I'm no math major but these "odds" seem fishy?

[IOMLS]

Which did you use? 'Do this for the fans'? I would imagine that wouldn't be particularly useful in this game. To be honest I only ever use that if playing a rival at home. What was the feedback from your Assistant manager regarding the teamtalk?

Did you use the same tactics every time? Maybe the backroom staff gave misguided information, in that they thought that Tottenham were poor against a 4-4-2 but in actual fact that wasn't the case. Did you check out Tottenham's results against 4-4-2 tactics and other tactics to see if in fact that was the correct thing to do.

I'm pretty surprised by that set of results to be honest. So far FM2010 has seemed fairly straight forward in terms of being intuitive and getting teams to perform well, usually leading to winning matches that you shouldn't rather than losing matches you should.

Out of interest, what were the actual lineups for the match?

[x42bn6]

are you really this dense? i mean seriously. 4-4-2 ... 3-5-2 .... 4-1-4-1 ... it shouldn't matter that much ... aside from fielding a side of 11 goalkeepers there is no "tactical formation" that should skew the result of two evenly matched teams much less one moderately superiour team to 1 out of 35.

Oh my! Go ahead, go play the old 2-3-5 then. I'm pretty sure there is no difference, as it is a formation.

And it's not quite true - you could argue that on average, 4-5-1 beats 4-4-2 (more midfield domination) which beats 4-3-3 (more midfield domination) which beats 4-5-1 (more space for the 4-3-3's wingers to run around). You wouldn't play 3-4-3 with Manchester United and you wouldn't play 5-4-1 with Spurs as they have no sweeper. Tactics are fundamental here.

4-4-2 is basic as it is by far the most popular formation. Doesn't mean it's the best. On average it covers the most ground on the pitch which is why it is easiest to use. Doesn't mean it's the best. Certainly doesn't mean it's the most appropriate to play against Spurs.

[ArseneLupin42]

Ok haven’t read the whole topic but we are testing how the match engine reflects the bookie odds given to you before this Spurs – Manchester United match. This first thing we have to do is convert the odds into percentages. This is done by divining the last number in the fraction by the total of the fractions. Spurs to win = 4-1 = 1/(4+1) = 20% for a spurs victory, if you run the match 100 times spurs should win 20 of them times.

The percentages of odds are as followed (of course this are not taking into account team-talk, ect, that effect player moral)

Spurs win 20% of the time

Draw 16% of the time

United win 64% of the time

Now back to you’re test run over the 35 matches Now united have clearly been winning over time, this is to be expected looking at the odds given to you by the bookies. As we would expect to have them win the match the majority of the time. The percentages after you’re 35 games are as followed.

Spurs Win 3%

Draw 12%

United Win 84%

As we can see united have a lead over the 35 matches as we would have expected from our rematches odds. The lead is slightly bigger than expected due to spurs only winning 1 game from the 35. The draws a little lower than expected but that can be put down to stand variance that happens when betting. I think it would be more interesting to post you’re results after playing the match for a full 100 times. Then see how the game reflects the pre match odds, mainly because I’m curious if Spurs can pull the win rate up to around the expected 20%.

this is A. very good on helping me understand the bookie odds to percentages but you maybe should have read a bit more carefully ... tho i admit my writing is not perhaps the most cogent ...

it wasn't Spurs that one only 1 out of 35 ... it was MAN UTD that won only 1 out of 35

thus my maintaining that such a result is way into the stratospherically improbably realm given what i know of statistics and math ...

but i really do appreciate your break down as it only further illustrates just how ludicrous the results from the 35 matches was ...

[Pob]

If you think injuries are unrealistic, then use the editor to make them less frequent... it is an area of the game that SI always say matches reality since they partnered with thephysioroom.

The match odds assume you are are a half-decent manager who will change your tactics to suit the situation, e.g. play defensively when holding onto a lead, or attack when you are losing.

The odds sometimes don't make sense mathematically (the odds should add up to just under 100%, any less than 100% is the bookies' profit margin; sometimes the odds add up to way less than this.)

[ArseneLupin42]

Which did you use? 'Do this for the fans'? I would imagine that wouldn't be particularly useful in this game. To be honest I only ever use that if playing a rival at home. What was the feedback from your Assistant manager regarding the teamtalk?

Did you use the same tactics every time? Maybe the backroom staff gave misguided information, in that they thought that Tottenham were poor against a 4-4-2 but in actual fact that wasn't the case. Did you check out Tottenham's results against 4-4-2 tactics and other tactics to see if in fact that was the correct thing to do.

I'm pretty surprised by that set of results to be honest. So far FM2010 has seemed fairly straight forward in terms of being intuitive and getting teams to perform well, usually leading to winning matches that you shouldn't rather than losing matches you should.

Out of interest, what were the actual lineups for the match?

i was on the road and only a moderate favourite so i used "you can win this" ...

and as i maintained before ... and as i have tried and failed to convey to the other "it is the magic tactics silly" crowd that there is NO formation and tactic short of something truly idiotic like 11 keepers in a 1-1-8 that should result in such a skewed and improbable set of outcomes ....

and for the record for the other commenter who keeps telling me that me playing a 4-4-2 should somehow reduce my chance to win a match against anyone into the 3% range is ludicrous ... what better formation would there be for Van Der Saar, Evra, Vidic, Ferdinand, Haregraves (trained up to be accomplished at RB), Darijo Srna, Juan Manuel Vargas, Carrick, Danielle de Rossi, Rooney and Berbatov? those are A all of my best and highest rated players (all 3 stars of CA or above) ... and it gets them all on the field in the exact positions they are great at ... i mean seriously ... and they were in a 4-4-2 as well so there was no (oh they had 5 in the midfield so blah blah blah)

[uistbhoy]

are you really this dense? i mean seriously. 4-4-2 ... 3-5-2 .... 4-1-4-1 ... it shouldn't matter that much ... aside from fielding a side of 11 goalkeepers there is no "tactical formation" that should skew the result of two evenly matched teams much less one moderately superiour team to 1 out of 35.

and a 4-4-2 is a pretty basic formation and tactic ... i mean to me who actually knows something about football ... one center back that is strong, good in the air and tough at marking tackling and with good positional sense and concentration (Vidic) ... one center back that is still good at those things tho maybe slightly less but is better at ball control and passing and looks to get the ball out to start offense (Ferdinand) ... two full backs that have decent pace/stamina/crossing and good (but not as good as the center backs) defensive skills (marking/tackling/anticipation/position/concentration) set to "automatic" which I assume means based on the good "decision" skills get forward when they should and back when they should most of the time (Evra and Haregaves-trained up to "accomplished RB) ... one central midfielder geared more towards defense or in this case an all aroud box to box stud (de Rossi who is a statistical monster in this game) ... and once central midfielder to act as a playmaker (Carrick, another statistical monster in the demo) ... two wide midfielders who in my case are of the D/WB/M variety ... Juan Manuel Vargas and Darijo Srna both of whom are top notch and I set to play defensive wingers who attack ... one all around striker leading the line (wayne rooney) and one striker looking to link between the midfield and the goal (berbatov) ....

that is a basic tactic ... perfectly fit for the players in it ... AND NO counter tactic or set of circumstance should result in that line up ... in that very sensible formation .... playing neither too offensive or too defensive to win only 1 out of 35 games against anybody ... even a team of Superman-Jesus hybrids ... much less Tottenham ....

you dont have some super mystical magic "tactics" powers that others of us mere mortals are unaware of ... the game does not require a phd in football or psychic tactic powers to see what positions the players are best at (there is a handy diagram with green glowing dots) and find the best players possible at those positions (there is even a handy diagram you can use to compare one to player to another directly where a visual geometric diagram of their relative strengths overlap showing how much and where one is superiour to another) and them put them in those positions when playing a game ...

WOOOOAAAHHH. big scary secret tactics ...

god you're an idiot

[IOMLS]

i was on the road and only a moderate favourite so i used "you can win this" ...

and as i maintained before ... and as i have tried and failed to convey to the other "it is the magic tactics silly" crowd that there is NO formation and tactic short of something truly idiotic like 11 keepers in a 1-1-8 that should result in such a skewed and improbable set of outcomes ....

and for the record for the other commenter who keeps telling me that me playing a 4-4-2 should somehow reduce my chance to win a match against anyone into the 3% range is ludicrous ... what better formation would there be for Van Der Saar, Evra, Vidic, Ferdinand, Haregraves (trained up to be accomplished at RB), Darijo Srna, Juan Manuel Vargas, Carrick, Danielle de Rossi, Rooney and Berbatov? those are A all of my best and highest rated players (all 3 stars of CA or above) ... and it gets them all on the field in the exact positions they are great at ... i mean seriously ... and they were in a 4-4-2 as well so there was no (oh they had 5 in the midfield so blah blah blah)

OK, yes that is probably the teamtalk I would give. What was the reaction to it?

Why the aggression? I wasn't suggesting that 4-4-2 was any worse than any other tactic to face them, I was just trying to understand the situation better. You are absolutely right there shouldn't be a situation where that team is only winning 1 out of 35. I'm a Tottenham fan and I think they are a bit over-rated on this game and some of the players are too effective (like Crouch being able to score more goals than games for example) but it still wouldn't lead to those kind of results.

Are the types of goals scored against you all different or are you conceding similar types of goals? For example if your full backs (particularly Evra) is getting forward a lot then that might be exploited by the opposition AI to a ridiculous amount. I had Lennon causing all sorts of trouble to opposition teams who had fullbacks attacking too much.

Even though it shouldn't have such a big effect there seem to be small tactical things that make a very big difference. What kind of goals to shots ratio are you having?

[ArseneLupin42]

god you're an idiot

yes. obviously putting those players into that formation makes me an idiot ... as opposed to not understanding the basic premise of "odds" and how the should and do relate to "outcome probabilty" ...

look i admit i have been a bit testy but in all seriousness ... you cannot argue that playing those players in a 4-4-2 formation, playing 'standard' to start and 'defensive' when up by 2 or more goals or in the last 10 minutes and 'offensively' when down i some sort of completely cruddy tactical mistake ... is there some "magic tactic" that would be better for Man Utd with these players that i am just not special enough to understand .. maybe but i doubt it ... but even so the above employed tactics under the given circumstances is no valid explanation for a result as far outside the A. initial "odds" given and B. so far outside even the realm of probabilty as was seen in the resulting data ... there is no logical or rational explanation for the result other than A. it was a one in a million event (which further matches run only 15 times thru showed is not the case) or that B. the 4-4-2 is super bad mojo tactic for man utd against spurs super duper fancy 4-4-2 tactic and the resulting 3% win probability (roughly) makes perfect sense given all of the facts ... that this pretty vanilla, straight forward sensible 4-4-2 alignment is somehow so awful it takes what at worst case scenario should be a 50/50 coin flip (despite the odds otherwise) and makes it 3/97 instead ... really? you want to stick with that?

i am in the camp that playing the whole formation backwards with 2 strikers in defence and 4 defenders up front shouldn't even skew the odds to 3% .. but that is just me in all of my "magic tactic denying" math and reality nerdiness i guess ....

[Pob]

Here's an easy way to prove the game isn't biased against the user (you)... play the game until you get to a situation like this (Man U vs Tottenham or similar ability), simulate the game 35 times again and get a similar result as before (1 win), but this time don't save over the game... upload the save, then I will play the game 100 times and show you the pre-game odds are achievable using no exploits of the match engine (e.g. corner bug) but just with sensible changing of tactics using just the touchline shouts and a standard tactic (no tweaks).

You are also misunderstanding the use of statistics... there may be 1000 other users who did a similar test to you who got sensible results from 35 games... you need to do a controlled test before you say you have concrete proof.

[x42bn6]

Well, in the last game we played Spurs, we were rather average as a 4-4-2 and we played better with 4-4-1 as Spurs collapsed.

In the 5-2 win before that, we went to a 4-2-4 almost with Caveman on for Nani, ripping them apart.

And the 0-0 draw before that saw a drab 4-4-2 made slightly better with 4-4-1-1, Giggs on for Caveman.

You don't give Spurs credit at all. They are a good side and in real life 4-4-2 hasn't worked terribly well.

[ArseneLupin42]

OK, yes that is probably the teamtalk I would give. What was the reaction to it?

Why the aggression? I wasn't suggesting that 4-4-2 was any worse than any other tactic to face them, I was just trying to understand the situation better. You are absolutely right there shouldn't be a situation where that team is only winning 1 out of 35. I'm a Tottenham fan and I think they are a bit over-rated on this game and some of the players are too effective (like Crouch being able to score more goals than games for example) but it still wouldn't lead to those kind of results.

Are the types of goals scored against you all different or are you conceding similar types of goals? For example if your full backs (particularly Evra) is getting forward a lot then that might be exploited by the opposition AI to a ridiculous amount. I had Lennon causing all sorts of trouble to opposition teams who had fullbacks attacking too much.

Even though it shouldn't have such a big effect there seem to be small tactical things that make a very big difference. What kind of goals to shots ratio are you having?

no no. sorry ... i was answering your question and the annoyance was for the "magic tactics make me super duper" crowd ... sorry for that

and the kind of goals i was conceding that led to my initial umbrage and frustration and thus initial urge to replay only to find other 'flukey' outcomes that led to the multiple iterations were goals like Van der Saar rolling the ball to Defoe who was 5 feet away from him so he could score ... and 35 yard screamers in 3 consecutive games from a guy with a 13 long shot rating ...

it wasn't that i lost one game ... that was just annoying and so i decided to 'cheat' by replaying the game ... it wasn't until it took 35 replays to win that i concluded something rotten in denmark was going on ...

but i am sure it was just my tactics ... i mean who would play those players in a 'standard' 4-4-2 and expect even up odds to beat the mighty tottenham ...

[uistbhoy]

4-4-2, kids down the park, jumpers for goalposts

marvellous

[ArseneLupin42]

Here's an easy way to prove the game isn't biased against the user (you)... play the game until you get to a situation like this (Man U vs Tottenham or similar ability), simulate the game 35 times again and get a similar result as before (1 win), but this time don't save over the game... upload the save, then I will play the game 100 times and show you the pre-game odds are achievable using no exploits of the match engine (e.g. corner bug) but just with sensible changing of tactics using just the touchline shouts and a standard tactic (no tweaks).

You are also misunderstanding the use of statistics... there may be 1000 other users who did a similar test to you who got sensible results from 35 games... you need to do a controlled test before you say you have concrete proof.

i am not saying that i conducted a super scientific experiment worthy of being published in a scientific journal ... i am saying that even given the 35 times i did run the match only winning once ... no matter what tactics or team talks or whatever i used is way outside the realm of probability and makes me, as the C & C music factory so eloquently put it, "go hmmmm"

are our really going to try and maintain that you are somehow so vastly superiour in your ingenious FM tactical acumen that someone like me could never figure out how and when to employ the ever so complicated touch line tactics of ... i don't know "control" when you go behind by one and want to attack a bit more but not go all out exposing yourself too much ... or that if it is late and you are still down you might want to use "attack" or ... gasp ... even "overload" ... i mean i assume that overload in the game has the same meaning as in the regular english language ... and let me guess ... if you have a one goal lead maybe "counter is a good touchline strategy? and if you are up by more than one or up late would using "defensive" or "contain" be what i should do Oh wise oracle of the FM tactical knowledge ... please enlighten me with your divine wisdom ... would you perhaps even throw me a pearl of wisdom of when to use "possess the ball" command when leading and wanting to keep the ball away from the opponent ... you know ... like in real life soccer

[midlothian]

Well, in the last game we played Spurs, we were rather average as a 4-4-2 and we played better with 4-4-1 as Spurs collapsed.

I think United were very good in the first hour as well, in a 4-4-2... absolutely deserved the lead.

[ArseneLupin42]

OK, yes that is probably the teamtalk I would give. What was the reaction to it?

Why the aggression? I wasn't suggesting that 4-4-2 was any worse than any other tactic to face them, I was just trying to understand the situation better. You are absolutely right there shouldn't be a situation where that team is only winning 1 out of 35. I'm a Tottenham fan and I think they are a bit over-rated on this game and some of the players are too effective (like Crouch being able to score more goals than games for example) but it still wouldn't lead to those kind of results.

Are the types of goals scored against you all different or are you conceding similar types of goals? For example if your full backs (particularly Evra) is getting forward a lot then that might be exploited by the opposition AI to a ridiculous amount. I had Lennon causing all sorts of trouble to opposition teams who had fullbacks attacking too much.

Even though it shouldn't have such a big effect there seem to be small tactical things that make a very big difference. What kind of goals to shots ratio are you having?

though on an entirely different note ... i will say that Crouchy was the object of the funniest part of the funniest comedy skit i have ever seen about football ... in which an average Joe was berating the English National Team for not getting the the European Championships and said to Crouch:

"Crouchy ... I look at you and everything tells me you should be rubbish at football ... but you're brilliant."

truer words were never spoken ... how that 6'5" frame of all gangly arms and legs is a world class footballer defies all human physiology and logic ...

[IOMLS]

no no. sorry ... i was answering your question and the annoyance was for the "magic tactics make me super duper" crowd ... sorry for that

and the kind of goals i was conceding that led to my initial umbrage and frustration and thus initial urge to replay only to find other 'flukey' outcomes that led to the multiple iterations were goals like Van der Saar rolling the ball to Defoe who was 5 feet away from him so he could score ... and 35 yard screamers in 3 consecutive games from a guy with a 13 long shot rating ...

it wasn't that i lost one game ... that was just annoying and so i decided to 'cheat' by replaying the game ... it wasn't until it took 35 replays to win that i concluded something rotten in denmark was going on ...

but i am sure it was just my tactics ... i mean who would play those players in a 'standard' 4-4-2 and expect even up odds to beat the mighty tottenham ...

It does seem a bit odd to me. I'd probably play 4-4-2 with that group of players too and as you said it probably shouldn't make that much difference with the quality of players you have. I did read somewhere a quote from someone who claimed that success of FM was 20% the players and 80% the tactics, i.e. with the right tactics you can take a poor group of players to the top. I guess the inverse is also possible, but those tactics shouldn't be bad enough for that to happen.

I guessed it might be that there was some loophole in your tactics, that if being used each time was being continually exploited, i.e. the attacking full back issue. There might still be, but if the goals are being conceded from different situations then it would seem a little odd. I'm guessing most settings are set to normal, so normal pressing and normal tackling etc. rather than soft tackling and stand off (which might lead to conceding long shots).

Also my other guess was that, as it seems there is a slight imbalance on the central to wide player goal contribution on the FM2010 demo that most of the goals are scored by AM's or Strikers. Depending on the role you have given your strikers this may lead to them shooting from distance too often and so they are ineffective. It is something that SI have already said they are looking into but maybe that has happened in this case.

If the setup is exploited by some tactic, then no matter how many times you repeat it, it is likely to have a similar end result.

[Pob]

I find you offensive, but I will put it down to your frustration. I've PM'd you.

[Pob]

Tactical exploits are no longer needed to do well in FM... man management is the key. The user has more ways than ever before to raise the morale of his players.

[grep]

Lupin, you are on the right side of the hill and I am with you, I anyway think that this thread will be closed like many other similar in the past with the excuse of being reiterate , offensive and all the BLA BLA the censure will like to label it.

So, let's fire your latest and best silver bullets now , be smart until the end, before this thread will be censured !!!!

[Pob]

There is a small chance that a user with the best team and good tactics and good man management could lose every game in a season. Just because you played a game 35 times doesn't mean you have produced any relavent statstics.

[davehanson]

I must still be missing something?

What exactly are you moaning about? Losing to Spurs as Man Utd, or the fact the odds were in your favour therefore, in your eyes, you must win in the majority of cases.

[ArseneLupin42]

It does seem a bit odd to me. I'd probably play 4-4-2 with that group of players too and as you said it probably shouldn't make that much difference with the quality of players you have. I did read somewhere a quote from someone who claimed that success of FM was 20% the players and 80% the tactics, i.e. with the right tactics you can take a poor group of players to the top. I guess the inverse is also possible, but those tactics shouldn't be bad enough for that to happen.

I guessed it might be that there was some loophole in your tactics, that if being used each time was being continually exploited, i.e. the attacking full back issue. There might still be, but if the goals are being conceded from different situations then it would seem a little odd. I'm guessing most settings are set to normal, so normal pressing and normal tackling etc. rather than soft tackling and stand off (which might lead to conceding long shots).

Also my other guess was that, as it seems there is a slight imbalance on the central to wide player goal contribution on the FM2010 demo that most of the goals are scored by AM's or Strikers. Depending on the role you have given your strikers this may lead to them shooting from distance too often and so they are ineffective. It is something that SI have already said they are looking into but maybe that has happened in this case.

If the setup is exploited by some tactic, then no matter how many times you repeat it, it is likely to have a similar end result.

i guess any of that is possible ... but i think my point is i should be able to throw those 11 guys out there and simply say "play however and wherever you like just win" and those 11 guys should beat the 11 from spurs more than 1 out of 35 times ... and i find it both amusing and insulting that given a simple set of factual data and a mild hypothesis that the game might use other factors than those you can control to effect the probability the response of some people is to basically say that "nah, you are just stupid and nowhere near as wise as the great and powerful Poobah me at the magic and mystical tactics of FM. if it had been me i would have employed some mystical and logic defying tactics known only to myself and my secret clan that would have changed the outcome ... the outcome could only be down to you being a complete boob and your use of a simple and straight forward 4-4-2, following the advice of the game and not using any tactic more adventurous that "more attack" when behind and "less attack" when ahead that caused what is in no way an alarming or improbable result of only 1 win out of 35"

of course i am paraphrasing ...

as for my fullbacks i put them on "defend" setting when facing a team that is averaging over 1.5 goals per game and on "automatic" when playing against those teams that average less ... i am talking other moderately equivalent Premiership or European sides of course and not FA cup games against League 2 scrubos in which i start the B or C squad ... silly me i assumed that the 'automatic' setting would somehow factor in my fullbacks fairly good "decision" and "anticipation" and "positioning" scores and cause them to get forward and back as would be most appropriate most of the time ... see how foolish and mortal i am, how misguided in the true zen of holy FM tactics

[Pob]

Is there anything anyone can do to convince you that the ME is not biased against the user? (The user actually has more freedom with tactics than the AI manager, so should be capable of better results than an AI manager.)

The only area in FM where the AI manager has an advantage is being able to see hidden attributes and knowing the minimum fee another team will sell a player for.

[davehanson]

but i think my point is i should be able to throw those 11 guys out there and simply say "play however and wherever you like just win" and those 11 guys should beat the 11 from spurs more than 1 out of 35 times ...

Right now I will say it then..............

It is your tactics. The pre-match odds are determined by numerous factors I guess, your teams form, injuries, your teams reputation etc, etc. What it doesn't do, and what it can't do is take into account the user making stupid decisions. You need to go and look at the tactic forum. You need to understand how to set a team up. It is no use just throwing a team out there, do you really think that happens in real life?

Also, you suggested you should be able to submit your team the day before thereby getting odds that are better? Again this doesn't happen in real life.

[IOMLS]

i guess any of that is possible ... but i think my point is i should be able to throw those 11 guys out there and simply say "play however and wherever you like just win" and those 11 guys should beat the 11 from spurs more than 1 out of 35 times ... and i find it both amusing and insulting that given a simple set of factual data and a mild hypothesis that the game might use other factors than those you can control to effect the probability the response of some people is to basically say that "nah, you are just stupid and nowhere near as wise as the great and powerful Poobah me at the magic and mystical tactics of FM. if it had been me i would have employed some mystical and logic defying tactics known only to myself and my secret clan that would have changed the outcome ... the outcome could only be down to you being a complete boob and your use of a simple and straight forward 4-4-2, following the advice of the game and not using any tactic more adventurous that "more attack" when behind and "less attack" when ahead that caused what is in no way an alarming or improbable result of only 1 win out of 35"

of course i am paraphrasing ...

as for my fullbacks i put them on "defend" setting when facing a team that is averaging over 1.5 goals per game and on "automatic" when playing against those teams that average less ... i am talking other moderately equivalent Premiership or European sides of course and not FA cup games against League 2 scrubos in which i start the B or C squad ... silly me i assumed that the 'automatic' setting would somehow factor in my fullbacks fairly good "decision" and "anticipation" and "positioning" scores and cause them to get forward and back as would be most appropriate most of the time ... see how foolish and mortal i am, how misguided in the true zen of holy FM tactics

I guess all I can say is that if it is the same set up each time then there seems to be some exploit in the system that is happening each time. As for the players doing there bit, I guess it depends on the rigidity of your tactics. How much freedom did you give to the players?

Out of interest did you have issues against any other teams or was this a one off?

[spankie]

Is there anything anyone can do to convince you that the ME is not biased against the user? (The user actually has more freedom with tactics than the AI manager, so should be capable of better results than an AI manager.)

The only area in FM where the AI has an advantage is being able to see hidden attributes and knowing the minimum fee another team will sell a player for.

I seem to recall a post from a SI bod saying that the AI doesn't actually "see" the hidden attributes, at least not when generating transfer targets for AI teams. Apparently they only see what we see.

I do suspect that there's an algorithm involved that may be better at working out the better players then some FM players are, myself included judging by some of the dross I've brought over the different FM versions.

[Pob]

I seem to recall a post from a SI bod saying that the AI doesn't actually "see" the hidden attributes, at least not when generating transfer targets for AI teams. Apparently they only see what we see.[/Quote]

I'd be interested in reading that post.... I thought it was accepted that the AI could see all attributes of players.

[grep]

He is not talking about an AI advantage or cheat , he is stating the routines are flawed cause the fate is already written , or at least not dependant on human player choices but from hidden routines instead.

I agree, we are dummies , even if hard to admit

[ArseneLupin42]

I must still be missing something?

What exactly are you moaning about? Losing to Spurs as Man Utd, or the fact the odds were in your favour therefore, in your eyes, you must win in the majority of cases.

i am doing neither. i am pointing out that losing any ONE game to Spurs as Man Utd under the given circumstances is nothing too alarming or outside what could be expected

i just happend to be annoyed at the "way" i lost the first game on two completely flukey goals and so i thought "well heck i saved right before this" so i decided to play it over and get a win ... because i hate spurs and cannot tolerate losing to them even in a game ... lo and behold another loss filled with flukey goals ... another replay ... another loss ... after 5 i thought ... hmmm ... i am not doing anything outlandish or foolish with my tactics ... i am following the prematch advice to the letter ... now i am intriged, by the prematch odds (knowing that they are not exact but only ball park and other factors and decisions by me can move them up or down from that 'estimate') i should have one at least one of these 5 i wonder how many it will take for me to win ... the answer was 35 ... which all i am saying is shockingly improbable not given some strange goings on in the mathematical formula used by the game engine .. whether that is some factors that a manager cannot control being "factored" in (like the longer your winning or unbeaten streak some increasing probability of losing is factored in ... not an entirely outlandish notion) or whether it is just a wildly poor or whacked out combination of the controllable in game factors (like the "it is the tactics crowd is saying" ... it is absurd to think that some tactic over some other tactic, excluding radically bad ideas like starting 11 goalkeepers, should result in such an imbalanced probability outcome ... keeping all of the factors the same (and fairly standard) two evenly matched teams (and lets just pretend for the sake of argument these were) should come out roughly 50/50 in results or somewhere in the normal 3 standard deviation bell distribution over a high number of iterations ... that did not happen .... something caused that to happen .. i wonder what ... and i reject the idea that "poor" tactics could or should lead to such an outcome ... the tactics and players i employed were in no way, obvious to an honest and objective player of both this game and even moderately informed fan of football, so bad as to deserve this outlandishly improbable mathematical outcome ... and if in fact it was "my tactics" that led to this observed outcome, given i have fully described that i took a pretty basic/standard and reasonable tactical approach, is in and of it self the problem ... as nothing short of lining up all 17 year old 0 star rated scrubs, having them play nekkid and out of position with the opposite of what would be "sane" tactics should result in only 1 win out of 35 ... i am of course engaging in a bit of hyperbole but you get my meaning i am sure

[davehanson]

i am doing neither. i am pointing out that losing any ONE game to Spurs as Man Utd under the given circumstances is nothing too alarming or outside what could be expected

i just happend to be annoyed at the "way" i lost the first game on two completely flukey goals and so i thought "well heck i saved right before this" so i decided to play it over and get a win ... because i hate spurs and cannot tolerate losing to them even in a game ... lo and behold another loss filled with flukey goals ... another replay ... another loss ... after 5 i thought ... hmmm ... i am not doing anything outlandish or foolish with my tactics ... i am following the prematch advice to the letter ... now i am intriged, by the prematch odds (knowing that they are not exact but only ball park and other factors and decisions by me can move them up or down from that 'estimate') i should have one at least one of these 5 i wonder how many it will take for me to win ... the answer was 35 ... which all i am saying is shockingly improbable not given some strange goings on in the mathematical formula used by the game engine .. whether that is some factors that a manager cannot control being "factored" in (like the longer your winning or unbeaten streak some increasing probability of losing is factored in ... not an entirely outlandish notion) or whether it is just a wildly poor or whacked out combination of the controllable in game factors (like the "it is the tactics crowd is saying" ... it is absurd to think that some tactic over some other tactic, excluding radically bad ideas like starting 11 goalkeepers, should result in such an imbalanced probability outcome ... keeping all of the factors the same (and fairly standard) two evenly matched teams (and lets just pretend for the sake of argument these were) should come out roughly 50/50 in results or somewhere in the normal 3 standard deviation bell distribution over a high number of iterations ... that did not happen .... something caused that to happen .. i wonder what ... and i reject the idea that "poor" tactics could or should lead to such an outcome ... the tactics and players i employed were in no way, obvious to an honest and objective player of both this game and even moderately informed fan of football, so bad as to deserve this outlandishly improbable mathematical outcome ... and if in fact it was "my tactics" that led to this observed outcome, given i have fully described that i took a pretty basic/standard and reasonable tactical approach, is in and of it self the problem ... as nothing short of lining up all 17 year old 0 star rated scrubs, having them play nekkid and out of position with the opposite of what would be "sane" tactics should result in only 1 win out of 35 ... i am of course engaging in a bit of hyperbole but you get my meaning i am sure

Ok, could you upload a saved game or at least the tactic you used against spurs so we can see?

[Pob]

say the prematch odds were evens to win (50% chance of winning). If you replayed that game 35 times, the chance of you not winning the first 34 games would be 0.5^34 = 1 in 17 billion (very unlikely). However, the prematch odds assume you are as good as an AI manager; factor in what we will call the ArseneLupindoesn'tunderstandhowtowin factor which changes the real odds of you winning to 1 in 4 games; the chances of you losing the first 34 games you play would be 0.75^34 = 1 in 17 thousand. I am sure that this forum gets at least 17000 users each day who play the game as badly as you, so the odds do pan out

[ArseneLupin42]

I guess all I can say is that if it is the same set up each time then there seems to be some exploit in the system that is happening each time. As for the players doing there bit, I guess it depends on the rigidity of your tactics. How much freedom did you give to the players?

Out of interest did you have issues against any other teams or was this a one off?

that is my point exactly ... there shouldnt be ANY tactic or anti tactic or exploit that is so powerful that it should come up with a 1 win in 35 for two ... let us call them ... evenly matched teams ... and if there is ... and the data would suggest so ... there is a problem ... or if not a "problem" an interesting bit of observation ... i understand there are a plethora of ways a manager can effectively tweak the probability of the match engine outcome in his favour .... beyond just how good your players are vs the other players ... but nothing should tweak the probability outcome that far ... short of doing something epically stupid like starting 11 goalkeepers or playing a 1-1-8 ... and even then i am not so sure that 1 out of 35 is a sensible outcome ... kidding a bit

and on the flip side nothing you do (or the AI does) should be able to flip the match engine winning probability odds that far in your/its favour ...

and for the record i pretty much just use 'default' for things like creative freedom and rigidity and really mostly enjoy scouting and just getting the best possible palyers in the best possible positions and then just enjoying the game simulation like watching the match for real ... doing obvious and not outlandish touchline tweeks like "attack" when i am behind and "defend" if ahead ...

i also more enjoy the shuffling of players, watching young kids develop and throwing them out there in cup games and the like and watching them come thru ...

like i said earlier ... it isn't as if the game tactics are so obscure and inscrutable that a reasonably intelligent person with a moderately good knowledge of actual football (which i assume anyone nerdy enough to play FM is) cannot figure them out ... i mean seriously

[erbgor]

I'm utterly amazed at how many people misunderstood the OP. His main point was NOT betting odds or regens. He played with Man United against Spurs, repeatet the match 35 times and only won 1 out of 35 tries. He finds this very odd.

Arsene, I'm curious about a few things:

A How did you watch the matches? On "key moments" setting?

B What kind of goals did you concede?

C How often did you manage to score the first goal?

D How many opportunities did you create - usually far less than Spurs?

I'd be grateful for estimates, I'm a bit irritated about some od my results too, though I never reloaded a game. FYI, I played Arsenal over too half seasons and seemed to get a lot of "fluke" losses, i.e 20:6 shots for me, still lose. (Now, before someone jumps at my throat, yes, these kind of results happen, and I'm glad they happen in this game too, just seemed a bit much for me. No, this is not a proof for anything.)

Speculating about the cause of your results, I'd say it's either just coincidence, which is very unlikely, or the game intentionally (cheating to streamline the game experience) or unintentionally (some sort of bug) screwing you over.

I'd be interested in more test data. Maybe I'll do some myself (probably only at "key highlights", though ).

[ArseneLupin42]

Ok, could you upload a saved game or at least the tactic you used against spurs so we can see?

i could ... or i could just describe it in painstaking detail as i have pretty much throughout this thread ... i will say this again ... the tactics are not wizardry people .... they have little diagrams that show you how good your player is graphically compared to other players ... they show what positions they play best in with little green dots ... you put them in those spots ... you put as many really good players on the field as you can ... you follow most of the 'advice' the game gives ... you use common sense and football knowledge to do things like call for more "possession" when you are ahead and want to kill off a game ... it really isn't rocket science ... no offence meant to those who think they are somehow blessed by the holy power of tactical knowledge from the great grand FM tactic deity in the sky ...

[Pob]

I'm not misunderstanding... if he has a real chance of winning a game 1 in 4 (fairly realistic considering how poorly he plays), getting 34 consecutive "non-wins" is not that surreal considering he is one of thousands of people using this forum.

[erbgor]

Playing poor? What makes you think that?

[Pob]

Playing poor? What makes you think that?

His logic skills!

[TheStoneMonkey]

Lol, aspie thread.

Hey, if I put 2+3 into a calculator 35 times, will the result change?

[ArseneLupin42]

I'm utterly amazed at how many people misunderstood the OP. His main point was NOT betting odds or regens. He played with Man United against Spurs, repeatet the match 35 times and only won 1 out of 35 tries. He finds this very odd.

Arsene, I'm curious about a few things:

A How did you watch the matches? On "key moments" setting?

B What kind of goals did you concede?

C How often did you manage to score the first goal?

D How many opportunities did you create - usually far less than Spurs?

I'd be grateful for estimates, I'm a bit irritated about some od my results too, though I never reloaded a game. FYI, I played Arsenal over too half seasons and seemed to get a lot of "fluke" losses, i.e 20:6 shots for me, still lose. (Now, before someone jumps at my throat, yes, these kind of results happen, and I'm glad they happen in this game too, just seemed a bit much for me. No, this is not a proof for anything.)

Speculating about the cause of your results, I'd say it's either just coincidence, which is very unlikely, or the game intentionally (cheating to streamline the game experience) or unintentionally (some sort of bug) screwing you over.

I'd be interested in more test data. Maybe I'll do some myself (probably only at "key highlights", though ).

wow. look at this. an actual understanding of what i was getting at and some intelligent and probing questions ... well wonders never cease

A. i watched the match on key moments settings ... i always do, and after 35 even that was too much

B. like i said earlier the whole thing started when in the first game and first couple of replays i was giving up really weird goals ... like van der sar saving a shot ... walking out into his penalty box and calmly rolling the ball directly to Defoe 5 feet away when no defender was in the vicinity ... or 35 yard screamers ... more than once ... from a player with a 13 long shot rating ... of course that is anecdotal and i did not obviously record every goal

C. this is only an estimate from memory but i cannot remember scoring the first goal in more than 5 or 6 matches ... which i found very odd against a team that had scored 17 in 14 matches ... and given up something like 24-27

D. a preponderance of the time i way way way outshot Spurs ... one sticks out particularly where it was 3 - 0 at the half and Spurs had scored 3 goals on 3 shots none of which counted as "clear cut chances" and i had like 19 shots with 4 clear cut chances etc ... again not proof or anything just anecdotal ... and i especially love the games where all of what shows in the game highlights shows you 'dominating' the possession stats show you winning like 65 - 35 or more and the shots being even or you even get outshot .... i mean how often does that really happen? it seems like that would be a pretty mean feat to pull off ..

and i really cannot say exactly WHAT is going on .... only that such data (and subsequent games that took 15 or more replays to get a win against even more decidedly "underdog" teams) shows that something IS going on in the formula and game engine that is more than just those factors that the game describes and that a manager can control that should/do go into the game result factoring ... or that one of these managerial controllable factors (like the magic super obtuse and inscrutable "tactics") is factoring in way way way to much into the outcomes in certain circumstances ... i tend to think that some sort of programming streamline was needed to prevent "long winning streaks" from being a common occurrence and i can totally see some factor that increases the longer you go unbeaten or winning that causes a decreasing probability of a win outside any of the usual match probability factors ... that is just a conjecture of course and not provable

[Pob]

i tend to think that some sort of programming streamline was needed to prevent "long winning streaks" from being a common occurrence and i can totally see some factor that increases the longer you go unbeaten or winning that causes a decreasing probability of a win outside any of the usual match probability factors ... that is just a conjecture of course and not provable

And doesn't stack up as there have been other users with very long win streaks.

[backpackant]

This is just going round in circles now, so will simply bow out with these words and count the minutes until this is closed.

Thanks to the OP for replying to my earlier message, but I feel you are not testing the game fairly in any way. To use the very same tactics every time is very flawed. The game simply HAS to have a limited basis for calculating match outcomes, it's not possible to programme something that is absolutely accurate to real life. I don't think anyone - fanboys included - would try to pretend FM mirrors real life exactly. It can't. Playing the same game over and over with the exact same tactics and players will always produce similar results. Look at how the ME works:

1st, the ME looks at the two lineups and tactics, and once the teams and tactics are submitted - before a ball is kicked - it makes thousands of calculations to work out what the score will be, say 2-0 to Man U, and displays these calculations in the form of pretty graphics for your viewing pleasure. Within that calculation, there will be certain variables every time, such as injuries.

2nd, whenever the human user or AI manager alters their tactics, it makes ANOTHER set of caluclations and works out what the score will be based on the new information and puts THAT on screen instead of its original calculations. If Man U (human) was having all the possession, then the AI Rednapp will make an adjustment, which may now mean that the score will end up 2-0 to Spurs. If Man U do not change anything and Spurs are winning, and don't change their tactics either, then that's how the game will play out.

So if your first game ends in a loss, it simply means that in that first game Rednapp predicted your tactics, came up with a counter-tactic, and beat you. When you replay with the exact same tactic, it is MORE likely that Spurs will win again.

If you were saying that no matter what tactics you use you can never beat them, then it may be worth carrying on with this, just to prove that "it's your tactics". Clearly this isn't your problem, but rather you seem to be of the unshakeable impression that FM should be exactly like real life and therefore something is very wrong with the game. Yes, players should have an influence, but in your example it you are replicated all the same conditions every single time - mentality, morale, determination, formation, tactics, and so on.

Okay, in summary, if the point of this thread is to say the ME is somewhat limited and relies more heavily on tactics than real life football, I tip my hat and say I agree. If you are saying this makes FM a bad game or that it's a mistake of some sort, I humbly disagree - it's a computer game, and as such it is limited by physics.

Hope you manage to enjoy FM in the future.

[ArseneLupin42]

Lol, aspie thread.

Hey, if I put 2+3 into a calculator 35 times, will the result change?

need i bother to point out that your analogy is entirely off base?

in the case of 2 + 3 = 5; you are talking about a static set of mathematical rules which dictate that through any number of infinite iterations of this function you are supposed to, by definition get the same result (tho interestingly enough there are mathematical systems, like Non-Euclidian geometry where in fact 2 + 3 does not equal 5 but that is neither here nor there)

in the case of simulating a real life situation whose outcome is not predetermined by mathematical definition you are NOT supposed to get the same result. in real life the same 22 players could play under the exact same conditions using all of the exact same formations and tactics generally and not get the same result every time ... in fact if they did over a large number of matches that would downright defy probability ... especially if those teams were evenly matched ... thus the game simulation should seek to do the same thing ... taking some starting probability of winning, adjusting that for controllable factors in the game either positively or negatively and using some mathematically permutation to account for what would be naturally occuring 'chance' and 'variation' on the level of individual detail in real life ... all of which means you should be able to run the game simulation 100 times with the exact same set up (no matter what that set up might be) and not get the same result 100 times but rather over 100 times some distribution of results that reflects some "probability" of winning (ie the probability or chance you had of winning any one of those particular iterations given the controlable influencing factors plus sheer chance and variation) ... all i am saying is there should be no confluence of controllable factors that should result in two evenly matched teams having their probability distribution skewed from 50/50 to 3/97 (short of say starting only 9 men) ...

[Pob]

I have shown how the probability distribution can appear to skew to that extent using your test due to your playing ability and the number of users on this forum. (see post #131) You are wrong. I can go into more detail if you like. I could create a thread in the Experiments forum if you still insist you are right.

[ArseneLupin42]

This is just going round in circles now, so will simply bow out with these words and count the minutes until this is closed.

Thanks to the OP for replying to my earlier message, but I feel you are not testing the game fairly in any way. To use the very same tactics every time is very flawed. The game simply HAS to have a limited basis for calculating match outcomes, it's not possible to programme something that is absolutely accurate to real life. I don't think anyone - fanboys included - would try to pretend FM mirrors real life exactly. It can't. Playing the same game over and over with the exact same tactics and players will always produce similar results. Look at how the ME works:

1st, the ME looks at the two lineups and tactics, and once the teams and tactics are submitted - before a ball is kicked - it makes thousands of calculations to work out what the score will be, say 2-0 to Man U, and displays these calculations in the form of pretty graphics for your viewing pleasure. Within that calculation, there will be certain variables every time, such as injuries.

2nd, whenever the human user or AI manager alters their tactics, it makes ANOTHER set of caluclations and works out what the score will be based on the new information and puts THAT on screen instead of its original calculations. If Man U (human) was having all the possession, then the AI Rednapp will make an adjustment, which may now mean that the score will end up 2-0 to Spurs. If Man U do not change anything and Spurs are winning, and don't change their tactics either, then that's how the game will play out.

So if your first game ends in a loss, it simply means that in that first game Rednapp predicted your tactics, came up with a counter-tactic, and beat you. When you replay with the exact same tactic, it is MORE likely that Spurs will win again.

If you were saying that no matter what tactics you use you can never beat them, then it may be worth carrying on with this, just to prove that "it's your tactics". Clearly this isn't your problem, but rather you seem to be of the unshakeable impression that FM should be exactly like real life and therefore something is very wrong with the game. Yes, players should have an influence, but in your example it you are replicated all the same conditions every single time - mentality, morale, determination, formation, tactics, and so on.

Okay, in summary, if the point of this thread is to say the ME is somewhat limited and relies more heavily on tactics than real life football, I tip my hat and say I agree. If you are saying this makes FM a bad game or that it's a mistake of some sort, I humbly disagree - it's a computer game, and as such it is limited by physics.

Hope you manage to enjoy FM in the future.

if in fact the game engine simply takes a simple set of inputs like you say, and cranks out the same formulae that does not account properly for chance/random variation than that in and of itself is a/the flaw ... any good simulation should understand and try to recreate mathematically this very 'chance' and 'variation' in the real event ... no two football matches are the same ... even if the same players and same formations and same tactics and same overarching conditions and circumstances are used (ie the things you can control as a manger in the game) ... a real football game is a series of millions of small chance, variation and decision within even the exact same larger setup and circumstances ... all of which means that the probability (and thus the outcome percentages of a large number of iterations of the game) of winning needs to accurately reflect this. Which means two things ... to be a good simulation you should A. be able to put in the same exact controls and get a different result due to the program accurately modeling this very "chance" and "variation" ... and the probability of this varied outcome (either a win, draw or loss) should be rationally and logically tied into various factors that the manager can understand and control ... and there is no set of input factors (ie the manager tactics, etc) that should alter the probability of simulating two evenly matched teams (and thus roughly 50/50 win probability) and radically alter those to 3/97 short of something entirely radical and egregious ... and it is certainly NOT the case that the tactics and other factors that I controlled for this match in question were anywhere near "bad" enough to have reasonably caused such a radical shift in the probability as it did ....

[Braundjoh]

Wait. So when you lost the first time, why did you replay it? Because the game beat you?

[Pob]

any good simulation should understand and try to recreate mathematically this very 'chance' and 'variation' in the real event ... no two football matches are the same ...

It does account for random football events, the ME selects a 'seed' at the start of each game to base the game on; whenever you alter your tactics midmatch (or reload a game and play it again), it selects a new seed. The random events happen because of the seed selected, the ME then adds in all of the variable factors like player attributes, morale, tactics, etc. to decide what should happen at each event. (Each event lasts about a minute of game time).

Some Football manager games (e.g. LMA manager) didn't select a new seed on reloading the game, so you could replay the same seed game many times until you won the game, changing your tactics to load a new seed at appropriate moments (e.g. when you scored, but without changing would have conceded).

[TheStoneMonkey]

It's your tactics.

[ArseneLupin42]

I have shown how the probability distribution can appear to skew to that extent using your test due to your playing ability and the number of users on this forum. (see post #131) You are wrong. I can go into more detail if you like. I could create a thread in the Experiments forum if you still insist you are right.

you are hilarious ... please explain to mean any one or combination of "tactics" and "factors" that i was sooo bad at that you think would reasonably decrease my probability of winning a match as described so low as to 1/35 being a reasonable and statistically non anomalous result ...

please o mighty wizard of FM tactics ... was it my incredibly high morale on each and every starting player that was such "poor" managing on my part?

maybe it was my radical Team Talk of "you can win this game" for a game in which i was only slightly favoured by the "prematch odds" on the road?

could it be my totally outlandish and completely nonsensical 4-4-2 formation with highly skilled/stated/attributed players at peak fitness filling each traditional position in said formation according to which position they were most skilled in?

or perhpas it was my highly controversial use of those "automatic" positional settings and "standard" touchline tactics except for those times in the game when i was ahead or behind by alot or late where i did the unthinkable and moved the touchline tactics either to "attack" when behind or "defend" when ahead ... how foolish of me

please enlighten me with your wisdom o great one ... how would you have altered my tactics and formation from the details i have earlier in the thread ...

what a conceited egomaniac you are ....

[Pob]

please explain to mean any one or combination of "tactics" and "factors" that i was sooo bad at that you think would reasonably decrease my probability of winning a match as described so low as to 1/35 being a reasonable and statistically non anomalous result ...

You are misunderstanding the probability. If you have a one in four chance of winning a game, fairly reasonable considering what you have shown, not winning 34 games in a row is not an anomalous result. It would occur approximately 1 in 17 thousand tests for you; as I said, this forum probably has that many users playing at your ability level.

I am not going to lecture you on how to play the game better; you won't listen anyway.

[ArseneLupin42]

It does account for random football events, the ME selects a 'seed' at the start of each game to base the game on; whenever you alter your tactics midmatch (or reload a game and play it again), it selects a new seed. The random events happen because of the seed selected, the ME then adds in all of the variable factors like player attributes, morale, tactics, etc. to decide what should happen at each event. (Each event lasts about a minute of game time).

Some Football manager games (e.g. LMA manager) didn't select a new seed on reloading the game, so you could replay the same seed game many times until you won the game, changing your tactics to load a new seed at appropriate moments (e.g. when you scored, but without changing would have conceded).

well yes. i would hope it does. please try telling that to the "but if you used the exact same tactics of course you got the same results" crowd ...

what i am saying is whatever factors they are using ... be they the obvious and controllable ones .... or as yet undefined background factors ... took what at worst was a 50/50 game ... in which i think that any fair minded person can say that i did not by any means use "wretched" or "abysmal" tactics (and which i might say were in fact very well suited to A. the instructions given by the game and also well suited to the players and situation) ... to say that somehow the tactics i described in detail are some "radical" and "completely" rubbish approach is utter nonsense ... even if i concede they are not the lofty uber-tactics of some of the very special FM holy men on here ... i think anybody can fairly say that they at ABSOLUTE WORST were ... what net overall neutral in effecting the "seed" probability or let's say even slightly negative in effecting that probability ... that still does not account for the results i got unless something else is at work ... either A. an unseen factor or B. my tactics which to all reasonable football fans and FM players were pretty acceptable did in fact for some unknown reason have a dramatically negative effect ... either way i see that as something interesting to note and a flaw ... maybe one that if you only play each game once and don't really care about the bigger concept of how accurately the game is simulating reasonable probabilities doesn't concern you but i am not one of those kind of people so it does ... not concern but intrigue me