FM Balance

Is it me or does it seem that the coding in the game in multiple areas lacks balance? Every year there seem to be a number of areas that are either too much or too less, with very little in between.

It seems to be the SI holy grail just to get some balance. Is this down to the core code itself? Is it difficult to regulate things to strike balance where we have to suffer extremes?

This to me is the regular theme when it comes to bugs, discrepencies and match engine flaws - striking the balance. Is it more difficult then people appreciate?

Some examples:

Long shots - Too many

Red cards - Too many

Striker goals - Too many

Hitting the woodwork - Too much

Raining in games - Too much

Penalties scored - Too many

Becoming a club legend - Too few

Injuries - Fluctuates massively each version! This year too few

Tackling - Too few

Closing down - Too little

Is it flaws in the way the game is coded that restricts the balance a lot of us all crave for!

Is it me or does it seem that the coding in the game in multiple areas lacks balance? Every year there seem to be a number of areas that are either too much or too less, with very little in between.

It seems to be the SI holy grail just to get some balance. Is this down to the core code itself? Is it difficult to regulate things to strike balance where we have to suffer extremes?

This to me is the regular theme when it comes to bugs, discrepencies and match engine flaws - striking the balance. Is it more difficult then people appreciate?

Some examples:

Long shots - Too many

Red cards - Too many

Striker goals - Too many

Hitting the woodwork - Too much

Raining in games - Too much

Penalties scored - Too many

Becoming a club legend - Too few

Injuries - Fluctuates massively each version! This year too few

Tackling - Too few

Closing down - Too little

Is it flaws in the way the game is coded that restricts the balance a lot of us all crave for!

I can undertsand what you mean. It seems for some reason one year a setting is off. The fix it in a patch or two. The next version it's too few. Don't know. Maybe legacy code.

Another one is offering players for sale!

This year I can't give players away for nothing, whereas last year lots of teams were interested.

The problem is, everything is so interconnected - you can't just fix one thing without throwing something else off. If you reduce the number of long shots, you reduce the number of goals, so you have to compensate elsewhere, which could then unbalance a couple of other things. Also, if you reduce the number of long shots, the player has to do something else with the ball, so you are either going to increase the number of passes or increase the number of tackles by opposing players or something else, so again that could throw other things off.

Its your tactics

Only joking, i do agree with the long shots.

Well, kinda of valid point I guess from the OP

But I have 3 remarks:

1. The amplitudes of the imbalances are often exaggerated.

It's all very well to state "Too many 35yards screamers in my game, Sort it out SI!!".

But then you'll have to state how many you have and how many are realistic. And no, I'm not saying there aren't any imbalances, I'm just saying that they often are blown out of proportion...

2. Often 1 imbalance might cause another one.

For example:

too many long shots" might be caused by "not enough closing down"

And then you also have the imbalances which can (at least partially) be caused by tactical errors by the manager.

Difficult to "prove" what is caused by

A/ faulty code

b/ player errors

C/ other imbalances caused by A/ or B/

3. I forgot...

The thing that worries me is if this 'inbalance' happens in areas of the game we can't even see, or is harder to spot.

What if the stat 'creativity' just doesn't work properly? How would we know? For all we know it may have next to no impact on the match engine - it's just we cannot see it.

There are visible areas of the game - minor areas - that are just broken. What if this extends to key areas that we can't even see?

The thing that worries me is if this 'inbalance' happens in areas of the game we can't even see, or is harder to spot.

What if the stat 'creativity' just doesn't work properly? How would we know? For all we know it may have next to no impact on the match engine - it's just we cannot see it.

There are visible areas of the game - minor areas - that are just broken. What if this extends to key areas that we can't even see?

Well, that's inherent to the nature of FM and football in general.

What were the parameters causing that disastrous second half of Wigan this weekend?

We simply cannot know 100%.

So no need to worry yourself with potential bugs.

Let's focus on the things we can deduct.

So yeah, we need real life examples to compare game examples with for stuff like this if the balance is to be right. But in some circumstances it just is.

1 possible explanation:

Only 1 real world

and several thousands of FM-worlds striving towards the averages of that 1 real world.

No real surprise that some of the thousands are experiencing sometimes odd results which would lead the player to the conclusion that the game is flawed. While in fact it's only (by chance) his save which is acting funny.

I've seen real life stats on here in a very good thread that PaulC has commented on.

However as a watcher of lots of football for many years it's obvious to see some of the flaws in the early version releases.

You can just tell if something isn't balanced very well - and I'm talking about things across the leagues rather than something I'm doing with my tactics.

It's just a familiar pattern that things become very unbalanced in 0.1/0.2 versions of the game. I am however sure it will all be well again come 0.3, which was the case last year.

I think part of the problem is 'what should it be instead'? Certainly for stats like hitting the woodwork and long shots - do you have any real life stats to back up what there should actually be? Likewise with the tackling - how many tackles should there be - and do the real life stats take interceptions and the like into account?

For some issues they match up quite well - for example IRL for the Premier League last season red cards per game is 0.17 per match (although is 0.21 this season) - in the game it was at 0.19 so almost identical.

In terms of striker goals, having a quick look on the Internet a study suggested: "The striker goal percentages of the best clubs are in the 40% range compared to the 60-70%"

In our game, striker goals is currently at 56% which is well within that range. So yeah, we need real life examples to compare game examples with for stuff like this if the balance is to be right. But in some circumstances it just is.

http://community.sigames.com/showthread.php?t=167497

http://community.sigames.com/showthread.php?t=168395

we provided a lot of statistics and analisys compared to real life stats. strikers score much more than irl in leauges run on full detail, world class wingers like Messi or Ronaldo score awfully less and there are far too many shots same players are able to attempt during a season.

The other thing to bear in mind is, that once the release date is fixed, you have to be very careful about making changes. If something is a relatively minor problem (like strikers scoring too high a percentage of goals, or the woodwork being hit too often), it's often safer to leave it like that than risk introducing a much more serious problem and not having time to fix it. For example they could have reduced the number of goals by strikers, but if that meant that far too many games were finishing 0-0, then that's a much more serious problem. Stuff like this is just a fact of life with software development.

I totally appreciate that - which brings me to the part about the code. Is it a flaw in the way that it is currently written that prevents this as the long shots issue should definitely be an easier area to tweak without severe repurcussions.

Another example is 'rain' - It's obvious it rains slightly too much in matches in England but this should be able be to tweaked without major knock ons.

Like I said before, if players are taking less longshots, what should they be doing instead? It's inevitably going to effect some of the other elements of the match, either increasing passes made, or increasing tackles made by the opposition.

It's the same with the weather conditions - it's probably relatively simple to reduce the amount of rain, but the weather conditions affect the condition of the players (players tire more easily in wetter conditions because the pitches are heavier), so that would need to be checked closely after any changes, and again, there being slightly too much rain is probably less of a problem that far too many player being far too fit at the end of a season. Plus you could have knock on effects with AI squad selection because the fitness of players has been altered and so on.

Isn't the effectiveness of long shots the thing that's being toned by 20%? If so they would still have a dig but less would go in. I might have that wrong but that's how I interpreted it.

Either way, last year it wasn't a problem so if it is the number of long shots, what were they doing last year instead???

The other thing to bear in mind is, that once the release date is fixed, you have to be very careful about making changes. If something is a relatively minor problem (like strikers scoring too high a percentage of goals, or the woodwork being hit too often), it's often safer to leave it like that than risk introducing a much more serious problem and not having time to fix it. For example they could have reduced the number of goals by strikers, but if that meant that far too many games were finishing 0-0, then that's a much more serious problem. Stuff like this is just a fact of life with software development.

I don't want to be pain in the ass here but what was the testing team been doing? I understand the ME is complicated thing and can easily unbalance, but most of this scoring issues were fixed with 930.

such minor problems might be game breakers for some people, for me they certently are.

The other problem that hasn't been mentioned is that SI aren't just trying to model the English Premiership, they're trying to create a model that will be accurate down to semi-pro and even amateur level and on several different continents, let alone different countries. Personally I've found the ME to be most satisfying (perhaps not most accurate, but closest to how I perceive football at that level) in about English League 2. This isn't to say that there aren't too many screamers at that level (there probably just about are) but the player quality cancels out a lot of the issues that appear in higher leagues. In other words, most L2 midfielders don't score long shots very often where many upper league midfielders do so when the long range goals do occur it tends to be from players you'd expect and relatively infrequently as opposed to a back-up DM scoring 4 in 3 games from 35 yards for a top team.

As Glenn has said in those threads, if you provide the stats alongside the saves then it'll most definitely help us take a closer look at this. Cheers.

I deliberatly didn't want to upload the save becouse I belive such statistics should have been done long before the game was released. I can't believe it was released without noticing that some players are able to produce 80% more shots in season than irl, for example.

also it is not an issue produced by human user, it's situation happening in every save.

Like I said before, if players are taking less longshots, what should they be doing instead? It's inevitably going to effect some of the other elements of the match, either increasing passes made, or increasing tackles made by the opposition.

exactly that. it's not that uncommon that player has more than let's say 50 passes in a match, or defenders making more than 2 tackles per game, isn't it?

The problem with long shots is 'what is a long shot?' Is it anything from outside the area, anything from over 20 yards etc - I don't know what it is IRL (likewise I still can't see any figures which have been posted by anyone specifically for long shots, likewise I haven't found any when trawling through the Internet) so it's difficult to compare.

As Glenn has said in those threads, if you provide the stats alongside the saves then it'll most definitely help us take a closer look at this. Cheers.

I might have a bash at this later, probably just do this weekend's 4 English Leagues and compare them to one weekend in the FM world. Will be anecdotal evidence but better than nowt.

The problem with long shots is 'what is a long shot?' Is it anything from outside the area, anything from over 20 yards etc - I don't know what it is IRL (likewise I still can't see any figures which have been posted by anyone specifically for long shots, likewise I haven't found any when trawling through the Internet) so it's difficult to compare.

As Glenn has said in those threads, if you provide the stats alongside the saves then it'll most definitely help us take a closer look at this. Cheers.

EDIT: And to echo what bigdunk has said, it's not as though we can just tweak something and the numbers all sync up nice and lovely - every slight change has the potential to have a knock-on elsewhere so it most definitely is a massive balancing act.

Exactly.

People don't appreciate what a fine balancing act this is.

For instance, you could make long shots less likely by "toughening" up the long shots statistic - but, for the lower leagues, that might make it next to impossible for someone to score from a long shot. They could cheat - but it would render the "long shots" statistic meaningless if they did.

When I was about 15, I wrote a football simulator for my Dragon 32 - you would type in the attributes for each player (how good they were at passing and shooting and tackling etc.) and the game crunched the figures and played out, in text form, a full game over 90 real minutes.I even included extra time and penalties!

At first, a typical game was like 7 - 5 by the end. But by carefully tweaking the variables, I got it down to a more realistic 4 -2 or 3 - 0, while still maintaining the integrity of the statistics themselves.

Even 2 points one way can result in always having 0 - 0 or 1 - 0.

Mine was a simple effort - take a complex behemoth like Football Manager 2010 and you can see how difficult it must be to get close to realistically emulating the improvisational game of football.

I think the programmers have done a bloody good job on this one. But, I would reduce the likelihood of 30 yard+ shots going in. The difference between 25 yds and 30 is massive. That should be reflected.

I don't want to be pain in the ass here but what was the testing team been doing? I understand the ME is complicated thing and can easily unbalance, but most of this scoring issues were fixed with 930.

such minor problems might be game breakers for some people, for me they certently are.

Just because it was balanced in 9.3.0 doesn't mean it will be balanced 6 months of work later. The only way the balance could be preserved is if the match engine wasn't touched at all. And who's to say the testing team didn't find these things. There have been plenty of cases of the testing team finding things and them not being fixed because the potential risks from making a change far outweighs any benefits from it.

exactly that. it's not that uncommon that player has more than let's say 50 passes in a match, or defenders making more than 2 tackles per game, isn't it?

You can't just arbitrarily reduce one stat and increase another though, it doesn't work like that. Every change could affect more or less anything - if players suddenly started passing when clean through on goal because they've been coded to prefer passing instead of shooting, then that's a far more serious issue. If every player that gets within 30 yards of goal gets tackled regardless, then that's a far more serious issue.

End of 2008 2009 EPL, the 17 teams who weren't relegated scored 838 goals between them

6 yards: 36.87%

18 yards: 51.43%

18 yards plus: 11.7%

2009/2010 EPL 11 'rounds' in give or take, 319 goals total

6 yards: 36.99%

18 yards: 49.84%

18 yards plus: 13.17%

Given that FM's stats measure long shots as > 22 yards then the FM statistics may actually be even higher. Off the top of my head the goals scored by strikers where they receive to feet near the box (inside 22 yards) and turn to shoot long range would push the 18 yard + percentages higher.