imo, The Match Engine is Failing Badly

This is a long post. I tend to post very rarely, but when I do, I spend a lot of time on it. I hope that's not a problem. Bear in mind everything I say here is with love for the game, and if any of it comes across as accusatory, it shouldn't. My questions are all genuine, they are not rhetoric or whinging.

This thread is for ideas regarding how to fix what I see as the failures of an old and hopefully dying match engine. With the move to 3-d, I think it's going to be far easier to see what's going wrong and why it's getting worse at modelling football now that you see more. Previously, in my opinion, the engine could get away with a lot of it's weaknesses because you couldn't see them directly and much of the time were just filling in the blanks with your imagination. Now that the stupidity of the engine is there to be seen, it's time to look at the problems and decide how best to fix them.

Problems:

1 - Passing is badly broken.

Even with high-tempo and short passing, I see far too many players taking a touch in a situation in which pretty much any top flight footballer would know better. Under pressure or when an obvious quick pass is on, receiving the ball is not an invitation to dally and wait for another option or fail some hidden "dice roll" relying on decisions or composure or some stat, and the player not knowing what to do so just standing there. The difference between good and world class is not that "good" players take longer to pass the ball and end up getting tackled in horrible situations, whereas World Class players instantly ping the ball 50 yards perfectly. It's the quality of the pass, the consistency with which they make it, and occasionally a moment of brilliance such as a pass that unlocks a defence. Yes, players do sometimes dally in posession and end up losing it. But it's a rarity at the top level, especially in dangerous parts of the pitch. Why? Because fundamentally, footballers know how to play football! Even League Two midfielders are actually really, really good at football! That's why people pay to watch them play. If I saw players taking as long on the ball in real life as I do in FM09, I would assume they had taken a brown envelope from the oppo.

Also, intricate one-touch passing is faintly possible, but it seems to come about only occasionally in what looks like a pre-scripted move which results in commentary like "that was a great move from so-and-so". How come I don't see said great moves more often, even in a good team? And how come I don't see such great moves starting but perhaps petering out before reaching the final ball, or in which the final pass goes behind the player or to the wrong side of him or is intercepted? It's like the "great move" is inserted into the engine. I don't know if this is the case or not, but it feels wrong. A good team will ping the ball about for some time in reality, whereas in FM even a good team playing a slick passing game against poor oppositions tends to see a lot of bad passes going astray and players losing posession because a defender gets to them. Which brings me to...

2 - Defenders are too good in some areas, and not good enough in others

If you watch a Premier League game, you may notice that players on the ball tend to have a little more time than in, say, the Championship or further down the leagues. Yes, there is someone in attendance marking them and closing them down, but that person is not diving in on them hell for leather unless they think they have a very good chance of winning the ball. Why? Because most players at that level will simply sidestep an onrushing player approaching at pace, or they will just pass the ball around with a nearby teammate and take advantage of them being out of position. Good players do not dive in constantly. Yet in FM09 I see players doing just this. You see so many tackles in FM it's completely unrealistic. In reality, you see far fewer tackles per game. Posession is more often given up via the ball going out of play or being intercepted, or a pass missing it's target. Defending is not really about winning the ball from the opponent, it's about forcing him to give it away via a mistake, or by winning headers and interceptions and recovering posession for your own team. It is definitely not a constant battle of tackling versus dribbling. A ball might be contested because it is not precisely under the control of one player or another, but it's not just a case of "I have the ball, take it from me" and some exchange of dice-rolls resulting in "oh yeah, you fall over now and I have the ball".

The areas defenders are not good enough is their awareness. I was playing a game just now and in the 85th minute we had clawed back to 2-2. I had five defenders on the pitch, had set a very deep defensive line, and had set as "marked" all the opponents attackers. They even had back-arrows and sliders set as far defensive as possible. I set closing down to very close to my area. Basically, we were in lockdown. There should be a solid line of five defenders facing any attack.

Until, of course, the opposition's lone striker managed - somehow - to draw all three centrebacks way, way out of position nearly into midfield towards him leaving a rampaging opposition midfielder to literally stroll past them all and collect a pass and slot home. I checked the defenders stats. I looked for anticipation, composure, decisions, concentration. All were what I would call good or better.

Of course, we will get a lot of FM fanboys saying "learn to play" and "your tactics are wrong, read cleon's thread" but I already did that. And the tactics here are not the problem because all the tactics in the world will not prevent these moments of madness that all teams see from their defenders.

Well it's true, defenders do make mistakes. But these kind of things happen match after match, when in reality you wouldn't see something that catastrophically bad in the top flight very often at all. Are you telling me all my defenders blindly follow some instinctive need to chase the ball? They aren't Rugby players for goodness sake. And in the last moments of a match, players are as focused on their jobs as they can be. The kind of wide open spaces you get magically appearing in FM are so rare, so incredible...yet every match I play, with a variety of teams, I see defenders making the same horrendous positional errors.

Why is this? I would suggest it's because the ATTACKING element of the engine isn't clever enough. The defenders can be programmed to be perfect, but the game engine is incapable of modelling clever attacking play. It cannot model the slide-rule pass properly, the perfect chipped through ball for an attacker who nips round the back and manages to just beat the offside trap, or even get a player to cross the ball well. Try a game as Barcelona and tell me - does the football in the engine even APPROACH the kind of thing you see from them? Do players make the same kinds of run? Have you ever seen a player score a goal like some of the goals you've seen Messi score? Do you see one-touch give-and-go football? I don't. I really, sincerely, do not. Whether I am playing as them or not, I see a lot of the same stuff bring me to....

3) Attacking play is Procedural and Dull, and goals are more about pace vs pace or jumping vs jumping than team moves.

More often than not I see the same tired moves played out. It's like the attackers only have one thing in mind when they have the ball. I see players making runs but they are bad ones. Huge gaps open up, and players don't see them. Obvious one-twos present themselves, and for some reason are not taken. Why? The engine - this engine which, by the way, SI have had iteration after iteration to work on and are still struggling with badly - is not smart. It's crude. You see more clever attacking play in a game of pro-evo soccer between two relatively skilled players, then you do in any game of FM.

Also - where are the dummies? Why, when my player is down by the corner flag and hoping to cross the ball, does he not shape to cross, cut back inside, and find himself with more time? This is one of those "moves" you see on a Sunday League pitch, it's not some exceptional piece of skill. Dummying a defender is pretty basic stuff. Why don't I see it? Why is it more often than not that an attacker tries to dribble straight THROUGH an attacker? Dribbling is about taking the ball around people, not through them! You do sometimes see nutmegs or clever pieces of skill that get the ball very close to an attacker, but more often than not, the dribbling is about using balance and control and your body to shield the ball from your marker whilst trying to get it closer to where you want to be. Watch a player dribble at pace - a lot of the time they tempt a defender into going to ground or attempting a tackle and boom - they are past them and gone. In FM I see attackers and defenders just hugging each until one is outpaced or some tackle dice-roll takes place. And the attackers are seemingly never left on their backside on the turf. Even very strong strikers seem to find it impossible to get in front of an attacker and use their body that way, yet this is such a basic, basic move for a striker, especially around the box. You get in front of the defender and now he has to foul you to win the ball.

(Speaking of which you don't see shirt-pulling and pushing very much in FM, yet it's pretty much endemic. Maybe at corners you'll see it a bit, but it goes on a lot...just a thought. It would also solve some of the engine's troubles because in reality players will "take" a cheap foul to win a freekick and let their team get back into position either attacking or defending. This isn't necessarily gamesmanship so much as standard game practice. If you can win a freekick when your team is under pressure or when you want to give them a chance to catch up to you, you'll do it, even if it's a weak one. I don't see this modelled effectively at all.)

You also don't see players crossing the ball in a sensible way. Most crosses, even from players with decent crossing skills (above 15 should mean you are crossing very, very well. Beckham crosses are not the sole preserve of Beckham, it's just that he does it 95% of the time whereas most players do it less than that) are seemingly sent in oblivious of the defenders or players in the area. There's plenty of pace on them but they lack any sense of purpose. A good player in real life can plant the ball onto the head of an onrushing striker, that's why crosses are so dangerous. And that's just a "good" player, because that's the real factor here. You shouldn't need all these myriad stats just to get a decent cross in. Skill on the pitch is NOT the preserve of the world class player. It's just that they do it more often, and more consistently, and slightly better.

4) Shooting....my God the shooting.

Even a rubbish striker can hit the target sometimes you know. And midfielders shooting from range don't HAVE to just blast it over the bar nine times out of ten. So why do they? In the boxed/released version of FM pre-patch I saw strikers taking a lot more of their one-on-ones and I was pleased with that element of the engine, because that's the point of a one-on-one. It's a striker's bread and butter. But with 9.2 it seems that they miss far, far too many and often opt to shoot when there is a ridiculously obvious opportunity to square the ball. Players will almost always opt to pass to a teammate who is stood two yards out unmarked. Sometimes they won't but more often than not they will play for the team. That's WHY they are professional footballers.

Also, a striker hitting the target in the six-yard box is going to score more often than not. That's just a fact of life. When they MISS one from there, the reason the crowd goes crazy is because it's unexpected. I see so many shots missed from ranges that should be near certainties that I have to wonder....what is going on with this engine? Why are excellent chances only being taken by players of certain calibre? Bear in mind that you wouldn't want to leave any striker with a shot in the area, let alone the six yard box. Yet in FM they seem to find it so, so hard to score what should be routine chances. I honestly cannot think of a reason for this to be so other than some artificial compensation. Too many goals in testing? Something like that?

And what is with long shots? Just HOW good do you need to be to score, unmarked, from the edge of the area? The answer is, not as good as Frank Lampard. There are other players who score good goals, but seemingly not in FM. You need some hidden "shoots from range" or "shoots with power" stat to even approach consistently score in situations most players would lap up. And of course you also need long range shooting, composure, consistency, big-occasion, decisions blah blah blah.

How would Ronaldo have scored his 40 odd goals or whatever it was, in FM? I've never seen him have a season that good, or many other players that "should" (Lampard for example) in this edition. That's whether I am playing as them or not, by the way. I watch them all closely. Instead it's someone like Adebayor - who has scored in my current game as Spurs 14 goals in 8 league games which is just.....insane. It's also not uncommon in FM! Why is he so good and someone like Drogba not? Why does a player like Milito at Genoa, who is actually a very good goalscorer, perform so poorly in FM?

Anyway, hopefully I won't just get the usual responses that I've posted too long a post, that I "just don't know how to play" or whatever. I watch a lot of football, and I play FM a lot, and have played every version of this game since the Amiga. And I see a game that is getting careless when it comes to modelling reality.

We - and by we, I think I can safely say I represent long-term fans of the game - really, really DO NOT CARE about media interaction. Please, please, please stop coming up with new ways to talk to the media, or set up training or what have you. The one thing, above all else, that we care about is the match engine. That's the game. It should be 75% of the time you guys spend coding the game. Really, very little else matters other than the engine. And once you've got it right, you can hone it, and hone it and stop saying "match engine rewritten for this version of FM!!!" because those words fill me with dread. I know that you've still got a whole host of things wrong because you've spent so long just getting it running at all, that you've forgotten all the detail.

Stop working on new crap. The game doesn't sell because I can tell a guy from The Mirror "I don't wish to talk about that" or actually do what I do and hit "send assistant". The game lives and dies by it's match engine. Please, get it playing like football. I'm sure it's hard as hell to program (I was a programmer for many years so I know what you're up against) but this is not about the code. This is about the design behind it which I think is deeply flawed.

Anyway, thanks for reading (or scanning at least).

Top quality post mate, read it all (some bits twice) I'd say I agree with 99% if not all of it.

I think the 3D match engine will give SI the chance to see a lot of these problems which have not been as obvious as others . We can at least hope for some improvement in 2010

I read it because it was worth reading. I don't agree with much, but I do agree with a lot. There are most definitely bugs/design choices which I disagree with in the match engine which I believe could do with sorting out. There does however seem to be a huge problem with how players understand the interface so that they can get their desired results from the match engine - the OP shows that several times. And that's really a huge problem when long term fans can no longer relate what they do to results ingame. Been there and did it myself for a few weeks before I decided I'd go and get some help with it.

The main point I really and wholeheartedly agree with is that it's obviously time to stop trying to rush a full new product out each year and then requiring 3 patches for even some very evident bugs and errors to be squashed which can only be put down to having to fulfill such a tight development cycle. (Overinflated Arsenal stats perhaps being a simple one which a bit more time in testing could have squashed.)

Tuppence of thoughts from me. It was good to read a well-written 'rant' (for lack of a better word) - thanks Tedder Road.

Nice post indeed and top marks for not simply having a one-line blast at the game but rather coming up with constructive criticism and suggestions. I especially like your comment about how clever attacking play is lacking in its presentation in the match engine. I would like to see dummies and flicks being incorporated in one day for sure.

However, you must acknowledge the fact that this is not "an old and hopefully dying match engine" as you had put it, but rather a brand-new one (as SI have mentioned) with 3D representation introduced for the very first time in FM history. As such, of course there will be hiccups and the lack of visual frills. We have to just give SI time and they should get the match engine and the 3D representation tuned towards perfection with each future installment of FM.

Lastly, you mentioned about your disinterest in media interaction and training and perhaps other things. While you are certainly entitled to your preferences, other gamers are entitled to theirs as well. So it is not too fair to claim that you represented a good majority of us gamers out there and want all those taken out or for SI to stop working and improving on those aspects of the game.

To be honest, I think separate teams work on both sides of the game, engine and interface, so your worries that SI are neglecting the match engine in favour of ornament-y interface features are slightly unfounded.

This is a great post. The faults which you have described lead to a situation where it's almost impossible to 'read' the situation on the pitch. I see my defenders running away from the ball when they could easily have won it, or my strikers miss easy chances time after time after time and I honestly don't know whether it's because of the usual inadequacies of the ME or whether I should be changing tactics/formation. Or maybe it's because I didn't give the right half time talk or possibly it's all as a result of the press conference.....And why, oh why, when I have set my team to defend deep and counter attack (using TT&F criteria for setting my formation up) do they bomb forward and swarm around the opposition area only to be caught by the inevitable long punt? Why do DCs suffer a sudden rush of insanity and charge out to tackle an oncoming opposition midfielder leaving a huge gap? Why do my tactical alterations appear to make no visible difference to the pattern of play on the pitch?

You are so right when you say that it's the ME which should be the most important thing. In their quest for 'realism' SI have provided more and more 'features' (half time talks, media, weather, etc). In doing so, they have lost sight of the basics. Their first priority should be a realistic match engine which is responsive to our input, where player behaviour is credible and where it is possible to read the game and alter one's tactics according to circumstances. After that, we can begin to concern ourselves with press conferences!

Personally I think there's too much goal action and not enough midfield action (if you get my drift).

Also "Key Highlights" are not "Key" enough. Way too much irrelevant stuff.

Totally agree with this. I've said elsewhere that the tactical side of things is too hard to get your head round, and part of the problem has to be the match engine.

Until watching the ME feels like watching a proper game of football it'll be too hard to see whats going wrong. If you saw your full back constantly miles away from the winger, you'd make him less attacking, or get him to close down or man mark the guy. But even when this is sorted you still see the pair of them running along together and the winger emerging with the ball, or just plain running through your guy... Well, why's he done that? Is the defender not good enough at tackling, or anticipation, or composure, or pace, or acceleration? Or his he annoyed because you said your midfield was your best position? Or did you wish him luck at the start of the game instead of telling him he could win?

Watching a real game, we could tell where the players weaknesses are - and more importantly the player would know and adjust his game, any full back who knows he's beaten for pace by a winger will try to make sure he's always in the wingers way, rather than consistently getting involved in sprints for the ball. Even if the manager has told him to mark closely he'll know enough about football to know "we'll get hammered if I mark him too closely", and adjust his game. Or he'll have a word with the captain, or the manager.

But until the ME looks and behaves like a real game we're left wondering whats happening. We can just about work the tactics to get our players in the right sort of position most of the time, but even then there's something else where I can't explain why the players do what they do.

It may just be that we're half way to a proper match engine at the minute, but until we get all the way the effect of the tactical and player interaction side of things needs to be toned down to reflect the limitations of the visual and written feedback we're getting on the game.

There are a lot of interesting points there, tedder.

But I have to say a couple of things.

Firstly, I think you've got to remember the enormity of the task facing SI with the match engine. You say they've had iterations and iterations to 'get it right', but simulating a football match is *incredibly* difficult and having all the time in the world isn't going to lead to a perfect representation of a football match. We can be constructive and suggest ways of improving the engine, but I think it's worth bearing in mind the sheer scale of this 'project'.

I think you're right in that attacking play (especially top-level attacking play) is the hardest thing to replicate and isn't perfect at the moment. I also think that SI have to, on a version to version basis, make sure that the match engine produces results and bottom-line statistics that are fairly represetnative of real life. So I suspect that when, for instance, some aspects of defensive intelligence are improved, you'll see better finishing in one-on-ones, because they'll need to tweak that to keep the goals-per-game ratios the same. Until the match engine is perfect, I think this need to produce believeable stats 'on the surface' will cause some parts of the engine to play out a bit unrealistically -- the 'artificial compensation' that you talk about might well have an element of truth, in parts.

Secondly, we have to remember that what we see (either in 2D or 3D) is just a representation of what's actually going on in the engine. It's hard to represent a player dummying for a cross, and it could well be happening but we just don't see it (I know that's a bit convenient, but generally there's lots more going on than we see). I think certain elements of defending are just shown badly on the engine -- defenders who lack concentration and positioning get caught out, but the way it's represented looks wrong, sometimes.

I think SI probably do spend a massive percentage of their time on the engine. But I suspect that, at some level, they *need* to come up with new features to keep people buying the game. It would delight me if all they worked on from one game to the next was the engine and certain parts of the tactical interface/regens, but it would make the game extremely difficult to market to a lot of people. The SI forums represent a core of FM gamers, but there are a lot of people out there who would think twice about buying "NEW FM10, with updated match engine and players!".

With regards jostling at corners and blocking attacking players off, I'm not convinced that the physics of the match engine really treat players as solid objects, at the moment -- players seem to run 'through' other players, but again that could be just the 2D representation playing tricks.

great post,but all is already known. No point in it because si wont reply you or even tell you what areas of the match engine they are fixing

great post,but all is already known. No point in it because si wont reply you or even tell you what areas of the match engine they are fixing

If SI replied to every criticism, they would be posting on this forum constantly and there's no point in adding a sticky because people take absolutely no notice of them.

That being said, the match engine and it's many foibles, imo, have been in a shocking state for a while. The difference with 09 is that the issues are more visible with 3D.

great post,but all is already known. No point in it because si wont reply you or even tell you what areas of the match engine they are fixing

A bit harsh I think.

There is a danger of going round in circles on these discussions, but IMO it is worthwhile getting as many views as possible. We're all using the same ME, but we're all seeing it from different views and have our own opinions on why what we think we're doing isn't what's happening.

Is it a problem with the ME?

Is it a problem with the on screen representation of the ME?

Is it a problem with some tactics in the ME?

All these things will lead to my tactics looking like they don't work, someone needs to get to the bottom of what's going on.

I wouldn't expect SI to defend every single criticism, they're meant to be developing the game after all! But we're all effectively testers of the current ME and I hope, if we're able to get to a general consensus, SI will listen to that. I hope...

Great post. I recently had a match where I was winning 2-1 with 10 minutes to go. Like I always do if holding a narrow lead near the end I set up defensive tactics. Ignoring this like always my players go up the pitch on another attack which i'm ok with at this point because my 4 defenders are sitting back picking up their 2 attackers (no problem here you would imagine). The ball is cleared to one of their attackers who has drifted into the left back area to pick the ball up then goes on a run back from the wing in towards my CB's. To my amazement my LB decides the best idea for him it to chase after the attacker who was already 5 yards away from him leaving a huge gap where there RM has pushed into, the ball is played back out to him who crosses to the back post where the original attacker who picked up the ball is standing open with no defenders around him for 10 yards and tucks the ball away.

This enraged me as it is such a basic mistake to make as any footballer with even half a brain would simply pass the attacker on to the CB who is in a much better position to close him down.

I wouldn't agree with you about the one-two's as i've seen quite a few instances of this, however it occured to me that defenders and midfielders never seem to go with the runs of attacking players and track back inevitably leading to the player in questions being clean through on goal with not a defender in sight. It reminds me of a game between 5 year olds who have not yet been taught about positions and all seem to gravitate towards the ball.

Good read, agree with pretty much all of it, we do want a realsitic match engine, i'm sick of trying to implement tactics which i think the engine see's as effective rather than as you would do in real life.

Thank you for the replies so far. Nice to see a long post that took me a while to write getting some considered replies - and some interesting points made too. A couple of things...

When I said "this old match engine", I feel that the ME we're looking at is a lot of the old code running on a new 3d engine. I know they claim to rewrite a lot from scratch, but they clearly don't because there's no way you could start with a blank screen every time. I think there are core modules they re-use, and of course it's all still based on the same approach to player statistics as it has been for iteration after iteration (player ability/potential etc.) also I suspect that a lot of the players decision-making logic is based on code from former editions of the game. After all, they got Ray Houghton in to help them and personally I suspect his involvement is the reason why the game can only really simulate English direct-ball football and not a 'continental' ball-keeping posession based approach to the game.

It's an interesting point that perhaps there's "more going on than we can see" with the ME and that a lot of what goes on is hidden from us, and just can't be represented by the simple 3d view. Perhaps there is, I would like to think that there is a lot more crunching going on, but how can there be a number crunched which results in the players being so bad so often? And showing dummies shouldn't be beyond the 3d view - the players already slide into tackles, and if the 3d engine is based on the old Virtua Soccer engine (a rumour I heard, possibly on these boards? Either way it was a game I had way back when) then it's surely not a big job to code in things like that? Maybe they could keep the engine they have from FM09 and just improve it for FM10. Not rewrite it, just tweak it and tweak it and add in new things.

I think that's one of the big problems with a truly massive overhaul every time. You claim to do it EVERY time! So you never get a settled engine that can be repaired, you just keep getting new problems and fans have to spend a whole bunch of time getting used to the 'nuances' to put it politely of a new ME. Yes, it needs work as I've suggested, but perhaps we can build on the FM09 engine, which is at least 3-d and therefore offers far greater feedback for testing.

The biggest problem with putting in player skills and interesting attacking play seems to be that players are purely reactive. There doesn't seem to be much going on upstairs with the players when they are there running about on the pitch. Maybe some numbers are being crunched and dice-rolls made, but that's not really enough for a player watching the engine. I don't see them making the most fundamental runs, and I also don't see players getting closed down and marked UNTIL they have the ball, which is a real problem. Often I have told a player "mark him closely, tight marking, aggressive tackling" but the guy is standing there in full view of my midfielder and only when he gets the ball does my player go haring towards him and dive in with a tackle. Again - I put this down to a lack of forethought in the simulation of the players minds.

3D was rushed-out.

I believe that SI never worked more on this version (they hired a lot of new people), but by introducing 3D...they had aditional enormous job.

It would be much better if the make 2D to perfection.

I also wonder why they just don't stick to one match engine code...and evolve it year by year. Why they are writing completely new, buggy ME?

Nothing wrong with the match engine imo,

All about managing expectations and just a little common football sense.

The Passing

My team is probobly an average side in league two so that means i have average players. So i've adopted a normal-quick tempo, direct, counter attacking style. It took 10-15 games for my team to adopt this and another 5 games for them to understand another (squad understanding in asst manager teamtalk feedback screen).

The Defending

This is probobly down to the level your playing against. My defenders are not amazing but i'm a cautious tactician and compensate that way. But you will find against better players that your defenders concentration and decisions can result in mistakes as the players you are against are better at both.

If you do not drop your defence against pace then of course you will stuffer (happens to me if i don't react) and also if a player is taller and stronger then you will lose some balls in the air.

Also make sure that you are more narrow against better teams, and also oppostion instructions to show them onto weaker foot can be a big help.

The Attacking

I'm happy with the way this is show as my style comes througn nicely on the 3D and seems to be knock it long for the big man to flick on.

My main striker scored 16 goals in a season with my other players chipping in with 2-3 goals to so i have no problem with this.

to sum this uip, the match engine isn't broke and like anything it will require tuning and improvement but still i think its perfectly good. You will also need to think about how motivated your players are. This can have a big effect in making them over achieve for collapse from over expectation.

are you sure you're playing FM09

I'd say the problems you see are mainly problems with showing the graphical engine. If you just play the game with commentary only it's not that different to the older FM/CMs. I'd say in terms of giving results it's not too bad.

Having said that the 3D engine does need a lot of work. While the OP is right to point out that attacking play seems to lack flair and imagination I think it's not that bad. I always imagine that the 3D graphics doesn't show things like dummies but behind the scenes they are considered. The main problem with the 3D representation I find is the woeful defensive displays. Most of which have been posted on these forumns many many times but there's a couple that seems to get fewer mentions.

It seems to me that defenders don't react to where where the opposition team is or what they are doing. Rather they react to where the ball is and where the person with the ball is. To illustrate: often the ball is passed long but sometimes the pass is poor and heads towards nowhere - this is not a problem, it happens in real life - however, the team that's cleared the ball in this way should be preparing to defend and win the ball back from the opposition who will certainly be launching an attack. Instead the team that's cleared the ball keeps running forward with several players running past the ball as if anticipating someone on their team heading/flicking the ball on. It seems to me here that the engine has considered where the ball is going but not where the opposition team (or in fact their own team) is positioned.

Similarly when the opposition passes the ball to a striker the nearest defender should be intercepting the striker's run by running to where the striker will be. Instead the defender keeps trying to run to where the striker is now and then when he's got there the striker has obviously run further forward. The defender now sets himself a new co-ordinate for where the striker is now. This kind of behaviour inevitably means strikers will get away from defenders 95% of the time since the defender is forever behind them.

Attacking wise, crossing is a problem. I don't expect many crosses to result in goals since that would be extremely unrealistic. Most crosses in real life don't result in goals but most crosses do get past the first defender. There are far too many crosses that hit the first defender and I think if the 3D engine is to be made as a credible representation of football then this is one of the many things that need to be changed.

There are a lot more problems but since they've all been mentioned on the boards a hundred times I won't repeat them. I suppose the one redeeming feature is that these problems are consistent. The computer oppositions suffer from them as much as I do. I'm suspicious with the number of complaints I see on the boards that *sound* as if the posters think the negative bugs are only happening to them. Though of course since I haven't played their particular games I don't want to say it's impossible. It's probably borne out of frustration which I can understand.

Nothing wrong with the match engine imo,

So basically, because your team playing simple cautious football is doing ok, the engine is fine? I am not saying the engine is broken because I am not succeeding in the game - I am doing fine. What I am doing is criticising the engine. Don't be results based!

4) Shooting....my God the shooting.

Even a rubbish striker can hit the target sometimes you know. And midfielders shooting from range don't HAVE to just blast it over the bar nine times out of ten. So why do they? In the boxed/released version of FM pre-patch I saw strikers taking a lot more of their one-on-ones and I was pleased with that element of the engine, because that's the point of a one-on-one. It's a striker's bread and butter. But with 9.2 it seems that they miss far, far too many and often opt to shoot when there is a ridiculously obvious opportunity to square the ball. Players will almost always opt to pass to a teammate who is stood two yards out unmarked. Sometimes they won't but more often than not they will play for the team. That's WHY they are professional footballers.

Also, a striker hitting the target in the six-yard box is going to score more often than not. That's just a fact of life. When they MISS one from there, the reason the crowd goes crazy is because it's unexpected. I see so many shots missed from ranges that should be near certainties that I have to wonder....what is going on with this engine? Why are excellent chances only being taken by players of certain calibre? Bear in mind that you wouldn't want to leave any striker with a shot in the area, let alone the six yard box. Yet in FM they seem to find it so, so hard to score what should be routine chances. I honestly cannot think of a reason for this to be so other than some artificial compensation. Too many goals in testing? Something like that?

The bolded bit is what concernes me about this match engine. Too often it feels like my players are intentionally messing up chances and acting daft to keep the stats believable. There's a thread in the bugs forum where someone replayed a match against a lower league opposition ten times and noted that the average shots on target percentage is around 40% while it's almost impossible to get over 50%. When it's usually remarkably higher against better opposition. When I see my players rattling the crossbar from two yards out in front of an open goal and doing it more than once it just does not look believable.

Oh, and the same goes for crossing. Marking is messed up for full backs which means wingers going past them and into the box is a common occurence. 90% of the time however these wingers go for a cut back when the obvious action (and one that would result in a goal more often than not) is to put one across the goal for an onrushing striker to tap in. Why would they cut it back if it wasn't for artificial compensation? It's much harder to pull it back than it is to cross it in reality and I doubt it would be the intuitive move for any half decent attacking player.

And yet another instance of this is when one team is supposed to be under pressure. Suddenly players in acres of space cannot control the ball or direct it instead of panicking and booting/heading it into touch. I don't know what the match engine is trying to simulate but what it looks like is one team deliberately conceding possession and putting themselves in a difficult situation.

Things like this make watching the match and changing tactics seem utterly redundant. Afterall, when creating good chances or getting players in good positions isn't rewarded what exactly are we expected to aim for?

Nice post

Really agree that the ME needs improving. As you rightly say, this should be the main focus of the game.

Heres hoping SI will take a good look at constructive criticism.

The bolded bit is what concernes me about this match engine. Too often it feels like my players are intentionally messing up chances and acting daft to keep the stats believable. There's a thread in the bugs forum where someone replayed a match against a lower league opposition ten times and noted that the average shots on target percentage is around 40% while it's almost impossible to get over 50%. When it's usually remarkably higher against better opposition. When I see my players rattling the crossbar from two yards out in front of an open goal and doing it more than once it just does not look believable.

Oh, and the same goes for crossing. Marking is messed up for full backs which means wingers going past them and into the box is a common occurence. 90% of the time however these wingers go for a cut back when the obvious action (and one that would result in a goal more often than not) is to put one across the goal for an onrushing striker to tap in. Why would they cut it back if it wasn't for artificial compensation? It's much harder to pull it back than it is to cross it in reality and I doubt it would be the intuitive move for any half decent attacking player.

And yet another instance of this is when one team is supposed to be under pressure. Suddenly players in acres of space cannot control the ball or direct it instead of panicking and booting/heading it into touch. I don't know what the match engine is trying to simulate but what it looks like is one team deliberately conceding possession and putting themselves in a difficult situation.

Things like this make watching the match and changing tactics seem utterly redundant. Afterall, when creating good chances or getting players in good positions isn't rewarded what exactly are we expected to aim for?

I've seen the correlation between the amount of shots, or how much your team is dominating the game, and your team's percentage of shots off target and being blocked. Incredibly frustrating to believe that a team with over 60% possesion and over 15 shots is only getting 2-3 shots on net and 1 or 0 CCCs. I've seen it often enough to believe it's not just one of those days for my team - it certainly feels like the ME is holding my team back. In fact, most of my high scoring games come from very tight games according to stats, with my team simply converting on a high percentage of their shots and CCCs. When I have really dominant games, it is more likely a low scoring win or draw.

I agree, I'd like to see some more 1-2s and dummies and all that stuff. You could say there are some 1-2s already in the game, but they aren't the same as what you'd see from Barcelona for example. The types of runs that Xavi and Iniesta make I would like to see. Hard to describe, but they just get in the right place and make it easy for others to pass to them.

There are a lot of interesting points there, tedder.

But I have to say a couple of things.

Firstly, I think you've got to remember the enormity of the task facing SI with the match engine. You say they've had iterations and iterations to 'get it right', but simulating a football match is *incredibly* difficult and having all the time in the world isn't going to lead to a perfect representation of a football match. We can be constructive and suggest ways of improving the engine, but I think it's worth bearing in mind the sheer scale of this 'project'.

I think you're right in that attacking play (especially top-level attacking play) is the hardest thing to replicate and isn't perfect at the moment. I also think that SI have to, on a version to version basis, make sure that the match engine produces results and bottom-line statistics that are fairly represetnative of real life. So I suspect that when, for instance, some aspects of defensive intelligence are improved, you'll see better finishing in one-on-ones, because they'll need to tweak that to keep the goals-per-game ratios the same. Until the match engine is perfect, I think this need to produce believeable stats 'on the surface' will cause some parts of the engine to play out a bit unrealistically -- the 'artificial compensation' that you talk about might well have an element of truth, in parts.

Secondly, we have to remember that what we see (either in 2D or 3D) is just a representation of what's actually going on in the engine. It's hard to represent a player dummying for a cross, and it could well be happening but we just don't see it (I know that's a bit convenient, but generally there's lots more going on than we see). I think certain elements of defending are just shown badly on the engine -- defenders who lack concentration and positioning get caught out, but the way it's represented looks wrong, sometimes.

I think SI probably do spend a massive percentage of their time on the engine. But I suspect that, at some level, they *need* to come up with new features to keep people buying the game. It would delight me if all they worked on from one game to the next was the engine and certain parts of the tactical interface/regens, but it would make the game extremely difficult to market to a lot of people. The SI forums represent a core of FM gamers, but there are a lot of people out there who would think twice about buying "NEW FM10, with updated match engine and players!".

this ME doesn't feel like watching football match, nobody is talking about perfect engine. if it's so hard hard to make a really good one, why do they make ''new'' engine each year? why didn't they try to iron out 07 ME for example? while there were too many new features introduced in last couple of years most of which don't work as meant or are purely cosmetical, ME hardy moved on. same old thing. it's biggest achievement since FM06 is that strikers don't score 50 goals a season no more. but it still doesn't feel like half decent football.

there's probably not a single feature in FM that doesn't need a serious look at (even player rating system which worked fine for a decade got messed now). fixing the most important features like ME, transfers, training, tactics counts more than a milion new features. will they manage it? i think FM is a sleeping giant who got poisoned by his own cooking. he needs to open a window for fresh air...

top post, i agree with all.

I totally agree with everything the OP has said. I'm not to bothered about press conferences and things like that. Match day is the all important thing. It's the same in real life. Managers train their players and have press conferences but what really counts is matchday. If have a near-worldclass player who sees the a ball that he can easily win but he manages to run pass it - which means that the on-coming striker gets the loose ball and scores. In the commentary it is " x missed his interception and x made him pay ". Complete nonsense

When your watching the 3d play, you are not watching a match unfold before your eyes, your watching the Match Engines attempt at making a realistic visual description of a pre-determined event.

The game takes data from several sources

(Pre-Match press conference, form, team morale, tactics, team-talk, which players have started etc.)

It then produces similar data for your opposition.

It then compares the 2 sets of data and calculates the "Most Likely Result" from that data

the ME then attempts to recreate the MLR visually.

If you pause the ME and change your tactics, make a substitution etc, the MLR is recalculated from that point, and the ME then attempts to visually recreate the new MLR.

I too believe that the match engine should be #1 priority (with training/development #2 and transfers/finances#3), but I also like the additional stuff (media etc)!

Anyway, the OP is really interesting. I agree that attacking play is kind of undeveloped still and Paul C should be working on it. I agree that defense often looks stupid. And I agree that possession football is lacking horribly (that's the hardest part anyway). But you are being a bit unfair imo, because the match engine is new. Why is it new? Because it simulates a 3D world now, so it had to be rewritten (much of it anyway). What does this mean? That's it is in its childhood now and it's only going to get more mature.

Do you remember CM4? It was SI's first effort on 2D and it was amazingly bad. Eventually it got there, and FM08 had a very good 2D engine. To be frank, I was expecting a similar disaster with the first 3D engine (simlar to CM4) but fortunately things are much better now. FM09's engine is a good starting point, I think.

So hopefully SI will slowly incorporate all the things you are asking for, and will also take care of what McMahon's Lover said about goal vs midfield action. You hit the nail on the head a few times in your post (tackling, shooting etc). Real football at top level is mostly Strategy not Tactics, but FM has a hard time simulating strategy. We are being given a million options on Tactics (Run, Shoot Long etc for EVERY player, count the number of sliders!), but very few strategic ones (width, depth, tempo, focus passing, free role, hold up ball etc). It's not that I don't know how to do it, it's that the few (strategy) options won't let me do it. I don't think that overall strategy should be decided by the sum of our tactical decisions (e.g. you tell the guy to never cross, so he never goes wide). Rather, tactics should be the slave and strategy should be the master.

You have a strong case. Hopefully, we are going to read more from you around here.

PS. Real Magpies, no, that was the CM3 engine.

When your watching the 3d play, you are not watching a match unfold before your eyes, your watching the Match Engines attempt at making a realistic visual description of a pre-determined event.

The game takes data from several sources

(Pre-Match press conference, form, team morale, tactics, team-talk, which players have started etc.)

It then produces similar data for your opposition.

It then compares the 2 sets of data and calculates the "Most Likely Result" from that data

the ME then attempts to recreate the MLR visually.

If you pause the ME and change your tactics, make a substitution etc, the MLR is recalculated from that point, and the ME then attempts to visually recreate the new MLR.

That doesn't mean there shouldn't be things like stepovers, dummies, or CMs moving off the ball. These things may not affect the end result, but remember FM is a football simulation, and stepovers, dummies, and CMs making runs is part of football.

PS. Real Magpies, no, that was the CM3 engine.

No, thats EVERY match engine SI have released.

Including FM09.

You don't see Step-overs, or dummies because they aren't included as options for ME to show.

No, thats EVERY match engine SI have released.

Including FM09.

You don't see Step-overs, or dummies because they aren't included as options for ME to show.

Then maybe they should add them in?

Isn't that the point of this thread? :confused:

When your watching the 3d play, you are not watching a match unfold before your eyes, your watching the Match Engines attempt at making a realistic visual description of a pre-determined event.

The game takes data from several sources

(Pre-Match press conference, form, team morale, tactics, team-talk, which players have started etc.)

It then produces similar data for your opposition.

It then compares the 2 sets of data and calculates the "Most Likely Result" from that data

the ME then attempts to recreate the MLR visually.

If you pause the ME and change your tactics, make a substitution etc, the MLR is recalculated from that point, and the ME then attempts to visually recreate the new MLR.

I have no idea what difference you think this makes, to be honest. Or even if that is true.

However the resulting match is represented to the player, it needs to be done better in my opinion. At the very least there needs to be far more visual feedback so that the player can see what they are doing wrong or right. And it also needs to look more like real football.

As for someone's point that this is their first attempt at a 3-d engine so it's a "baby" - the visual side is new, but the underlying core values of the FM engine are still there. FM still overvalues certain attributes and STILL fails to adequately model certain player types because the engine is incapable of doing so. This is not about 3-d, this is about serious design flaws in the base model (again, in my opinion!).

I'm merely pointing out the underlying reasons for why we see what we see when the ME plays out the game. I'm not disagreeing with the OP, I'm just looking at the problem from a different perspective.

The ME itself is fine and works as intended, the visual representation is just eye-candy.

I would prefer SI to work on more fundamental behind the scenes issues than increase the level of eye-candy.

Glad you took the time to write that out, pretty much summed up how I feel about the match engine.

What annoys me is that I rememver a time when SI wouldn't have dreamed of putting something half-finished into an FM game. I also remember in 1999 that they said they would not put a 3D engine in until it was perfect.

How times have changed.

The FM09 match engine is just appalling all round. I don't understand it could get so much worse since FM08. The old match engine wasn't perfect, but it only had a few small problems. The new one doesn't have any redeeming features at all. SI would be better off scrapping it and going back to the FM08 match engine. If that means scrapping 3D, it would be well worth it.

Have to agree with all of that to be honest, one thing I did notice is that you mentioned the amount of one-on-ones missed. I agree that there are loads missed but, I think there are too many and instead of fixing that aspect of the match engine they've balanced it by making your strikers miss more.

What you have to remember is that the match engine is one incredibly complex piece of programming, and at the end of the day it is just a programme, and Football is alive and it flows.

It could be better but on the other hand it could be a damn site worse. and IMO the issues we have with the match engine are in there because fixing them causes issues with the balancing, and it's not that easy to fix the issues whilst keeping the balance, for instance strikers score a lot more one on ones, but you end up with 7-7 draws , now what would you prefer rubbish strike rates from one on ones or over inflated scores?

When your watching the 3d play, you are not watching a match unfold before your eyes, your watching the Match Engines attempt at making a realistic visual description of a pre-determined event.

The game takes data from several sources

(Pre-Match press conference, form, team morale, tactics, team-talk, which players have started etc.)

It then produces similar data for your opposition.

It then compares the 2 sets of data and calculates the "Most Likely Result" from that data

the ME then attempts to recreate the MLR visually.

If you pause the ME and change your tactics, make a substitution etc, the MLR is recalculated from that point, and the ME then attempts to visually recreate the new MLR.

This might be true up to a certain point. But I think that the MLR, as you propose it here, can then be influenced by changing your tactics. The MLR would be re-calculated and might differ in another result then it was initially. This seemed to work fine from FM05 until FM08. Attacking by 1 down could result in a draw or win and defending a 3-2 could result in a just win.

How different is this with FM09, when I am 1 down with 15 minutes to go, I can attack all I want, change mentality, set up 3 or 4 attackers instead of my standard 2 strikers, letting everyone have forward runs including the goalie, closing down to the maximum etc. NOTHING WORKS! It is even worse, when I change something in tactics for the last 15 mins, I do not get any key highlights at all!!! So my game is doomed anyway, whether I change my tactics or not.

Therefore I tend to agree with Real Magpies that in FM09 the MLR is calculated beforehand and is not changed during the match. At least not accordingly to changing tactics.

This is frustrating, but can nonetheless be corrected in my view, because it worked fine in previous versions. I expect it to be good again with FM10.

I too believe that the match engine should be #1 priority (with training/development #2 and transfers/finances#3), but I also like the additional stuff (media etc)!

Anyway, the OP is really interesting. I agree that attacking play is kind of undeveloped still and Paul C should be working on it. I agree that defense often looks stupid. And I agree that possession football is lacking horribly (that's the hardest part anyway). But you are being a bit unfair imo, because the match engine is new. Why is it new? Because it simulates a 3D world now, so it had to be rewritten (much of it anyway). What does this mean? That's it is in its childhood now and it's only going to get more mature.

Do you remember CM4? It was SI's first effort on 2D and it was amazingly bad. Eventually it got there, and FM08 had a very good 2D engine. To be frank, I was expecting a similar disaster with the first 3D engine (simlar to CM4) but fortunately things are much better now. FM09's engine is a good starting point, I think.

So hopefully SI will slowly incorporate all the things you are asking for, and will also take care of what McMahon's Lover said about goal vs midfield action. You hit the nail on the head a few times in your post (tackling, shooting etc). Real football at top level is mostly Strategy not Tactics, but FM has a hard time simulating strategy. We are being given a million options on Tactics (Run, Shoot Long etc for EVERY player, count the number of sliders!), but very few strategic ones (width, depth, tempo, focus passing, free role, hold up ball etc). It's not that I don't know how to do it, it's that the few (strategy) options won't let me do it. I don't think that overall strategy should be decided by the sum of our tactical decisions (e.g. you tell the guy to never cross, so he never goes wide). Rather, tactics should be the slave and strategy should be the master.

You have a strong case. Hopefully, we are going to read more from you around here.

PS. Real Magpies, no, that was the CM3 engine.

But in general I agree with Lyssien, CM4 was hell to play, but I worked my way through with it for I think half a year of 'FM fun'. How different is it with FM09, it feels like hell, but has some soft touches to it that makes it worth play it and still experience the FM fun (without the dubble quotes!). Playing lower leagues is my favorite, because ME flaws can be due to lower leauge playing abilities :-)

So, this year is a year of patience, grinding teeth and bad sleeping (and nightmares) all whilst hoping that next year will be more than great.

Merging the two posts in one final conclusion leaves tactics and strategy about equal to eachother. Where tactics can be trained and strategy is in the eye of the manager. And we all know that managers are nothing without players who cohere and stick to tactics ;-)