and is made is not possible in this ME due to lack of final third movement and decisionmaking 

Guardiola in FM is far below par.

[DNV]

2 hours ago, Svenc said:

So the point is? That it is frustrating yes. That it never happens in real, no. All three goals from Angers after a sending off as well. If this wouldn't be in the game, it were a massively poor simulation of football. As a contest of who is really the superior side, one-off matches in football have never sufficed any, in particular as it's a) an ultra low scoring sports, and b) not for the lack of trying given that it is 90 minutes of kicking, and a significant amount of attempts per minute average. This is a silly sports in which silly men try to get a ball on a target using their feet of all things! That is, minus the keepers, who are allowed to use their hands, dirty cheats. It's a miracle that only North Americans seem to resist it in numbers really.

I would never argue that the game "perfectly" recreates football. I've already argued the opposite. Aside of ME issues, AI issues and more: Inherently there tends to be far less randomness in the game engine, overall -- same as there tends to be fewer injuries, as injuries cause massively player kerfuffles, at which point the game tends to be "blamed" for being "unfun", which is a fair point. Honestly, football management oft likely isn't very fun at all. It's an extremely high pressure job of thinnest of margins, with people who don't have an inkling about how competitive sports work from the go judging your work day in day out.)

The way the game tends to portray it is that it's neither as competitive as the real thing, nor as "random", so that you can feel like the uber special one. The Clubs Own Personal Jesus turning the fortunes around. Evidently in football teams can have actually better chances rather than just moar for weeks rather than one-offs and still drop points in sequence. Given that modern day analysis suggests that even the biggest chances may be roughly comparable to tossing a coin (which your average football scoreline and winning margin of +1/+2 goals at best has always somewhat hinted at ever since), that perhaps should be not much of a surprise.

Conversion Rate  is suddenly excellent happens a lot in real life - yes. But looking City (under Pep managing) playing this way? Have you ever seen Pep's team got humiliated this way by an underdog team? HELL NO. In his nearly 10 years of managing career, only Real Madrid, Bayern and Liverpool could do this to him but he has been learning to adapt to it recently already. This, only the first season and I am not managing Everton

[mp_87]

I think there are definitely issues with the Guardiola style of football. It is hard to translate it into match engine instructions/attributes and how the ME and some roles work (eg. the ultra direct and largely one dimensional wide players in the AM strata that work unsatisfactorily in this type of football on FM).

Some of the concepts, not just with but also without the ball, are though just beyond the capabilities of the game at this point in time. They probably will continue to be for some time yet. You just have to accept that sometimes.

Also this approach does require considerable in-game intervention. Constant little tweaks to player roles, instructions, where/how the ball is moved etc etc. depending on how the match is unfolding. I think that's largely beyond the AI, and lets be honest probably 99.9% of human managers cant be bothered to do all that.

However I think there are some things that probably should be scrutinised. As has been mentioned the last couple of editions of the game have seemed to shift the balance towards a rather more simple, get it wide at tempo and get it in the area type football that at times is a bit on the messy side. And, purely my impression, I might be wrong and I hope this isn't taken as criticism, I do get the impression that within those who make the game, and even with all the consultants they have, there is a lack of understanding really to this type of football and more (complicated and unnecessary for some....... ) heavily tactical/possession styles that have become en vogue in the last few years, and so the game struggles a little to keep up. That shouldn't be a surprise though I suppose, as most making the game will be (casual) football fans, and most of the theory/ideas behind it will go straight over the head of the vast majority of football fans.

 

[Mensell76]

If only it were possible to select human tactics in the FM editor (if there was a link between the editor and the tactics folder) and inject them into a manager profile. That would be a major step forward for AI representation in the ME. 

I'd love to inject Guardiala with the tactics of some of the masterminds on here @Rashidi or @Cleon to name just two. I myself would love create a typical Mark van Bommel 4-2-3-1 PSV  tactic to be used by AI Mark van Bommel so when playing against AI PSV you'd all be playing a little bit against me.

Of course this would have to be just the main tactic without set piece instructions (so no corner exploits and such). If this was possible SI would create a giant customer flexibility. If for instance you do not like my version of a Van Bommel tactic, create your own and select it in the Editor in the Mark van Bommel profile, start a new career and le voila. Or even better, select one of the best used and hailed tactics from fmscout or soritoutsi.

We need more diversity in playing styles between teams. We need more distinctiveness. More flexibility is needed to create great AI. Again SI make us your accomplice. The loyal dedicated fanbase is there. Help us help you! Think outside of the box, rework the link between Editor and FM or make it possible to inject human tactics in FM through the in game editor!

[Spurs08]

12 minutes ago, Mensell76 said:

If only it were possible to select human tactics in the FM editor (if there was a link between the editor and the tactics folder) and inject them into a manager profile. That would be a major step forward for AI representation in the ME. 

I'd love to inject Guardiala with the tactics of some of the masterminds on here @Rashidi or @Cleon to name just two. I myself would love create a typical Mark van Bommel 4-2-3-1 PSV  tactic to be used by AI Mark van Bommel so when playing against AI PSV you'd all be playing a little bit against me.

Of course this would have to be just the main tactic without set piece instructions (so no corner exploits and such). If this was possible SI would create a giant customer flexibility. If for instance you do not like my version of a Van Bommel tactic, create your own and select it in the Editor in the Mark van Bommel profile, start a new career and le voila. Or even better, select one of the best used and hailed tactics from fmscout or soritoutsi.

We need more diversity in playing styles between teams. We need more distinctiveness. More flexibility is needed to create great AI. Again SI make us your accomplice. The loyal dedicated fanbase is there. Help us help you! Think outside of the box, rework the link between Editor and FM or make it possible to inject human tactics in FM through the in game editor!

Nice idea, but where I think it might struggle would be that the AI wouldn't know what it was trying to do, and therefore wouldn't be able to effectively adapt the tactic at all - it would have to be nearly entirely plug-and-play.

[Mensell76]

Just now, Spurs08 said:

Nice idea, but where I think it might struggle would be that the AI wouldn't know what it was trying to do, and therefore wouldn't be able to effectively adapt the tactic at all - it would have to be nearly entirely plug-and-play.

AI has the same roles, same tactical instructions and settings to its disposal as we. Fundamentally there should not be any difference. Now this is me not knowing for certain but basing this on several reactions by MODS and SI who structurally state that AI can make use of the exact same settings and roles as we can.

@themadsheep2001 could you perhaps chime in?

[themadsheep2001]

8 minutes ago, Mensell76 said:

AI has the same roles, same tactical instructions and settings to its disposal as we. Fundamentally there should not be any difference. Now this is me not knowing for certain but basing this on several reactions by MODS and SI who structurally state that AI can make use of the exact same settings and roles as we can.

@themadsheep2001 could you perhaps chime in?

They do indeed have the same access as we have. it's making use of all those combinations that we're simply better at (overall anyway)

It's not as simple as hardcoding AI Pep to play like his real counterpart, otherwise there is no ability to apart

[Mensell76]

1 hour ago, themadsheep2001 said:

They do indeed have the same access as we have. it's making use of all those combinations that we're simply better at (overall anyway)

It's not as simple as hardcoding AI Pep to play like his real counterpart, otherwise there is no ability to apart

But would it be possible for AI teams to use human made tactics? Because of course they are now using human tactics as well but with not enough attention to detail 

[kingking]

14 hours ago, themadsheep2001 said:

They do indeed have the same access as we have. it's making use of all those combinations that we're simply better at (overall anyway)

It's not as simple as hardcoding AI Pep to play like his real counterpart, otherwise there is no ability to apart

 

14 hours ago, Mensell76 said:

But would it be possible for AI teams to use human made tactics? Because of course they are not using human tactics as well but with not enough attention to detail 

@Mensell76 Mate i Love your passion to improve the current system of FM, your analysis and ability to give solutions to current issues without holding back because you are afraid people will assume you hate the game.

Possible Solutions. 

Improve ME for Attacking Possession-Based Football

Issue in ME. Attacking Possession-Based needs improving

Improve A.I Managers 

Improve squad management, building and development. this can always be improved

Improve the tactical styles that A.I uses. The A.I cannot utilize roles and instructions like humans, therefore provide them better tactical styles to use.

 

Introduce 2 formations Tactical System

Possession Based formations (4-2-3-1 & 4-3-3) are Flawed for A.I. In real life when top teams use a 4-2-3-1 or 4-3-3 they defend using a 4-5-1, 4-1-4-1 or 4-4-2 for A.I that is not the case

Reason.

• Weak Defense. Defensively 4-2-3-1 OR 4-3-3 is weak, it leaves gaps in defense.
• Realism. Teams change and switch defensive shape to a 4-4-2 or 4-1-4-1 or 4-5-1 when using a 4-2-3-1 or 4-3-3  because of the defending benefits of having a 4 or 5 in midfield 

[Spurs08]

1 hour ago, Mensell76 said:

But would it be possible for AI teams to use human made tactics? Because of course they are not using human tactics as well but with not enough attention to detail 

What I'm think is that the AI probably uses templates close to the basic tactics available to humans e.g. a 4-4-2 Wing Play. It probably knows what to do in various situations with that e.g. if trying to hang on, go to 4-5-1, keep on the big bloke as a target man and sub off your poacher. It would be a lot harder for it to know what to do with a tactic it hasn't been coded for, has no idea about and has to try and analyse from scratch. It could fail to realise, for example, the importance of a certain type of pressing to Guardiola's game, and decide to abandon it as soon as it goes a goal down, ruining the tactic in the process and going further behind.

[Mensell76]

51 minutes ago, kingking said:

 

@Mensell76 Mate i Love your passion to improve the current system of FM, your analysis and ability to give solutions to current issues without holding back because you are afraid people will assume you hate the game.

Possible Solutions. 

Improve ME for Possession-Based Tactical Styles

Imbalance in ME. Possession-Based Football need improving

Introduce 2 formations Tactical System

4-2-3-1 & 4-3-3 are Flawed for A.I. In real life when top teams use a 4-2-3-1 or 4-3-3 they defend using a 4-5-1, 4-1-4-1 or 4-4-2, In fm that is not the case

Thanks for your nice compliment. SI offers the most adult platform to raise concerns, issues or ideas. It is all about the way we raise those to them.

I trust SI and the mods fully in understanding my passion for FM. And we are dealing with professionals. They can surely accept unbiased feedback.

Regarding your input , AI manager profiles in FM Editor are actually set up with a preferred attacking and defensive formation. Guardiola's preferred defensive formation for instance is 4-1-4-1 DM and his attacking formation is the wide version of that. It shows that SI already has defensive shaping in place but the execution of it isn't quite there yet and the transitioning towards ball retention and dominating possession currently lacks even more.

 

[richj24]

@Mensell76 I think that is definitely an option that the developers can explore. It is a tough question for them to address I know, since they would have to make a decision between using a somewhat arbitrary measure that the users will not be able to replicate.

@kingking brings up another big issue, which is squad management of the AI. Very often it is easy to see AIs opting to play players at 72% condition, and the team's morale is shattered because they refuse to address some minor issues from a key player. This has a lot to do with the game's difficulty, and that is another tough question for the devs. 

I understand that these issues are not easy, but I just want to encourage the devs to push harder for some improvements.

[kpsia518]

FM19.1.2 ME is tiki-taka unfriendly.
that why...

[Svenc]

13 hours ago, themadsheep2001 said:

It's not as simple as hardcoding AI Pep to play like his real counterpart, otherwise there is no ability to ap

Wasn't that parts of what the presets were for? There absolutely need to be some stuff hardcoded to them, such as possession based systems:

- always employing at least one role/player/duty in central midfield who tends to hold the centre of the park / control the pitch
- always employing a front line combination that actually links up with midfield, rather than pushing further forward ASAP

That's very easy to do. Furthermore, it isn't even anything special. Any AI manager that fails to "grasp" this could be argued to not even ever had an UEFA Z License of coaching badges (if that existed).   Given that SI must have received suggestions for like the better part of a decade, I'm more inclined they approach this with cautious, as they may realize competition may be a tad more tough for players. As, in tendency, a few additionally points that else may be dropped by such AI would now be taken over a season.

[Mitja]

1 hour ago, kpsia518 said:

FM19.1.2 ME is tiki-taka unfriendly.
that why...

this ME is football unfriendly for anything higher than balanced mentality and lowest tempo

[pauly15]

Back to back to back EPL champs on my save...

[Mitja]

2 hours ago, Svenc said:

- always employing a front line combination that actually links up with midfield, rather than pushing further forward ASAP

in 2D this ME looks more like barbecue party than football match. there are bunch of guys sitting in the box waiting for food, having fun, drinking and chating. those couple of guys in the middle that should be MCs look like they're doing something (they move actually), i guess it's them who are making the barbecue maybe. central defenders are also just chating with opposition forwards and it looks particullary funny when there are two opposition strikers. they give the feeling like they are not enjoying the party too much or maybe their team-mates forbid them to come to the barbecue (to cross the halfway line).

and fullbacks i think they are hard working waiters who want everyone alse make happy 

[richj24]

29 minutes ago, Mitja said:

in 2D this ME looks more like barbecue party than football match.

1

Poets have long pondered to write analogies like that.

Add to that picture the host (the manager), who is trying out different seating arrangements and strategies to make sure that everyone has a good time, but ultimately always unsure whether they are working.

[Vicz]

City are incredible so far on my save so...

[Mitja]

2 minutes ago, Vicz said:

City are incredible so far on my save so...

who's taking corners?

[Vicz]

2 minutes ago, Mitja said:

who's taking corners?

Couldn't tell ya but when I played them they scored from a counter attack. Sterling.

And won 1-0

[Mitja]

29 minutes ago, richj24 said:

Poets have long pondered to write analogies like that.

Add to that picture the host (the manager), who is trying out different seating arrangements and strategies to make sure that everyone has a good time, but ultimately always unsure whether they are working.

i think that's someone who paid for the party but isn't allowed to be there. me?