Is there really too much randomness in the game, or is it all a myth?

This thread is following on from a discussion some of us had in another thread about whether or not the games in FM09 are too random.

Some people have claimed that they can replay the same match and get wildly varying results, from losing 4-0 in one game to winning 4-0 in the next one.

Obviously there is something in the game that generates a certain level of randomness, in real life this may be called luck. So in any game a butterfly effect can take place in that a team may get a piece of luck which can change the whole game. For this reason I would never expect a game replayed over and over to have the same outcome, there will be some variations. But at the same time I wouldn’t expect there to be wildly differing results over a number of retests.

The only way to test the affect of this random element properly is to save the game directly before a match, after the team has been selected and everything, and then re-load and replay that game as many times as possible. To make the test as fair as possible I used the exact same line up every time, I avoided opposition instructions completely, used the same pre-match team talk every time and as much as possible the same half time team talk. I also only made subs if forced to do so through injury.

In my current save I’d recently taken over Derby and at that point I was 8th in The Championship (1st season). The match I would be replaying is a home game against 14th place Blackpool. Because my Derby team was struggling a little before I took over, and because Blackpool are overachieving, I’d expect these games to be fairly even, with my team edging it slightly due to home advantage.

Results:

................Derby .Vs .Blackpool ...........Half Time score

Game 1........1........-.........2...........................0-2

Game 2........2........-.........2...........................2-0

Game 3........2........-.........1...........................2-0

Game 4........2........-.........2...........................1-0

Game 5........1........-.........0...........................0-0

Game 6........2........-.........1...........................1-0

Game 7........4........-.........2...........................2-0

Game 8........1........-.........1...........................1-1

Game 9........4........-.........1...........................2-1

Game 10......1........-.........3...........................0-0

I wanted to carry on and replay the game 20 times, but unfortunately I didn’t have time. The results above translate as 5 wins for me, 3 draws and 2 losses. But more importantly, apart from a couple of games, all of the score lines were very similar. In the games that I won 4-2 and 4-1 the butterfly effect could easily be seen to occur as in both games I scored early goals, possibly with the aid of some luck.

So from this first test it seems to me that there is no problem with the games random element being too prominent. If anyone else has done similar tests, or would like to, then please post more results here so we can build up a bigger picture. And all opinions of course are welcome.

That's my plan for tonight Paul, and after that Im hoping to try a similar test using a poor team against very difficult opposition.

Interesting. There have been a number of threads in the past that have attempted to tell us that no matter how many times we reload, the result and match incidents will remain the same i.e. get a red card in one match, reload it and the same red card will happen.

I think it's good to see a blend of results in your test, of course you won more games than you lost, but that is probably down to your management, which should always be the overriding factor. However, it shouldn't be the definitive factor and it doens't look like it is, as the high scoring games, draws and losses prove.

The game engine is questionable sometimes.

I once forced the AI to play with wide players only against me: -

i.e. a formation of 2 Full Backs, 2 Wing Backs 2 Wide Mids 2 Wingers and 2 Wide Strikers.

Results against them despite us players a basic 442 were still roughly the same.

I've tried this with every incarnation of FM and it would seem that this time around the AI is severly boosted.

I'd be interested to see a thread showing the specific tests you carried out and the results SMSL, if you have the time.

I scrapped the game however, if I have time over the weekend I'll post screenshots of what I mean but basically the AI can only play one stupid formation.

A 2-2-2-2-2 all down the wings. Now you would think a nice big gap in that tactics mean plenty of easy wins for us mere mortals?

Er... No! The results and performances end up being roughly the same as if the AI was playing it's normal tactics.

Conclusion? Tactics mean very little in FM09.

A variation between 4-1 and 1-3 seems pretty similar to the 4-0, 0-3 spread you quoted earlier, Chopper99.

Quite honestly, I think that's too big a variation to be considered satisfactory. While the 'random element' is likely to even itself out over a season, that doesn't, in itself, satisfy those who feel that there should be a reasonable level of control for the player over what happens in individual matches.

In your experiment, the results showed a level of consistency which I would regard as satisfactory in 7 cases, there was one marginal case (the 1-3) and 2 which were some way out of kilter. I believe that some adjustment is needed to the 'randomisation process' on that evidence, although you clearly don't.

Just shows that you will never satisfy everybody!

In the two games that you mention Rupal it was obvious to me that what happened in the early part of the match effected the outcome considerably. In both of those games I managed to score early goals and from then on the AI manager was trying to get back into the game for long periods, allowing me to extend my lead.

In the 1-3 defeat I hadn't had the rub of the green allowing me an early goal, meaning that we went into half time at 0-0. Team talks then have an effect as well, as the game is still very even. And because I wasn't changing tactics the AI came out and played better because they weren't forced to play in a reactive manner by earlier events.

Some conclusions can be drawn from my test, but it's still a pretty small and limited one. I plan to do more to build up a bigger picture, and would welcome the input of others who feel like doing the same.

So far though the overall picture doesn't appear particularly random to me.

The explanations of what happened and why are very interesting. However, they don't alter the fact that in only 70% of the cases were the results within what I think are acceptable limits of consistency.

To make a cheap point, I wouldn't regard my car engine as satisfactory if it only ran well 70% of the time; likewise, I don't really see why a match engine should be regarded as fine if it is only consistent for that proportion of the time.

As you say, the test is very limited so maybe drawing conclusions either way is rather premature.

Regrettably, I just don't have the time to join in the tests - I've only played 2 FM matches in the past week . But I will view the results with real interest!

which is worse having to much randomness or having no randomness so the favourites always win?

But Rupal, why should the results be the same? Or even similar? chopper's explanation is an excellent one - the teams are very evenly matched and the outlying results (ie. at the edge of the data range) can be explained by the element of "luck" which gives an early goal. An early goal changes the whole pattern of the match. That is totally reasonable.

Think of it this way - when you are playing, and in the first ten minutes your team has three decent attacks which produce nothing. There is very little apart from luck separating that from an opening ten where you score three goals - slight deflection, one astonishingly out-of-the-blue piece of brilliant skill, keeper makes a stupid mistake... So although there has been very little difference in actual play, the rest of the game is going to go very differently courtesy of those three goals.

Damn.

I wanted to talk a bit about this in the original thread before you went off and did your experiments Chopper, but sadly Paul closed it a bit too promptly (and unnecessarily in my opinion, but nevermind).

If you are controlling one of the teams in your reload/repeat experiments, then surely you are not actually measuring what you're setting out to determine - namely how random the outcomes of matches are. You are in fact just measuring how good you are at the game! Once that initial flap of the butterfly's wings has occurred, the outcome is now based on how good (or bad) you are at reacting to changes in the match vs how good the AI is at doing the same. And not making any tactical changes/substitutions, or making the same team-talk at half-time etc. is just as bad as making bad decisions/changes.

To really measure the effect of randomness in the game, you would be better off either (or both):

- Doing the same experiment but with you in control of the both sides.

- Doing the same experiment but with the AI controlling both sides. The benefit of this being that you can do a whole day's fixture list instead of just one match, therefore giving you a massive amount more data in the same space of time.

But Rupal, why should the results be the same? Or even similar? chopper's explanation is an excellent one - the teams are very evenly matched and the outlying results (ie. at the edge of the data range) can be explained by the element of "luck" which gives an early goal. An early goal changes the whole pattern of the match. That is totally reasonable.

Think of it this way - when you are playing, and in the first ten minutes your team has three decent attacks which produce nothing. There is very little apart from luck separating that from an opening ten where you score three goals - slight deflection, one astonishingly out-of-the-blue piece of brilliant skill, keeper makes a stupid mistake... So although there has been very little difference in actual play, the rest of the game is going to go very differently courtesy of those three goals.

I agree totally with that. The original poster explains perfectly how certain variables alter the outcome of a match, the first goal in a match, especially between two teams as equal on paper as Derby and Blackpool, is absolutely critical. I think the 5-3-2 split of results is perfectly acceptable for the teams involved. I'm sure some would have ridiculed Liverpool 9-0 Crystal Palace, Crystal Palace 4-3 Liverpool from 1990 had it happened in the game. I remember a cup tie from the early 90's where Peterborough played Kingstonian in which the first match was drawn 1-1, Peterborough then won the replay 9-1 only to have to replay again after a coin throwing incident - in the third game Peterborough scrapped through 1-0. These three matches took place within a couple of weeks.

Let's face it, football can throw up some pretty random, unpredictable results.

Squirmy - What I want to know is whether a match that the user goes into has too much of a random factor in it. Because the AI teams generally perform how you would expect when playing each other, and because their positions in the league at the end of a season generally reflect this, I assume that there is not a problem when it comes to AI vs AI games.

The original debate was born out of people claiming that they can replay the same game against an AI opponent and get wildly differing results, which I feel may be down to what they're doing during the game. By eliminating this (and for my next test I will not be using pre-match or half time team talks at all) then the only thing that now effects the outcome of the game is the random element. The team and tactics have already been set and do not change, and neither does anything else. Because I don't react to changes in the match at all in any of the games, the fact that I'm controlling the team sort of becomes irrelavent for this experiment.

If it was done doing AI vs AI then both managers would constantly be changing things to react to what happens and I think you'd actually get less valid results.

If someone wants to do an AI vs AI test though they are more than welcome, but I'll be sticking to reloading the exact same Human Vs AI game for the time being as that's where people seemed to claim that there was a problem.

Kewell08

It's a matter of BALANCE. Remember that this is supposed to be a GAME. If there is going to be a very wide spread of results dependent on factors (let's call them 'luck' just for simplicity) which are outside the player's control then this detracts from gameplay.

It may well be that you are right and that the 'luck' factor is 'realistic'. Unfortunately, when you translate this into the GAME scenario it means that you are not in control of matters. The concept of having a certain 'problem' (that is, playing the opposition) and using one's skill and experience to try to 'solve' that 'problem' is undermined by the 'luck' element which overrides the player's input. If I play a match and win 4-0 is it because my tactics are good? Or was it just the fact that I had 'luck' on my side? I don't know. If I lose 3-0, are my tactics wrong? Should I alter them? Or was the 'luck' the other way this time? Again, I don't know.

If we have what amounts to almost total unpredictability, then, however philosophically realistic according to chaos theory this may be, the concept of an enjoyable, playable GAME is beginning to be lost. I believe that to provide a good gameplaying experience it is necessary to provide an element of consistency within reasonable parameters so that one's decisions can be made on a rational basis, even if this is theoretically less 'realistic'.

It may well be that you are right and that the 'luck' factor is 'realistic'. Unfortunately, when you translate this into the GAME scenario it means that you are not in control of matters. The concept of having a certain 'problem' (that is, playing the opposition) and using one's skill and experience to try to 'solve' that 'problem' is undermined by the 'luck' element which overrides the player's input. If I play a match and win 4-0 is it because my tactics are good? Or was it just the fact that I had 'luck' on my side? I don't know. If I lose 3-0, are my tactics wrong? Should I alter them? Or was the 'luck' the other way this time? Again, I don't know.

Why can't it be a bit of both tho?

If both teams have good tactics, what would seperate them?

Rupal, I do agree with you on this point. If the results of this thread point to the fact that there is too much of a random element then you're right in the fact that this will retract from the enjoyment of the game.

So far (in my opinion) my test has not shown that. Almost all of the results were very similar which points to a trend of my tactics, players etc being the important factor. Luck only seemed to play a part in a couple of games, had I been making changes to react to a lucky goal we may have seen a different result and my tactical changes may have won through.

As I said, more tests are needed before we can start to see a pattern, tonight I will be using my Man City team to play a small team to see what results I end up with.

which is worse having to much randomness or having no randomness so the favourites always win?

To much is worse.

If we only could have the perfect amount of randomness.

Have mixed feelings about the Butterfly effect. Agree to the fact that it is real and afects games but feel it plays to big of a role in the game currently.

It's strange but i can't relly find words to describe why i don't like it or what aspect of it i feel is unrealistic, but in the end i don't like the fact that i feel like i don't have any control over the games.

It's hard to try and show these kind of things as it is very dependant of users own opinion. My experience has been very inconsistant as i try and get a feel for the game before the new patch comes. So far i have played 4 seasons worth of games, but i have only progressed 1.5 sesons in the game. Thats how often i reload and test different formations.

I have no problem winning or losing games. I'm not happy when i win and the other team has dominated because i feel that i have my tactics wrong. (usually i am heavy favourite when i play)

I just want to feel more in control.

To much is worse.

If we only could have the perfect amount of randomness.

Have mixed feelings about the Butterfly effect. Agree to the fact that it is real and afects games but feel it plays to big of a role in the game currently.

It's strange but i can't relly find words to describe why i don't like it or what aspect of it i feel is unrealistic, but in the end i don't like the fact that i feel like i don't have any control over the games.

It's hard to try and show these kind of things as it is very dependant of users own opinion. My experience has been very inconsistant as i try and get a feel for the game before the new patch comes. So far i have played 4 seasons worth of games, but i have only progressed 1.5 sesons in the game. Thats how often i reload and test different formations.

I have no problem winning or losing games. I'm not happy when i win and the other team has dominated because i feel that i have my tactics wrong. (usually i am heavy favourite when i play)

I just want to feel more in control.

Do you feel you lack of control is due to a random element though? I feel I have a certain lack of control, but for me it is because I find the tactical system in the current game counter intuative; i.e. I find it difficult to know why my tactics are or are not working.

Why can't it be a bit of both tho?

If both teams have good tactics, what would seperate them?

Luck, of course (presuming, as I take it that you are, that the teams are about of equal standards, have equally good morale, etc).

But that isn't really the point for me. I'm looking at this as a game and the player enjoyment factor, not as an attempt to portray reality. And if I don't know what's down to my own incompetence and what is down to bad luck when I lose again, then that enjoyment, for me, is diminished. It's not as much fun as it ought to be and it's that which bothers me.

Kewell08

It's a matter of BALANCE. Remember that this is supposed to be a GAME. If there is going to be a very wide spread of results dependent on factors (let's call them 'luck' just for simplicity) which are outside the player's control then this detracts from gameplay.

It may well be that you are right and that the 'luck' factor is 'realistic'. Unfortunately, when you translate this into the GAME scenario it means that you are not in control of matters. The concept of having a certain 'problem' (that is, playing the opposition) and using one's skill and experience to try to 'solve' that 'problem' is undermined by the 'luck' element which overrides the player's input. If I play a match and win 4-0 is it because my tactics are good? Or was it just the fact that I had 'luck' on my side? I don't know. If I lose 3-0, are my tactics wrong? Should I alter them? Or was the 'luck' the other way this time? Again, I don't know.

If we have what amounts to almost total unpredictability, then, however philosophically realistic according to chaos theory this may be, the concept of an enjoyable, playable GAME is beginning to be lost. I believe that to provide a good gameplaying experience it is necessary to provide an element of consistency within reasonable parameters so that one's decisions can be made on a rational basis, even if this is theoretically less 'realistic'.

Personally I wouldn't want to play the game the way you're putting it forward. Say for example I battle through to the 3rd round of the FA Cup as Wrexham and draw Arsenal. Rationally I know I will never outplay them and only a huge slice of luck will give me a chance - removing the luck factor from the game virtually eliminates any chance of a major shock result. In my fictional role as Wrexham manager I might as well pull my team out of the Cup if I draw a top team - where's the fun in that?

Nearly all sport involves an element of luck, if FM purports itself to be a realistic simulation of course it should include a certain random element.

Luck, of course (presuming, as I take it that you are, that the teams are about of equal standards, have equally good morale, etc).

But that isn't really the point for me. I'm looking at this as a game and the player enjoyment factor, not as an attempt to portray reality. And if I don't know what's down to my own incompetence and what is down to bad luck when I lose again, then that enjoyment, for me, is diminished. It's not as much fun as it ought to be and it's that which bothers me.

I agree with you on main point, but like chopper said, I think that might have more to do with the shortfalls of the current tactics system and not the random element.

To go back to the point of randomness, somebody posted a screenshot of the Liverpool - Stoke game this season where Liverpool had 30 shots on goal, if just one of those had gone in, early in the match then who knows what the score might have been.

Of course that game was real life, but I'm trying to explain as well as I can about how the randomness (or luck) plays a big part in real life games too. Admittedly tho, I lack the skills to explain myself as well as I would like tho.

For me this element of randomness is a non-issue. My irritations come from what appears to be too much randomness from game to game - e.g. when I play a similar strength of opponent with a similar tactic and against one we create loads, against the other we create next to nothing. Then also a recent example in which my Scunthorpe team usually win at home and lose away (fair enough, no problem with that), then suddenly we start losing home games and playing like Brazil away from home with seemingly nothing having changed and leaving me feeling like I'm just a passenger, left at the whim of hidden attributes or random elements that are influencing results far more than any changes I make in tactics.

That is probably just frustration on my part, but it does highlight, for me, the difference between a simulation and a game. Inexplicable things happen all the time in real life and thus in a simulation, but for a game I prefer a situation in which I feel it is down to me to work out what is wrong, try to fix it and observe the results of that, etc. Maybe I just have a load of players with high hidden inconsistency attribute or too many of whom don't like big matches or some other strange goings on that I can't see. Too big a random element is very frustrating though

Back to the original point, and why I said it isn't an issue and why I went slightly off topic after that - the situation described in the original post is a totally theoretical one that has no parallel in real life. You can't replay the exact same game from the exact same starting conditions in real life so we have absolutely no data against which to measure how likely it would be that the result would be the same each time. I know you can say that big team should beat small team 99 times out of 100 etc, but it is just speculation.

People who are reloading games and then complaining about getting different results are heading off into a totally unrealistic situation that should never realistically be encountered. We should be in blissful ignorance of how the game might have gone on another 9 occasions just like we are in real life.

I think the point about the tactical system is very, very important too. I'd still not like the 'randomness' of the thing but if there was a clearer relationship between what I saw on the pitch and what my changing the sliders did, then I would feel more able to change things effectively if they were going pear shaped.

I might turn a game round from losing 2-0 to drawing 2-2 (occasionally I amaze myself and can do this) by messing about with the sliders but I honestly am thoroughly confused as to exactly what I did and what effect it had. It's rather like poking about in ignorance under a car bonnet wiggling all the wires in the hope that this time the damn thing will start......

The team and tactics have already been set and do not change, and neither does anything else. Because I don't react to changes in the match at all in any of the games, the fact that I'm controlling the team sort of becomes irrelavent for this experiment.

No, because as I said, you not reacting to changes is just as bad and will produce flawed results, because the AI should and will be reacting appropriately to take advantage of the situation.

Like I was saying in the other thread. If the smallest random difference early on in the game (e.g. the ball going in off the post, or bouncing off the outside for a goal-kick) means the difference between the AI going a goal infront or not, then from then on the game will take a drastically different course and the AI team will change it's style of play appropriately throughout the match, whereas you will not. For this reason, the AI could then go on to win by 3 or 4, because your tactics at the very beginning (that you are not changing) play perfectly into the AI's hands in this situation of you being behind.

If it was done doing AI vs AI then both managers would constantly be changing things to react to what happens and I think you'd actually get less valid results.

Well, the results should be valid in showing that regardless of any random-number-generator, the match-engine and manager AI is such that the expected results generally happen the expected amount of time. e.g. Manchester United vs West Brom at OT, United would win something like 70% of the time or whatever (no disrespect to Baggies fans).

If someone wants to do an AI vs AI test though they are more than welcome, but I'll be sticking to reloading the exact same Human Vs AI game for the time being as that's where people seemed to claim that there was a problem.

Well, that's up to you, I'm only here discussing it because even though I don't know you, I don't want to see you waste your evenings.

At the very least, I would suggest making your tests Human vs Human, and making the absolute minimum changes for each side (as you were already doing).

Well my tactics are working. (why i don't know... trail and error is my way to go)

They play as i want and i feel the ammount of posessin and amount of chances i get is realistic.

Play with Man Utd.

But because this is my trial campaign before the comming patch i reload games to try different formations and to make sure that one tactic when dominating and scoring lots wasn't just luck. When i have done these tests i found i could go with United from 0-1 losses where i dominated (lost because of bad luck) to wining in the next 3-0. Realistic if you look at the stats.

Then play it once mor and find you lose 1-2??. Again the representation on the pitch and the statistics say you dominated and you where heavily favourd, but "luck" made you lose.???

Or is it tactics?? There is no way of knowing as the results differ so much from game to game for me.

Some good points there Squirmy, I think I may make my next test human vs human. At least then neither manager will be making changes and then the difference in results will, as you pointed out, be solely down to randomness and the butterfly effect

Hey Chopper,

It doesn't matter if this supposed randomness is a myth or not. The fact that the myth is there makes the question void. If we perceive it as such it's a problem. No matter who comes up with tests and results to prove that it's a myth. I find it a hard and frustrating thing, because I can't pinpoint the solution to the problem.

I remember a cup tie from the early 90's where Peterborough played Kingstonian in which the first match was drawn 1-1, Peterborough then won the replay 9-1 only to have to replay again after a coin throwing incident - in the third game Peterborough scrapped through 1-0. These three matches took place within a couple of weeks.

I was actually at the 9-1 game, and the reason Peterborough won 9-1 was because the coin hit the Kingstonian goalkeeper on the head and was unable to continue. With Kingstonian not having a substitute goalkeeper on the bench a comfortable Peterborough win was on the cards.

Can this be put down to randomness or luck?

I was actually at the 9-1 game and the reason Peterborough won 9-1 was because the coin hit the Kingstonian goalkeeper on the head and was unable to continue and with Kingstonian not having a substitute goalkeeper on the bench a comfortable Peterborough win was on the cards.

Can be put down to randomness or luck

Or the stupidness and disrespect of the opposition fans.

I got really annoyed losing 1-0 to barnet after a bizare own goal by my keeper. I played this match about 10 times and lost 1-0 most times regardless of my tactics i kept them the same at first. Ocasionally 2-0 when i went all out attack. I even played one game 3,3,4 and lost 3-2 (in injury time). I gave up then.

I was actually at the 9-1 game, and the reason Peterborough won 9-1 was because the coin hit the Kingstonian goalkeeper on the head and was unable to continue. With Kingstonian not having a substitute goalkeeper on the bench a comfortable Peterborough win was on the cards.

Can this be put down to randomness or luck?

Tony Philliskirk maintains that the keeper faked it and was pinting it in the club afterwards. I think it can probably be put down to randomness...

To my mind this experiment is totally pointless and can't prove or disprove anything.

Mainly because the conditions of the test can never happen in real life and therefore there can be no true parameters to measure against.

In fact testing to prove correct randomness in a game such as FM is a direct contradiction in terms, even if one of those results published above was a 10-0 win for one team it would neither prove or disprove that the random effect was wrongly applied in that match, because that's how random rl football sometimes is.

Luck truly is a key part of the game. In fact entire saves have come down to pure blind luck for me before.

Last year I had 5 odd saves, and this is how they ended up:

1st: Derby - Relegated by 1 point, average championship season, missed out on the playoffs, falling into financial trouble. 3rd season, starts terribly I resign.

2nd: Derby - Survived (by a couple of points, can't exactly remember) and won the FA Cup. Came 7th and won the UEFA cup in the second. Came 4th and knocked out of the UEFA cup in the first knockout round. Won the Premier League and Champions League. Success...

3rd: QPR - Rollicking start, signed a lot of very good players and it showed early. At christmas 1st by 6 points. Ended up coming 3rd and I made it to the playoff final; I went 2-0 up before Ipswich came back at me with 3 goals, I equalised in the 93rd minute before they won it in the 95th. Signed some very very good players (at least at Championship level), but in the end failed to get promoted again, coming 3rd again, but losing in the first playoff match.

4th: Burton - Destroyed the Blue Square Premier and won one of the non-league cups. Made it to the 5th or 6th round of the FA cup before losing to Pompey. Firing my way up League Two, but then I had to give up on it due to lack of time.

5th: Queensland Roar - 5th after 10 matches... Gave up.

As for FM 2009 so far:

1st: Derby - Promoted on 96 points with Derby. Then came 12th and won the FA Cup (Ipswich won it the year after, still in the Championship), followed it up by coming 2nd and winning the UEFA Cup. The next season I came 2nd again, won the FA Cup and Champions League, then followed that up by winning the Premier League and Champions League.

2nd: Burton - 13th after 30 matches... Gave up.

There is an awful lot of luck in the game, and that's a good thing.

To my mind this experiment is totally pointless and can't prove or disprove anything.

Mainly because the conditions of the test can never happen in real life and therefore there can be no true parameters to measure against.

In fact testing to prove correct randomness in a game such as FM is a direct contradiction in terms, even if one of those results published above was a 10-0 win for one team it would neither prove or disprove that the random effect was wrongly applied in that match, because that's how random rl football sometimes is.

If it is established that there is a high random element in the matches then this may indicate that the game is reflective of reality (although I feel that this is open to question). The question is, though, is a high degree of randomness good in a computer game? The two things are not necessarily the same.

I'm not that worried about whether FM is a minutely accurate simulation. I do want it to have good gameplaying qualities. If the randomness makes it less enjoyable as a game then the fact that it may be 'realistic' doesn't help.

I'm reading a lot of "I don't like the butterfly effect because I have no control over my team" in this thread. That's not what the butterfly effect is or means - yes it might be true if the butterfly effect was absolute maximum and user inputs actually had no effect. chopper's tests, at least to me, prove that the randomness is not that huge.

I do agree to some extent with Kriss and glam that the tests are difficult to analyse because they have no correlation to real football, as this situation is not one you would ever have (ie. replaying the same game under the same conditions) but we aren't really testing whether it is realistic. chopper's tests are more of the actual gaming element, because there are and have been complaints that the game is too random and you have no control. I think chopper's tests disprove that, in any case that is the point up for discussion.

Just playing a few matches for FM09, but from my FM08 experience, I'll get weird scoreline when refs ****ed up the match, either against my team(usually lost by a big margin) or favors my team(usually win by a big margin). Whoever gets more favoring decisions will usually wins(not always)..

What's a reasonable degree of randomness will, of course, be a matter of opinion. On the other side of the coin, posts like stonegate's concern me too. Whether it's 1) keeping the same tactics seems to lead to too wide a spread of results or 2) using a whole set of different tactics DOESN'T lead to a wide ENOUGH spread of results, the overall effect of things like this is to make me question just how much effect my input (ie things I can actually DO something about) actually has on the result of a match at all. This is where that lack of control feeling comes in.

Paul C, you take an overall view of the league tables at the end of a season and if they seem reasonable then, from your point of view, your simulation is a success. I completely understand why you take that view of things.

But I'm looking for a game Paul, not an attempt to replicate 'reality'. If I feel that my input is overwhelmed by factors over which I have no control, then, however accurately FM 'simulates a football world and puts you in that world' then, as a game it isn't as good for me as it should be.

I suppose it depends on what your customer base is looking for. Many people are clearly satisfied with your approach - but some others aren't. Currently, FM09 is the only management simulation available (I use that word advisedly). Those of us who would prefer a game will await the new CM and hope (rather against hope) that it will provide more fun even at the cost of less 'realism'. Or maybe some new competitor will emerge with a product which more closely matches what we are looking for.

Horses for courses.

I have to say, I think this is a pointless exercise. As I said - look at the league tables at the end of the season. They reflect the engine in a way that can be compared to real life. If the whole thing was random then Man U etc wouldnt be up there and Stoke etc down there.

The whole point of FM is that it simulates a football world and puts you in that world. That is the game. To somehow make the match engine "less random" in order to somehow facilitate "gameplay" makes zero sense to me. Part of the beauty of football is its cause and effect, which itself implies you can make a difference through your actions as manager.

I agree to a certain extent, with both this response and that of Kriss.

However, the reason I did this in the first place is because a lot of people have been claiming that they have no control over their games and that the results of human vs AI teams is too random. People have tried to justify this belief by replaying the same game over and over and then claiming that they had wildly differing results that were far too random.

I'm simply trying to disprove this, as I believe the user's feeling of randomness or lack of control is down to the tactical side of the game.

And although unrealistic I don't see the experiment as being pointless. If I do this test again using a strong human team Vs a weak human team, with both teams using the exact same tactics, making no changes and not bothering with team talks, then you would expect there to be an overall pattern of the stronger team winning the majority of games. This is what I fully expect to happen, and hopefully this will put to bed peoples feeling that their lack of control is down to some over-cooked random element.

However, if it doesn't then in my mind that in itself proves that something may be wrong.

Unfortunately I didn't have time to do the tests I was planning to last night, I'll hopefully do them tonight.

I think FM08 (I still haven't bought FM09 yet so don't know how different it is) simulates real life very well if you "zoom out" to a certain level of looking at things - i.e. like Paul C was saying, looking at the end of season league table in my game I see nothing that is wildly ridiculous. What irritates me slightly (although not necessarily a bug, real football managers also have similar irritations I'm sure ) is how you get there. If you zoom in far closer, looking at a game to game level it is not always so obviously realistic.

As I said in my earlier post, maybe I just have a team of players whose hidden attributes cause them to be prone to making mistakes or not liking big matches or being generally inconsistent, but there are so many times when I feel that I am not managing the team, rather I am just watching them. Of course I may have just lost all my skill completely from when I was able to see cause and effect of my actions in earlier versions.

Managing Scunthorpe I seem to have inexplicable long bad patches every season that just appear out of nowhere just as a good run is starting. Morale in the team is high, tactics are the same (with minor tweaks) as those that were working against similar strength teams previously (and do so again later on in the season so it isn't just teams learning my tactics), the players are mostly the same and yet we start missing all our chances. Usually at the start of the run we dominate a game, but still lose, then it degenerates into us just getting played off the park and losing. Then, out of nowhere we get a good result and suddenly we are off again. I react differently on different occasions that this happens - sometimes I stick to my tactics because they worked before and should come good again, sometimes I make major changes to try to shake things up. Sometimes the latter works, sometimes it doesn't, sometimes the former works and sometimes it doesn't.

From a simulation perspective maybe this is quite accurate. I can't really complain about overall results because I am in my 5th season in the Championship with Scunthorpe so I have done better than they did in real life, it is just that from a game perspective I can't seem to get any handle on cause and effect of my actions. Towards the end of last season our usual "win at home, lose away" routine was turned on its head - we kept losing at home whilst somehow pulling off great results on the road. Losing at home to a lowly team one week then thrashing Ipswich 7-2 away from home the next week confused me hugely since I was doing very little different to when we had been winning at home and losing away

I am sure that the exercise does have a point.

The mods have just closed a thread where somebody was complaining that FM'09 was the worst of the series because nothing that he did seemed to make any difference. It wasn't very specific in what was wrong, which is why it was closed. Nevertheless, before closure, the advice given was to visit the tactical forums, possibly download some tactics and that this would help to sort things out. Very sensible advice indeed....IF the tactics you choose do have a significant input.

When you get concerns like those which have been expressed, or posts like stonegate's which imply that any old tactics will get the same result, you're bound to begin to wonder about the whole process.

I really, really hope that chopper99's experiments prove that there isn't a problem. I think that people need to know one way or the other though.

And great post, glamdring. That's exactly how I feel.

I have to say, I think this is a pointless exercise. As I said - look at the league tables at the end of the season. They reflect the engine in a way that can be compared to real life. If the whole thing was random then Man U etc wouldnt be up there and Stoke etc down there.

The whole point of FM is that it simulates a football world and puts you in that world. That is the game. To somehow make the match engine "less random" in order to somehow facilitate "gameplay" makes zero sense to me. Part of the beauty of football is its cause and effect, which itself implies you can make a difference through your actions as manager.

Paul, with all due respect mate, this is where your looking at it in completely the wrong context.

The thread of mine which you closed, was all about the fact that although i was UNBEATEN in my Man City save and topping the league(which i won) it was the 38 matches and endless hours playing these matches that have become a chore and ruined my experience of the game.

If you think its all about the final table and making that table appear realistic, then you may as well forego playing all the matches and just have the game calculate a final league table instead. Of course then you would say "where is the fun in that?" which is exactly what i am saying to you is how i feel about the game in its current state.

What is the point of me playing each and every game, when the Match Experience appears to be all about making sure that the final table turns out believable and realistic?

I understand that my post is not about the random element in the game, but you did close my thread in favour of starting this one, which is why i have posted here.

As an example of my issue with the game, i found that although unbeaten, my players were having to create 5 or 6 times as many CCC's as the AI just to score a goal, as well as the fact that my supposed World Class defenders would make more errors(largely missed interceptions) than even the poorest of opposition.

As i was dominant in over 95% of the games i played, i went through the entire season without the slightest bit of luck, which i believe had i had my share, the whole season would have been a non event and i would have had the EPL wrapped up after 30 games.

Now as you know, i am not claiming CHEAT but rather a ME and AI that has been coded to produce a believable set of results and final table(the problem being, the match itself becomes largely unbelievable, as do the match stats the game has produced)

Ergo, my enjoyment of the game is massively reduced.

Garry,

It was closed because it had served its purpose ( some key fixes for 9.3 ) and had veered off topic.

No problem

One thing - regarding CCC's.....I've fixed a bug for 9.3 that assigns these for some headers that probably arent valid CCC's. Not sure if it will apply to you, but worth pointing out all the same.

Sadly, probably not, as most of the CCC's i create are simple 1 on 1's that players tend to pass back to the opposing GK(have tried teaching "places shots" and "rounds keeper" with little effect)

Btw I dont really see how you can have a believable set of results and league table without the match engine reflecting the data sent into it, but there you go.

Cheers,

Paul

My point here is that, the performances and stats produced should reflect the results and league table, rather than a pre coded influence to have the results and table appear realistic at a cost of making the match experience appear unrealistic.

It appears that there is little correlation between these key elements and any attempt to marry these up, would make for a much better and believable experience.

Just to add, Rich(wwfan) and i, had a long established discussion about this very thing and he assured me that these kinds of results and performances would largely dissapear in FM09 because of the much improved ME, but if anything, i have witnessed the exact opposite, which is very disconcerting.

(The above is not a dig at wwfan, he remains very respected by myself and regarded as a mate, albeit one i have never met, i am just pointing out that he too was willing to accept that these kinds of games were much too frequent and were counter intuitive to the whole experience)

Interesting debate, I haven't had chance to get the information off my PC from the tests I ran.

One interesting stat I did find was that results were very similar despite team selection and team formation, which made me cry a little inside.

If it helps at all... I feel pretty much the same way. I don't mind losing games - I've never complained about that. What I do feel like, is that there's no relation between the way I lose the games and the tactics/input I put into the game. Just like there's no relation between the way I win games and my input.

I usually see how all the games are decided by small mistakes by the players, some ridiculous ones, such as stupid 'rebounds' that happen way too often since the introduction of the 3D ME (to me, they seem related to the addition of some sort of physical 'mass' to the players), or unbelievable long shots being constantly made by crap players.

The worst of all, is that I don't mind losing the occasional game because of bad luck, but I can't take winning all my games because of luck as well!! Unrealistic things such as my lower league striker with 10 in long shots scoring an impressive 40 yards shot every match. Those are things that can happen from time to time, but I know (or at least I feel) that I can't base my team's gameplay on such an unbelievable, unpredictable and unrealistic event. Yet THIS kind of events are the ones which are deciding the results of all my matches. Every time.

It's ok if I win/lose matches by stupid random mistakes by any players, or the occasional bit of luck that makes a 30 yard shot go in. I'll try to reduce them/exploit the opposition ones. What's definately NOT right, is that despite those stupid mistakes deciding the results of ALL my matches (i.e.: being decissive goals), the final positions in the league table are yet the expected ones in the end!!

I am able to make my team play as I want them to play. I've achieved that. However, I find this has no impact at all on the match result. I keep having this weird feeling that the games are following some kind of script, and that no matter how I make them play, if I'm playing too bad I will score a stupid goal just to compensate things, while if I'm playing too well, I'll concede a goal the same way just to keep it going in the 'scripted' direction.

Of course, at the end of the season, you'll take a look at the league table and you'll see that things look as they should. Teams are in the place they should be, and everything seems fine. But knowing that your games were ANYTHING BUT NORMAL (either for the good, or for the bad of it), you would expect the final result to be somewhat 'altered' from what would've happened in a regular situation.

Yet, in the end, it seems that the stupid amount of unbelievable mistakes made by your defenders, etc. misteriously COMPENSATES with the amount of unbelievable amount of spectacular goals scored by your strikers/mistakes made by the opposition, thus producing a final result which seems pretty realistic.

It's like a complete chaos, but one which has some kind of 'order' or probability setting that makes it all fit together and produce the expected result.

In other words, it's as if the national lottery, after 10 years, had given the exact same amount of € per square meter throughout the whole country. It would be random each individual time, yes, but in the long term it would prove that the 'randomness' was somehow orchested to keep itself inside some road to produce a desired result. Only that FM shouldn't be THAT random in the first place, because there are different managers and different players giving different input to the ME. And then, if randomness happened the way it happens, then the produced results should vary wildly and prove that the season was a very unusual one.

However, every FM09 match lives in this randomness that keeps producing realistic results despite the games being completely unreal and apparently driven by luck.

I'm not surprised by the people claiming that a 2-2-2-2 tactic (all flanks) produced the same result than a usual 4-4-2. I've sometimes set up stupid formations that completely exposed my team to receive lots of goals, and the opposition created lots of chances because of that. Yet, 'magically' enough, most of those chances were unbelievably saved by my crap lower league goalkeeper, or hit the post, or were somehow missed by their strikers ("how could he miss that?"), and in the end I even scored the weird goal in some counter to produce a realistic result.

I'm not sure what's going on, but it just doesn't feel right to me and I'm not enjoying the game as much as I expected. I loved FM08, I have a game still going on in year 2023 and I still feel like it produces realistic results, despite the 'arrowed' match engine. But despite the big, BIG effort I'm making to try and understand the logic behind this new matche engine (as I do every year), I'm finding it impossible to get into the game, as I don't feel the game is taking into consideration much of what I do towards producing the final results.