Boring, Boring Crossing Manager Simulator

[Mikke]

7 minutes ago, MBarbaric said:

I feel like Dave has a point here. It isn't about tactics, there is an excess of numbers of crosses per match in FM for at least three versions (just like the number of offsides). I am not talking about human teams, but AI.

If you take Barcelona as an example, we can all agree they play with very aggressive full backs. They move forward and hug the line in order to stretch the opposition in the attacking third. They aren't there to cross but to provide more space for the winger. In FM though, these aggressive full backs are inherently destined to cross. So far this season Barcelona has the least number of crosses per match in La Liga (7, average for league is 13). I don't have FM 18 but it would be interesting if someone went into team stats in his save game to see the average crosses per match that ME produces.

Is that completed crosses or attempted?

[Svenc]

16 minutes ago, MBarbaric said:

I feel like Dave has a point here. It isn't about tactics, there is an excess of numbers of crosses per match in FM for at least three versions (just like the number of offsides). I am not talking about human teams, but AI.

If you take Barcelona as an example, we can all agree they play with very aggressive full backs. They move forward and hug the line in order to stretch the opposition in the attacking third. They aren't there to cross but to provide more space for the winger. In FM though, these aggressive full backs are inherently destined to cross. So far this season Barcelona has the least number of crosses per match in La Liga (7, average for league is 13). I don't have FM 18 but it would be interesting if someone went into team stats in his save game to see the average crosses per match that ME produces.

There has also (and will never be) an instruction that translate to NEVER do "x", which influences too. That said, whilst that's never been something I focused on (I'd take a wide/crossfield ball successfully connecting over a central/no angled one on one from a no angled through ball every time though, at least in terms of how likely it justified is a goal )I just tried it and had a match of 70% possession, mostly in the opposition half/final third with about 20+ crosses attempted (8 of which corners, about 4-5 of each from the backs still from open play apparently).  Narrow formation, wb/s (crossed on "mixed", further reduced thus by "work ball into box",  thingamambob. Plus similar to the FM 2012 video, 2 guys centrally sitting deeper to stretch the central space rather than choking it. I'd be careful with purely numbers, naturally. Stats come off play, not the other way around (there is still a bad theory that the engine would create arbitrary sequences of play to match an event / even a result, when that is the exact opposite of what's going on). 

That all, naturally is still a viable concern. I'd be surprised if SI would't be much aware of it, but recent and former history suggests major overhauls won't come until the next major ME rewrite. I second the stance though taken by the mods that somebody spreading this as some kind of game breaker and going with hyperbole is obviously taking things a tad far, at least by FM 2018. [Even though everybody should be allowed to dislike a game based on it... I personally considered the wide midfielder stuff on FM 17 to be infinitely, ridiculously more serious, as flanks are but flanks, and central spaces are central spaces, shortest distance goalside and all* but to each his own!]. Such spreads also completely wrong ideas about how the game works too in the end, as usual.

 

A few great points have been brought up though. Namely the UI being a few opaque, the options having limitations, and as such players oft finding it reasonably diffifcult to somewhat translate their wishes into the UI. If that isn't straight-forward for human players, how could it be for AI?

[MBarbaric]

11 minutes ago, Mikke said:

Is that completed crosses or attempted?

In short, attempted.

Longer answer.

AvCrosses per match                                     attempted/completed

The most crosses p/m Real Madrid                21/7 (31%)

La Liga Average                                                  13/3,5 (27%)

Least crosses  p/m Barcelona                          7/1,8 (26)

9 minutes ago, Svenc said:

There has also (and will never be) an instruction that translate to NEVER do "x", which influences too. That said, whilst that's never been something I focused on (I'd take a wide/crossfield ball successfully connecting over a central/no angled one on one from a no angled through ball every time though, at least in terms of how likely it justified is a goal )I just tried it and had a match of 70% possession, mostly in the opposition half/final third with about 20-25 crosses (8 of which corners).

I am not saying it isn't possible to create relatively plausable scenarios. However, it frequently takes more advanced user, not an average player and certainly not the AI.

[Mikke]

1 minute ago, MBarbaric said:

In short, attempted.

Longer answer.

AvCrosses per match                                     attempted/completed

The most crosses p/m Real Madrid                21/7 (31%)

La Liga Average                                                  13/3,5 (27%)

Least crosses  p/m Barcelona                          7/1,8 (26)

 

After 8 games of La Liga in FM18:

The most crosses p/m Real Madrid, completed crosses 9.6 p/m (31% of attempts)
Barcelona, completed crosses 7.5 p/m (29% of attempts)
Least crosses p/m Girona, completed crosses 4.0 p/m (22% of attempts)

So yes, more crosses in FM than IRL, but actually not as much more as I would have thought. Less difference between the two extremes, also. Barcelona crossing 4x more in FM compared to real data....

[HUNT3R]

50 minutes ago, Dave1990 said:

 

I don't agree, this is exactly the point. Because, as I said before, on FM just it's enough put a offensive full-back and maybe "look for overlap" and you easily abuse crosses. In reality, however, with similar instructions there are certainly crosses, but in a much more normal way.. And repeating what I said before, there are no half-measures in the game. With "look for overlap" I would like a constant thrust of full-backs and obviously crosses, but not 37 per games, much less. Another user showed that he has an average of 57 crosses per game, which with any kind of instruction, remains totally unrealistic. That is why I complain essentially with ME.

You are narrow and your choice of role and duties will make everyone bunch up in the middle. Your choice of Attack fullback on the right and the addition of Look For Overlap will ask the other fullback to also get forward more and cross more. You're specifically asking your fullbacks to bomb forward and whip in crosses. You're asking most of your midfield to bunch up with the attackers, which will bunch up the defenders against you and drastically reduce proper passing options. The only real clear option is the wide fullbacks as everyone else is quite central and narrow. Your choice of TIs are all aiding this. You're selecting narrower, shorter, slower passing which will mean by the time that ball gets into advanced areas, most of your players are in advanced positions all bunched up. There's very little use of space here. Whether you're doing well or not isn't the point. The point that was made is that you are complaining about "crossing simulator" when it seems that you've specifically set up to do so.

[MBarbaric]

9 minutes ago, Mikke said:

After 8 games of La Liga in FM18:

The most crosses p/m Real Madrid, completed crosses 9.6 p/m (31% of attempts)
Barcelona, completed crosses 7.5 p/m (29% of attempts)
Least crosses p/m Girona, completed crosses 4.0 p/m (22% of attempts)

So yes, more crosses in FM than IRL, but actually not as much more as I would have thought. Less difference between the two extremes, also. Barcelona crossing 4x more in FM compared to real data....

What you should really look at are the number of cross attempts per match rather than completion rate. I wouldn't argue the completion rate or number of goals scored from crosses are wrong. What appears as a problem is the sheer number of crosses taken in the ME. 

I feel the mods (wrongly) concentrate on human tactics here. It is the AI that should be looked at in order to determine if this is "a crossing simulator" or human player creating a cross friendly tactics. If it is indeed problem with AI teams, then "crossing simulator" might have a case. If it is only the human, then it certainly isn't a crossing simulator but a human factor

[Mikke]

Just now, MBarbaric said:

What you should really look at are the number of cross attempts per match rather than completion rate. I wouldn't argue the completion rate or number of goals scored from crosses are wrong. What appears as a problem is the sheer number of crosses taken in the ME.

I provided the amount of completed crosses, and also the completion rate. Everyone can calculate the number of cross attempts from there, so I didn't bother.

Also, the completion rate is very close to real life statistics, so just the fact there are more completed crosses should be enough.

[herne79]

7 minutes ago, MBarbaric said:

I feel the mods (wrongly) concentrate on human tactics here.

Possibly.  However reading the opening post it is purely about the OP's own tactical system, nothing about how the rest of the AI clubs are getting on.

As the OP talks about their own system, it's pretty much the only thing we have to go on so I'll point out the short comings in the system to show that the reasons have little to do with the game being at fault in that particular instance.

If, on the other hand, anybody has conducted hard and fast soak tests over multiple saves and seasons that can demonstrate the number of crosses attempted/made is way over the top compared to real life, the best thing to do is open a report in the bugs forum with all the data and let SI take a look.  At the moment all we really have is one tactical system that is more flawed than the OP realises (in terms of promoting crosses) and a lot of conjecture.

[MBarbaric]

24 minutes ago, Mikke said:

I provided the amount of completed crosses, and also the completion rate. Everyone can calculate the number of cross attempts from there, so I didn't bother.

Also, the completion rate is very close to real life statistics, so just the fact there are more completed crosses should be enough.

Well not really, 

While we have the best and the worst team we can get sort of average for the league:

av Crosses p/m         attempts/completed (%)

La Liga FM                             18/6,8            (27%)
La Liga                                   13/3,5             (27%)
 

If I've calculated well (quite possible I didn't so correct me if I am wrong), there's around 50% increase in average number of cross attempts per match in your save. This might suggest the ME produces more crosses than it really should and by a big amount. However, as Herne says, someone should really look at numbers over multiple saves to get any reliable conclusion over this debate. There's also a question of what game stats consider as a cross and what does any outside source consider as a cross.  So, basically, only SI can test this in any conclusive manner. 

[themadsheep2001]

11 minutes ago, herne79 said:

Possibly.  However reading the opening post it is purely about the OP's own tactical system, nothing about how the rest of the AI clubs are getting on.

As the OP talks about their own system, it's pretty much the only thing we have to go on so I'll point out the short comings in the system to show that the reasons have little to do with the game being at fault in that particular instance.

If, on the other hand, anybody has conducted hard and fast soak tests over multiple saves and seasons that can demonstrate the number of crosses attempted/made is way over the top compared to real life, the best thing to do is open a report in the bugs forum with all the data and let SI take a look.  At the moment all we really have is one tactical system that is more flawed than the OP realises (in terms of promoting crosses) and a lot of conjecture.

This. I'm not sure what MBarbaric wants us to look at here, when the current hypothesis set out by the OP are based off his own flawed setup. 

With the absence of multiple soak tests to go along with qualitative results from a well balanced setup, there's actually precious little else to look at 

[Mitja]

maybe real life La Liga teams all play with ''work the ball into the box'' ticked on? 

as being mentioned on fm17 even 100 crosses per match can be achieved.

 

[MBarbaric]

16 minutes ago, themadsheep2001 said:

I'm not sure what MBarbaric wants us to look at here

Well, people have complained for years about too many crosses. If you really want to add to a conversation you don't just feed the debate/troll with "it's your tactics".  Unless somebody gets into numbers behind AI the question will always come back and the myth will preserve. If, on the other hand, somebody does the math and shows that numbers fall reasonably inside real stats, then it is indeed, human tactics.

Sadly, I don't have new FM but it shouldn't be difficult to load a league, run it for 10 years, repeat it 10 times and get averages to compare with real stats.

[HUNT3R]

14 minutes ago, MBarbaric said:

If you really want to add to a conversation you don't just feed the debate/troll with "it's your tactics".

No one did. There was proper explanations for why it produces what it does and if you read back, we're all saying the same thing. We don't have anything else from the OP as he posted his experience and his tactic.

[tkg]

I've been collecting some data as part of a bug report I'm writing up at the moment that could be of interest here.

I've ran 5 simulations of the Premier League upto match week 11, 3 of which i've collated the data for so far:

The total number of crosses in the simulations have been:

• 6639 total crosses with a 11.0% success rate
•  6660 total crosses with a 10.6% success rate
• 6748 total crosses with a 10.2% success rate.

In real life there had been 4233 attempted crosses with a success rate of 22% at the end of match week 11.

Now obviously there are some slight issues with comparing those stats directly as corners and free kicks appear to be considered cross attempts in FM. I haven't got round to taking that into account yet but I suspect it should bring the number of crosses IRL in line with the match engine. What does seem off however is cross completion rates. it's currently 1/2 the expected value in real life and if we take into account corners then I'd actually expect the average values in the game to be higher than real life, as corners are typically more accurate that crosses.

[themadsheep2001]

20 minutes ago, MBarbaric said:

Well, people have complained for years about too many crosses. If you really want to add to a conversation you don't just feed the debate/troll with "it's your tactics".  Unless somebody gets into numbers behind AI the question will always come back and the myth will preserve. If, on the other hand, somebody does the math and shows that numbers fall reasonably inside real stats, then it is indeed, human tactics.

Sadly, I don't have new FM but it shouldn't be difficult to load a league, run it for 10 years, repeat it 10 times and get averages to compare with real stats.

We have one set of information to work with right now. And based on what has been offered from it, it is tactical in this case, with multiple proper responses given as to why if you read the thread from the start. Which makes any general conclusion from it flawed. 

[themadsheep2001]

4 minutes ago, tkg said:

I've been collecting some data as part of a bug report I'm writing up at the moment that could be of interest here.

I've ran 5 simulations of the Premier League upto match week 11, 3 of which i've collated the data for so far:

The total number of crosses in the simulations have been:

• 6639 total crosses with a 11.0% success rate
•  6660 total crosses with a 10.6% success rate
• 6748 total crosses with a 10.2% success rate.

In real life there had been 4233 attempted crosses with a success rate of 22% at the end of match week 11.

Now obviously there are some slight issues with comparing those stats directly as corners and free kicks appear to be considered cross attempts in FM. I haven't got round to taking that into account yet but I suspect it should bring the number of crosses IRL in line with the match engine. What does seem off however is cross completion rates. it's currently 1/2 the expected value in real life and if we take into account corners then I'd actually expect the average values in the game to be higher than real life, as corners are typically more accurate that crosses.

That's a good start. Be good to see what the various other leagues look like 

Next step would be looking at individual cross examples (how and why they aren't being completed)

[tkg]

Yeah totally, that's the plan. 

At the moment my suspicion is that the flight of the ball is too often moving forward from the position the cross was taken, as opposed to backwards or perpendicular. There is potentially  a lack of curl to the flight of the ball too. I think there may also be an area of the pitch from which these failed crosses are more likely to occur (roughly from the 6 yard to penalty spot) in terms of vertical distance. If the player gets tight to the byline then they perform a cross more akin to a cut back which actually seems to work quite well.

[enigmatic]

1 hour ago, tkg said:

Yeah totally, that's the plan. 

At the moment my suspicion is that the flight of the ball is too often moving forward from the position the cross was taken, as opposed to backwards or perpendicular. There is potentially  a lack of curl to the flight of the ball too. I think there may also be an area of the pitch from which these failed crosses are more likely to occur (roughly from the 6 yard to penalty spot) in terms of vertical distance. If the player gets tight to the byline then they perform a cross more akin to a cut back which actually seems to work quite well.

This would also be a partial explanation of the apparently absurd number of crosses which goalkeepers end up saving (part of this is down to sometimes ludicrously poor decision making by the goalkeeper and probably part down to football fans underestimating how often it happens IRL)

-

I suspect cross numbers may be a little high due to fullbacks in recent FM versions apparently being coded to sit a little narrower than real life counterparts to counter the fact neither AI players nor AI managers are quite smart enough to deal with straightforward overloads of the penalty area in a way humans would. Though I was under the impression this is something FM18 might have reworked a little with all those new roles using underlaps and channels...

 

[Dave1990]

4 hours ago, HUNT3R said:

You are narrow and your choice of role and duties will make everyone bunch up in the middle. Your choice of Attack fullback on the right and the addition of Look For Overlap will ask the other fullback to also get forward more and cross more. You're specifically asking your fullbacks to bomb forward and whip in crosses. You're asking most of your midfield to bunch up with the attackers, which will bunch up the defenders against you and drastically reduce proper passing options. The only real clear option is the wide fullbacks as everyone else is quite central and narrow. Your choice of TIs are all aiding this. You're selecting narrower, shorter, slower passing which will mean by the time that ball gets into advanced areas, most of your players are in advanced positions all bunched up. There's very little use of space here. Whether you're doing well or not isn't the point. The point that was made is that you are complaining about "crossing simulator" when it seems that you've specifically set up to do so.

I didn't want to come to a tactical discussion, but since you are dancing, let's dance! 
That the Christmas Tree has a choke of central spaces is intrinsic to the disposition on the pitch. So rightly I expect to suffer especially on counterattacks and long passes, especially against two strikers (one takes the space wide and the other slits in the area) and is one of the reasons why I have several goals on crosses. 
But in the offensive phase I don't expect to "bombard" with crosses, I expect a good number of crosses yes, but why a bombardment? I expect a bombardment with tactics with two wingers and two complete wing-backs behind them, for exemple either 4-4-2 or 4-3-3 or 4-2-3-1 or 4-2-4. And choosing "wipped crosses" I'm not choosing to cross more, but I'm choosing a particular cross type.
The midfield and trequartist roles are designed to create continuous ball overlays and movements. The CF / S should open the spaces at the SS / A and behind this there is the BBM that offers off the ball movements and good coverage (although it would be better the carrillero, but it moves too little vertically). On the left side there is a normal AM / S that is often toward the center of the field, almost a "fake playmaker", that release the space for the offensive mezzala. The Wing-backs as I said before are set to give width, the one set to support on the side of the offensive mezzala is slightly more cautious, the one set to attack on the side of the BBM that covers slightly more the midfield. It remains a tactic that seeks to impose ball possession (with FM17 I easily reached an average of 63%, this year possession is more difficult and I have registred about 58%). And like all offensive tactics, some defensive-looking difficulties should be considered, especially when you don't use top clubs. In short, I don't think that I giving so aggressive input on crosses with this tactic, but apparently we have different views, is in the nature of things.
Now I will do another test with a different tactic and without using "look for overlap" and Wing-backs, let's see what comes out.

[BigRoboCrouch]

This thread is praiseworthy, but ultimately useless. 

SI just isn't going to change a match engine that makes them millions every year, no matter how much data you can submit to them.

[MHovel]

1 hour ago, BigRoboCrouch said:

This thread is praiseworthy, but ultimately useless. 

SI just isn't going to change a match engine that makes them millions every year, no matter how much data you can submit to them.

I agree, they won't bother much because football is a game which is very popular, they won't care about what users in the forums will say about the ME.

The game is probably selling after all, not because it's that good, it's because football as a whole is a sport which is maybe the most popular and there will be always demand for football pc games, 2nd is because the lack of competition.

I don't understand people who buy the game and being disappointed of the crossing issue.

incase you had fm17, you probably witnessed the crossing issue. Now, you bought FM18 and you witnessed this issue not being fixed, you have 2 options:

1. Refund the game (which is what I did).

2. Accept the fact that SI cannot fix this issue and keep playing.

I came to the conclusion that there is no point to mention this on the forum anymore (the crossing issues), it's not worth to put an energy on it and provide your feedbacks about the issue.

[Rashidi]

11 hours ago, themadsheep2001 said:

I'm not home so it will have to wait but I'll hopefully demonstrate creating width without an overreliance on crossing this week. Failing that @Rashidi might already have something up his sleeve 

Sit narrower cut inside with ball, and tell your players to play with lower width and exploit the middle?  That's what I do when I don't want crossing. Depends on whether you have attacking roles punching through the middle to work. And I don't even understand the point of this thread. The tactic is blatantly focused on going down the flanks..and thats it.

[pjrydo]

On 11/20/2017 at 06:31, lemeuresnew said:

now, im not defending the game in any way. this is the worst released fm game for at least 5 years in my opinion, and i will not be purchasing the game in advance EVER again. if a demo impresses me i will buy the full, sure. but the days of fm getting my money blindly are over. but if you have a problem with those number of goals, there are at lkeast 7 team instructions you can remove to make it more to your liking mate

Agreed. Most disappointingly under-cooked release in many, many years

[Svenc]

I think there is a few angles to this.

1) Aforementioned tactics. Obviously influence.

2) AI. AI is too limited either way. I am shocked to see AI duty allocations on this right at the very top that by FM 2012 were classed a "bug" --- and made PaulC state then when "fixed" come FM 2013 on the forums it may make the game a few more difficult all by itself. Already posted in GD. Hopefully isolated stuff. If AI can't handle duty, it sure can't handle anything even approaching holistic systems.

3) Feedback/Statistics. And their definition. In one of the above graphics, Barcelona are ranked as averaging 7 crosses per match. However, they also average up to 7 corners per match. So either the likes of Dani Alves never ever make/made a cross from open play (which is highly doubtful), or it is mainly FM that counts every single corner towards the cross count too. Speaking of which, the corner count averages on FM could easily be higher in real football too. I know guys that averaged 15-20 corners alone per match as recent as FM 17. This is basically a corner for every 3rd minute the ball is effectively in play. Which required tactics however that no side in football plays (compressing the attacking space equally to the size of a tuna fish can going forward, also very popular on dowload sides naturally that try to "game" the engine, rather than let their sides just play "realistic" football. It's not just corners. There are events flagged crosses from my stance if you click through the dots that arguably are but simple cutback passes from out wide into the box.

4) Balancing/Effectiveness of crossing. Numbers are but numbers. There are actually way more suspicious in-game from my end, which also may be highly relevant as to the consistency of top sides (or lack thereof). What's arguably more important in the grander scheme of things is that the balance is right. Traditionally, in particular on older releases, the amount of "tackles" per match was ridiculously higher on FM in real football too. However, on FM's level of match play it was required for the defensive phase to be efficient/balanced, which related to a limitation as to defensive engagement/movement/anticipation, of which there is still some left, some of which may relate to the numbrs of crosses going off too (a cross after all, requires a player to be in some space). Speaking about balance, some "crossing" frustration expressed from my experience seems to relate to that crossfield balls, when they successfully connect (most should be wasted) are perceived to be too efficient in the scheme of things. "Crossing manager" is not seldom short-hand for the frustration that such a move successfully pulled off  seems more likely to result into a goal than certain one on ones off through balls.  From my end, that's justified. This is laws of physics. A cross field pass successfully connecting, oft from the first touch, instantly changes the direction of the ball, turning that "one on one" rather in a "one to close to zero", as the keeper has limited chance of reacting. Compare that to a one on one from a narrow/central through ball at no angle. Not only can the keeper fully anticipate, and the forward oft has the ball played into his back rather than his running path. The forward will also find it hard to impossible to get any angle off that could be considered to actually handicap the keeper.

5) UI. Does it specifically state anywhere in-game that "work into box" could significantly reduce the amount of crosses per match? In its first inception, I think it even mainly reduced the amount of long shots and it's been reworked since in subsequent iterations. You only get this knowledge from following the tactical community. Not ideal.

I think anybody is on the same side here -- ultimately, everybody wants this to be a more realistic core sim, really. However, it's a bit more nuanced than just filing "crossing manager" every season, and then bumping some numbers. In particular on FM's level of numbers, which  have no definition whatsoever.

[HUNT3R]

8 hours ago, Dave1990 said:

I didn't want to come to a tactical discussion, but since you are dancing, let's dance! 
That the Christmas Tree has a choke of central spaces is intrinsic to the disposition on the pitch. So rightly I expect to suffer especially on counterattacks and long passes, especially against two strikers (one takes the space wide and the other slits in the area) and is one of the reasons why I have several goals on crosses. 
But in the offensive phase I don't expect to "bombard" with crosses, I expect a good number of crosses yes, but why a bombardment? I expect a bombardment with tactics with two wingers and two complete wing-backs behind them, for exemple either 4-4-2 or 4-3-3 or 4-2-3-1 or 4-2-4. And choosing "wipped crosses" I'm not choosing to cross more, but I'm choosing a particular cross type.

No dancing needed. You're setting up to cross as several people said already. I'll put it in simpler terms for you:

FBR - Wingback Attack   ---  Cross More Often by default.

FBL - Wingback Support  --- Cross Sometimes by default + Look for Overlap = Cross More Often.

So, you're specifically telling one of the fullbacks to cross a lot. You're also reducing passing range and their dribbling options, so IIRC the Wingback Support will now not dribble rarely.

 

You have the Mezzalla and BBM joining the front 3, giving you 5 players up top and compressing play. It's especially happening because your build-up play is slower and shorter so you're giving the team time to all move/run forward. Your FBs won't then have too many passing options and they'll be left with the option to cross, hence bombardment of crosses. 

[Svenc]

6 minutes ago, HUNT3R said:

 

FBL - Wingback Support  --- Cross Sometimes by default + Look for Overlap = Cross More Often.

 

 

Similar to work ball into box specifically affecting cross frequency, this is also isn't stated anywhere in-game, btw. It's new to me too actually, likely one of the reworks only communicated to mods by SI. The original inception simply encouraged backs to surge forward ASAP plus made the wide guys in front of them more cautious in their forward runs plus holding the ball so that they may likely feed the eventually overlapping run. There's a whole host of topics to tackle here, and this is one of them (in terms of AI, naturally, too).

[HUNT3R]

5 minutes ago, Svenc said:

Similar to work ball into box specifically affecting cross frequency, this is also isn't stated anywhere in-game, btw. It's new to me too actually, likely one of the reworks only communicated to mods by SI. The original inception simply encouraged backs to surge forward ASAP plus made the wide guys in front of them more cautious in their forward runs plus holding the ball so that they may likely feed the eventually overlapping run. There's a whole host of topics to tackle here, and this is one of them (in terms of AI, naturally, too).

Might be my imagination, actually. Thought it was in Lines and Diamonds, but it's not listed as changing crossing tendencies (https://community.sigames.com/topic/345094-lines-and-diamonds-the-tacticians-handbook-for-football-manager-2015/?do=findComment&comment=10322561). Will scratch that entry, although the rest of the points I made will remain the same. My apologies.

[Svenc]

5 minutes ago, HUNT3R said:

Might be my imagination, actually. Thought it was in Lines and Diamonds, but it's not listed as changing crossing tendencies (https://community.sigames.com/topic/345094-lines-and-diamonds-the-tacticians-handbook-for-football-manager-2015/?do=findComment&comment=10322561). Will scratch that entry, although the rest of the points I made will remain the same. My apologies.

NP. Wasn't actualyl questioning if it was legit, but curious -- stuff had been reworked multiple times before, which typically is only communicated from my experience after the Beta stage by mods --- if you happen to be around the relevent threads. The online manual states the same here (work ball into box too) as the in-game tooltips if you hover the mouse over the instruction. Nowhere at all does it state that work ball into box would reduce crossing frequency as such (and as said, AFAIK initially it only affected ranged attempts, but that could be memory playing tricks too). The important thing is,  work ball into box heavily affects crosses either way by now.

[Mitja]

3 hours ago, Svenc said:

The important thing is,  work ball into box heavily affects crosses either way by now.

for long time. but it does make sense?

AI uses it alot too.

[Dave1990]

5 hours ago, HUNT3R said:

No dancing needed. You're setting up to cross as several people said already. I'll put it in simpler terms for you:

FBR - Wingback Attack   ---  Cross More Often by default.

FBL - Wingback Support  --- Cross Sometimes by default + Look for Overlap = Cross More Often.

So, you're specifically telling one of the fullbacks to cross a lot. You're also reducing passing range and their dribbling options, so IIRC the Wingback Support will now not dribble rarely.

 

You have the Mezzalla and BBM joining the front 3, giving you 5 players up top and compressing play. It's especially happening because your build-up play is slower and shorter so you're giving the team time to all move/run forward. Your FBs won't then have too many passing options and they'll be left with the option to cross, hence bombardment of crosses. 

The slow tempo is to develop a game of short passes and possession of the ball more easily, and when the action developed in central spaces, I scored several goals with penetrative passes and through ball.

I keep thinking that cross "bombardment" isn't given by my instructions, but it is simply the ME set in this way. Because in these years, I have NEVER seen a tactic that, with the right instructions to do so, has bombarded the IA with penetrative passes and through balls. Why if do I set up some players with "more risky passes" and others players who attack the space have I never a bombardment of penetrative passes? Strangely, this always happens when we put "look for overlap" and offensive full-backs. 
Simply because the game succeeds in having a balance when exploiting the central spaces, balance that disappears whenever you want to exploit the flanks. And this, for me, is a fact for 3 years.
 

[HUNT3R]

2 minutes ago, Dave1990 said:

Simply because the game succeeds in having a balance when exploiting the central spaces, balance that disappears whenever you want to exploit the flanks. And this, for me, is a fact for 3 years.
 

No. YOU don't have any balance AT ALL in this tactic. It's 100% geared toward flank play and crosses. You've had multiple explanations as to why this is.

[herne79]

43 minutes ago, Dave1990 said:

I keep thinking that cross "bombardment" isn't given by my instructions, but it is simply the ME set in this way. Because in these years, I have NEVER seen a tactic that, with the right instructions to do so, has bombarded the IA with penetrative passes and through balls.

Sorry dave, just because you haven't seen it doesn't mean it doesn't happen.

The first picture is from my current FM18 save, about 65% (two thirds) of my goal assists are from central areas.  I play a 4123DM wide.

The second picture is from FM17.  That's about 57% of assists from central areas using a 4-4-2 narrow diamond with 2 x attacking fullbacks.

Hardly "crossing manager" due to the ME being set that way.

Is there a wider issue with the AI?  I don't know and tkg above seems to be doing some interesting analysis in that area.  But you keep banging on about your own tactic which you fail to accept is flawed from a crossing perspective and offer nothing further than pure conjecture.

[herne79]

43 minutes ago, Dave1990 said:

I keep thinking that cross "bombardment" isn't given by my instructions, but it is simply the ME set in this way. Because in these years, I have NEVER seen a tactic that, with the right instructions to do so, has bombarded the IA with penetrative passes and through balls.

Sorry dave, just because you haven't seen it doesn't mean it doesn't happen.

The first picture is from my current FM18 save, about 65% (two thirds) of my goal assists are from central areas.  I play a 4123DM wide.

The second picture is from FM17.  That's about 57% of assists from central areas using a 4-4-2 narrow diamond with 2 x attacking fullbacks.

Hardly "crossing manager" due to the ME being set that way.

Is there a wider issue with the AI?  I don't know and tkg above seems to be doing some interesting analysis in that area.  But you keep banging on about your own tactic which you fail to accept is flawed from a crossing perspective and offer nothing further than pure conjecture.

[Dave1990]

6 hours ago, HUNT3R said:

No. YOU don't have any balance AT ALL in this tactic. It's 100% geared toward flank play and crosses. You've had multiple explanations as to why this is.

I didn't talk about my tactic, I was talk in general. In a system properly set up to "bombarding" the central part, I will never have an equal number of assists with a through ball to what I got from the crosses with my tactic.

 

5 hours ago, herne79 said:

Sorry dave, just because you haven't seen it doesn't mean it doesn't happen.

The first picture is from my current FM18 save, about 65% (two thirds) of my goal assists are from central areas.  I play a 4123DM wide.

The second picture is from FM17.  That's about 57% of assists from central areas using a 4-4-2 narrow diamond with 2 x attacking fullbacks.

Hardly "crossing manager" due to the ME being set that way.

Is there a wider issue with the AI?  I don't know and tkg above seems to be doing some interesting analysis in that area.  But you keep banging on about your own tactic which you fail to accept is flawed from a crossing perspective and offer nothing further than pure conjecture.

It's not a conjecture born out of nowhere, I see the same things for three years. Am I wrong? It's possible. But I don't seem to be the only one who has been complaining of too many crosses in recent years.
The assist data that you have posted are "fake", and I explain why. FM considers as assist a goal-keeper punch and then put on net by another player. Example: Sterling crosses for Aguero, Aguero's shot is punched, Gabriel Jesus arrives and marks. The goal is born from a cross, but the game gives assist to Aguero and considers it as assist from inside the area. And this happens often, as it often happens that from a cross there is a flank play inside the area for a third player to score. Here too the action is born from a cross, but the assist is inside the area for the game. And then I often see crosses coming from inside the penalty area, that's why I don't rely on FM data, but only on my eyes.

[herne79]

Not sure which emoji to use after that.  ?  ?  ?

Think I'll just settle for .

I'm out.

[enigmatic]

49 minutes ago, Dave1990 said:

The assist data that you have posted are "fake", and I explain why. FM considers as assist a goal-keeper punch and then put on net by another player. Example: Sterling crosses for Aguero, Aguero's shot is punched, Gabriel Jesus arrives and marks. The goal is born from a cross, but the game gives assist to Aguero and considers it as assist from inside the area. And this happens often, as it often happens that from a cross there is a flank play inside the area for a third player to score. Here too the action is born from a cross, but the assist is inside the area for the game. And then I often see crosses coming from inside the penalty area, that's why I don't rely on FM data, but only on my eyes.

This is exactly how the real life assist data you're insisting FM should resemble works...

[bennytee]

This is like Brexit all over again...........

[themadsheep2001]

1 hour ago, enigmatic said:

This is exactly how the real life assist data you're insisting FM should resemble works...

Yep. On that note @tkg I'm really interested in what you'll be putting in the Bugs section. It'll be more useful than the rest of the information in this thread. Will run some soaks too. I've already started looking at crossing accuracy and delivery in my saves as bug reports. Not sure much else from this thread will be useful 

[Cougar2010]

57 minutes ago, Dave1990 said:

I didn't talk about my tactic, I was talk in general. In a system properly set up to "bombarding" the central part, I will never have an equal number of assists with a through ball to what I got from the crosses with my tactic.

 

It's not a conjecture born out of nowhere, I see the same things for three years. Am I wrong? It's possible. But I don't seem to be the only one who has been complaining of too many crosses in recent years.
The assist data that you have posted are "fake", and I explain why. FM considers as assist a goal-keeper punch and then put on net by another player. Example: Sterling crosses for Aguero, Aguero's shot is punched, Gabriel Jesus arrives and marks. The goal is born from a cross, but the game gives assist to Aguero and considers it as assist from inside the area. And this happens often, as it often happens that from a cross there is a flank play inside the area for a third player to score. Here too the action is born from a cross, but the assist is inside the area for the game. And then I often see crosses coming from inside the penalty area, that's why I don't rely on FM data, but only on my eyes.

 

So basically what you are saying is that every goal ever scored is as a result of a cross, right?

[Dave1990]

1 hour ago, enigmatic said:

This is exactly how the real life assist data you're insisting FM should resemble works...

Why mix things up? In IRL a shot that turns to goal after a goalkeeper's punch isn't assist, in the game instead it is. I should laugh at this that has never been corrected, but let's go ahead... An assist in the game is also the touch that precedes a own-goal, that's why the game assist data has no value.
The other cross reference is more general. It is obvious that a flank play in the area is considered rightly as assist made in the area, but the incidence of the cross isn't considered in the action that is there. This doesn't finish in game statistics as a goal born from cross, though in fact it's just this way. 

 

1 hour ago, Cougar2010 said:

 

So basically what you are saying is that every goal ever scored is as a result of a cross, right?

No, I've already explained it. But the problem was my tactic, I'm sure that with a tactic that doesn't do a crossing bombardment everything resolves and becomes normal.......

[Mikke]

21 minutes ago, Dave1990 said:

Why mix things up? In IRL a shot that turns to goal after a goalkeeper's punch isn't assist, in the game instead it is. I should laugh at this that has never been corrected, but let's go ahead... An assist in the game is also the touch that precedes a own-goal, that's why the game assist data has no value.
The other cross reference is more general. It is obvious that a flank play in the area is considered rightly as assist made in the area, but the incidence of the cross isn't considered in the action that is there. This doesn't finish in game statistics as a goal born from cross, though in fact it's just this way.

And when my two strikers get a 2 vs 0 counter attack, which ends with one of them crossing the ball from the edge of penalty area to the other striker who volleys it in, that also counts as a cross in FM. And it bloody well shouldn't, because the the goal wasn't born from a cross, but from a long ball. Right?

[Dave1990]

53 minutes ago, Mikke said:

And when my two strikers get a 2 vs 0 counter attack, which ends with one of them crossing the ball from the edge of penalty area to the other striker who volleys it in, that also counts as a cross in FM. And it bloody well shouldn't, because the the goal wasn't born from a cross, but from a long ball. Right?

Absolutely not. A long pass from the back has a much less direct result than a cross. For example, if a player scores with a shot from the outside after that the defense has rejected a cross, this I doesn't consider it as a goal came from cross. We need to think about direct and indirect consequences.

[Mikke]

7 minutes ago, Dave1990 said:

Absolutely not. A long pass from the back has a much less direct result than a cross. For example, if a player scores with a shot from the outside after that the defense has rejected a cross, this I doesn't consider it as a goal came from cross. We need to think about direct and indirect consequences.

Are you seriously trying to claim, that in my example the scoring chance was mostly a direct consequence of the cross, more than a direct consequence of the long pass? No wonder you don't understand even your own tactics...

There was a clear cut scoring chance even before the cross, and the fact that my target man still decided to cross to my poacher is quite irrelevant. Except, of course, FM presents that as a goal from a cross in statistics.

[Dave1990]

2 hours ago, Mikke said:

Are you seriously trying to claim, that in my example the scoring chance was mostly a direct consequence of the cross, more than a direct consequence of the long pass? No wonder you don't understand even your own tactics...

There was a clear cut scoring chance even before the cross, and the fact that my target man still decided to cross to my poacher is quite irrelevant. Except, of course, FM presents that as a goal from a cross in statistics.

If your striker made a cross instead of shooting directly, it means there was not such a big and real chance to score without it, don't you think? 

And if you don't understand that the long pass is part of a transition phase while a cross is purely an offensive phase, it's not my problem... A long pass can only be considered for direct assist, otherwise there is always a next most important step to consider.
 

[Mikke]

50 minutes ago, Dave1990 said:

If your striker made a cross instead of shooting directly, it means there was not such a big and real chance to score without it, don't you think? 

And if you don't understand that the long pass is part of a transition phase while a cross is purely an offensive phase, it's not my problem... A long pass can only be considered for direct assist, otherwise there is always a next most important step to consider.
 

No, it doesn't mean that. There was a real chance to score, but this time he chose to pass instead of shooting, and luckily the other striker didn't miss his chance. I'm surprised that you still haven't noticed that players in FM do not always choose the most logical course of action.

And passing someone clean through on goal is part of an offensive phase in my books. But yeah, it's not my problem if you disagree... A cross can only be considered for direct assist, otherwise there is always a next most important step to consider. Sound familiar?

[Dave1990]

10 hours ago, Mikke said:

No, it doesn't mean that. There was a real chance to score, but this time he chose to pass instead of shooting, and luckily the other striker didn't miss his chance. I'm surprised that you still haven't noticed that players in FM do not always choose the most logical course of action.

And passing someone clean through on goal is part of an offensive phase in my books. But yeah, it's not my problem if you disagree... A cross can only be considered for direct assist, otherwise there is always a next most important step to consider. Sound familiar?

Yes, it sounds familiar like a badly turned around omelet... If in your books a long pass has the same incidence in offensive phase than a cross, you should update them.
But if you don't understand the difference between the transition phase and the offensive phase, this discussion doesn't make sense to exist.

[shadowraiden90]

to be honest this discussion doesnt make sense at all. countless people have given examples how that specific tactic and a lot of other people's tactics are actually set up to attack the wings. by attacking the wings your asking for more crosses than you normally would.

 

you then add in the fact FM counts even a 5 yard side pass accross goal as a cross which in real life it wouldnt be although it could be argued by definition of a "cross" it is a "cross" as it means to pass the ball across goal which even a 5 yard pass technically is.

 

doesnt matter if you pack your team to be central players having fullbacks quite aggresive and having so much space infront of them will lead a lot of open space down the wings which guess what your players will make use off.

 

only one role i know off that doesnt have players roam and that is the Engache as hes essentially just a static guy in the AMC spot at all times. even target men move out wide when their is space to attack as guess what they would in real life. 

you then have the aspect of box to box players who push forward INTO SPACE which guess what lots of space out wide.

[Mikke]

1 hour ago, Dave1990 said:

Yes, it sounds familiar like a badly turned around omelet... If in your books a long pass has the same incidence in offensive phase than a cross, you should update them.
But if you don't understand the difference between the transition phase and the offensive phase, this discussion doesn't make sense to exist.

In my example the team is already in the attacking phase, but I can see why it's difficult for you to understand when I didn't provide a pkm or a video of that play.

Of course, not all long passes have the same incidence in offensive phase than a cross, and vice versa.

[Dave1990]

22 minutes ago, shadowraiden90 said:

to be honest this discussion doesnt make sense at all. countless people have given examples how that specific tactic and a lot of other people's tactics are actually set up to attack the wings. by attacking the wings your asking for more crosses than you normally would.

 

you then add in the fact FM counts even a 5 yard side pass accross goal as a cross which in real life it wouldnt be although it could be argued by definition of a "cross" it is a "cross" as it means to pass the ball across goal which even a 5 yard pass technically is.

 

doesnt matter if you pack your team to be central players having fullbacks quite aggresive and having so much space infront of them will lead a lot of open space down the wings which guess what your players will make use off.

 

only one role i know off that doesnt have players roam and that is the Engache as hes essentially just a static guy in the AMC spot at all times. even target men move out wide when their is space to attack as guess what they would in real life. 

you then have the aspect of box to box players who push forward INTO SPACE which guess what lots of space out wide.

But we have already discussed this fact.

The tactic exploits the advances of the full-backs, of course. Make use of their crosses, undeniable. But a team that bases all of its game on the flanks in IRL  makes a much smaller crosses number than what we get instead in the game. If no one can understand this, I am not surprised that no one does anything to improve the ME.

Now I'm using a tactic with two normal full backs and two inverted wings, without "look for overlap". For now we are at 34 crosses per game, only 3 less than the tactic that exploits a crossing bombardment.....

[enigmatic]

13 minutes ago, Dave1990 said:

But we have already discussed this fact.

The tactic exploits the advances of the full-backs, of course. Make use of their crosses, undeniable. But a team that bases all of its game on the flanks in IRL  makes a much smaller crosses number than what we get instead in the game. If no one can understand this, I am not surprised that no one does anything to improve the ME.

Now I'm using a tactic with two normal full backs and two inverted wings, without "look for overlap". For now we are at 34 crosses per game, only 3 less than the tactic that exploits a crossing bombardment.....

But then, as pointed out, the gap is a lot smaller than you've suggested because SI considers corners and set piece to be crosses. (Not sure about pullbacks from inside the penalty area; OPTA doesn't count them as crosses though)

La Liga average crosses per game is about 37ish which is low by top 5 league standards, plus about 10 corners, plus a handful of crossed set pieces, maybe plus pullbacks too. So around 50ish crosses between the two teams playing.

Divide by a number less than two (because I bet you usually have the most possession by a fair margin) and you'd probably expect about 30 crosses (by SI definition) per game which really isn't anywhere near as different from the raw numbers as what you're suggesting...

 

I built a very dominant FM17 team that managed around 20 crosses per match including several corners. Though funnily enough, the teams I built that scored goals from crosses were a lot more attractive to watch...