Jump to content

Key Tackles, Clear Cut Chances, Dribbles, Key Passes etc. All Broken!


jdmk88
 Share

Recommended Posts

  • Replies 113
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

1 hour ago, bielsadidnothingwrong said:

I plan on loading up some older versions of FM and finding out just how bad it is this year vs others. 

In FM19 there are 1 or 2 less important stats that suffer from this. There is also a difference on some of the stats discussed in this thread but the difference is much smaller and is manageable.

The obvious solution to this problem is for SI to make the QME mimick FME stats instead of trying to make both the FME and QME  mimick the real match stats; so im happy SI are looking into this. This is the only way FME  and QME will be completely comparable.

Link to post
Share on other sites

As I have demonstrated, the QME absolutely fails at replicating real-life stats, so I agree with your assessment.

I’m curious why they consider what the QME does right now as mimicking real-life if that’s the case. Unless we are to understand this was not looked at by the team at SI until one of us noticed it, which emphasizes my point that we have bore the brunt of finding these flaws at the expense of our enjoyment and experience. 

Link to post
Share on other sites

  • SI Staff
On 17/01/2021 at 00:10, bielsadidnothingwrong said:

Here's screenshots of my statistical discrepancies @Jack Joyce, you mean to tell me you're satisfied with this end product? 

The first are the key passes in my league, Vanarama South. As you can see, Jacob Mendy lead the league with 193 key passes. Pretty decent work by Mendy I say, especially compared with the Vanarama North leader's 144. Looking at how he compared with the Premier League, Mendy even bests Kevin de Bruyne! I am truly blown away by his performance, but how does he compare to league leaders elsewhere? Jacob Mendy scorched the league leaders of La Liga, Bundesliga, and Ligue 1 combined! I wonder how much these results mirror reality... KDB had 131 key passes last season. While 182 vastly eclipses it, its not that far off. Thomas Muller last season crafted 82 key passes, but digital Thomas Muller was shackled by the match sim engine and could only muster a dull 28. Yikes. Not only do the stats look awful when compared to other leagues in the game, they look even worse here.

I'm sorry, but to dismissively consider these differences as mere "statistic discrepancies" is an insult to your customer base. These are HUGE flaws with BASIC features in the game. While none of us expect you to talk-down SI or the developers, acknowledging the severity of this issue would at least buy you some good-will with us.

It is infuriating that you at SI decided incorporating xG in FM21 was more important than some other "legacy" features which have remained broken since the beta release. Things like analyst reports, player heat maps, passing networks, player-count dropoff, the youth intake bug, and now this have all had some explanation related to the introduction of xG. Clearly, SI had ambitious visions for FM21 that included successfully incorporating it in the game, but unfortunately you all failed at implementing them bug free due to COVID or otherwise. Honestly, that's quite alright, however, you send people on here to explain away all these legitimate issues as if there's something we the players are doing wrong (" If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail (obviously this does use extra processing so be aware of that). This will ensure that the selected matches are fully played as opposed to being simulated.") rather than acknowledging that there have been many unforeseen consequences implementing xG into the statistical system that has ultimately caused you to release a product which your customer base has bore the brunt of finding major flaws (some of which are obvious within an hour of playing and some, like this, that come later). 

I half-expect this critique to be removed for its honesty, but I am not going to treat you all with a light touch since that apparently gets met with disingenuous replies that do not sufficiently address the concerns of the posters and instead seek to place the blame on us for having expectations that the FME and QME produce similar results.

Sorry but you're completely misrepresenting what I said:

"Stats discrepencies between FME and QME - We will look at ways to improve this to bring them closer together, but in reality there will always be some difference between the two as a fully played match will always be at least slightly different to a quick ME simulated match. If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail (obviously this does use extra processing so be aware of that). This will ensure that the selected matches are fully played as opposed to being simulated."

I clearly stated that we are looking at ways to bring the stats closer together, so there's no need for exaggerated (and inaccurate) accusations like "you mean to tell me you're satisfied with this end product?" or "to dismissively consider these differences as mere "statistic discrepancies" is an insult to your customer base". It's needlessly antagonistic which doesn't help anybody.

"however, you send people on here to explain away all these legitimate issues as if there's something we the players are doing wrong 'If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail'"

Again, we've not dismissed these complaints at all and have already said we're looking into it. But regardless of any potential changes, my statement is true. Even if we were to use FME generated data to simulate the QME stats, those stats are still being 'spoofed' and if you're comparing those stats directly against FME generated data, then it's always going to be less reliable than directly comparing two sets of data generated in the FME. That was the point I was making.

Thanks,

Jack

Link to post
Share on other sites

7 hours ago, Jack Joyce said:

Even if we were to use FME generated data to simulate the QME stats, those stats are still being 'spoofed' and if you're comparing those stats directly against FME generated data, then it's always going to be less reliable than directly comparing two sets of data generated in the FME.

Yes definitely it wont be as reliable as FME vs FME 

Link to post
Share on other sites

Going through the player detailed stats comparing full detail with no detail/quick sim and it feels like the below are issues between the two:

1074267385_Screenshot2021-01-18214053.png.a28c77038a1aca3763d5bf402a541114.png

Some of them I don't really know if it's an issue with the Full or Quick because to be honest I don't really know what us as users are supposed to be expecting.

Link to post
Share on other sites

2 hours ago, yanloup64 said:

We only expect given stats to be something we can trust, that's it. Obviously 90 clearances per games is a joke, as the various discrepancies are too, which cause teams and/or players very difficult to assess. 

Yes, sorry, of course we want reliable stats. I meant more along the lines of I'm not sure which end of the spectrum we should be seeing for some of them (to know whether full or quick are more expected within the game), e.g. key tackles. I would have said key tackles were far too low on full detail (still think they are - but most clear cut chance defences are probably blocks rather than tackles), but I never knew the SI 'definition' of what a key tackle was until I read the response, so now I doubt/query what some of the others are supposed to represent, and how frequently they're likely to happen. 

Link to post
Share on other sites

22 hours ago, Jack Joyce said:

Sorry but you're completely misrepresenting what I said:

"Stats discrepencies between FME and QME - We will look at ways to improve this to bring them closer together, but in reality there will always be some difference between the two as a fully played match will always be at least slightly different to a quick ME simulated match. If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail (obviously this does use extra processing so be aware of that). This will ensure that the selected matches are fully played as opposed to being simulated."

I clearly stated that we are looking at ways to bring the stats closer together, so there's no need for exaggerated (and inaccurate) accusations like "you mean to tell me you're satisfied with this end product?" or "to dismissively consider these differences as mere "statistic discrepancies" is an insult to your customer base". It's needlessly antagonistic which doesn't help anybody.

"however, you send people on here to explain away all these legitimate issues as if there's something we the players are doing wrong 'If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail'"

Again, we've not dismissed these complaints at all and have already said we're looking into it. But regardless of any potential changes, my statement is true. Even if we were to use FME generated data to simulate the QME stats, those stats are still being 'spoofed' and if you're comparing those stats directly against FME generated data, then it's always going to be less reliable than directly comparing two sets of data generated in the FME. That was the point I was making.

Thanks,

Jack

Really pleasing we have someone to communicate with but I do feel like you’re ignoring/not answering direct questions. Me, everyone here and you I hope see the problem. There is a clear problem for all to see, when can we expect a hotfix/new patch?

I understand you might not have the answers so I don’t really wanna shoot the messenger but it would help if you could keep us in the loop of things so we all know what’s going on cos of lot of peoples game are on pause right now and I feel you’re just coming on here when you’re misquoted and not for the people if you like. 
 

I don’t mean to come across rude or anything at all as I know your job has been hard as it is.

Link to post
Share on other sites

On 18/01/2021 at 12:36, Jack Joyce said:

Sorry but you're completely misrepresenting what I said:

"Stats discrepencies between FME and QME - We will look at ways to improve this to bring them closer together, but in reality there will always be some difference between the two as a fully played match will always be at least slightly different to a quick ME simulated match. If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail (obviously this does use extra processing so be aware of that). This will ensure that the selected matches are fully played as opposed to being simulated."

I clearly stated that we are looking at ways to bring the stats closer together, so there's no need for exaggerated (and inaccurate) accusations like "you mean to tell me you're satisfied with this end product?" or "to dismissively consider these differences as mere "statistic discrepancies" is an insult to your customer base". It's needlessly antagonistic which doesn't help anybody.

"however, you send people on here to explain away all these legitimate issues as if there's something we the players are doing wrong 'If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail'"

Again, we've not dismissed these complaints at all and have already said we're looking into it. But regardless of any potential changes, my statement is true. Even if we were to use FME generated data to simulate the QME stats, those stats are still being 'spoofed' and if you're comparing those stats directly against FME generated data, then it's always going to be less reliable than directly comparing two sets of data generated in the FME. That was the point I was making.

Thanks,

Jack

I'm surprised that the FME vs QME has generated so much debate! They are two different MEs designed to do two different things and so will never be directly comparable. Not sure what is so controversial about this? Also not sure what was controversial about explaining how to essentially just use one ME if you are really keen to have directly comparable stats and are willing to sacrifice performance / have the computer set up to handle it.

Recognise that you'll want to play it safe, @Jack Joyce... but how confident are you guys that a fix for registering the missing match stats in the FME isn't too far away?

Link to post
Share on other sites

5 hours ago, yorkie87 said:

Are the things like key tackles, dribbles etc happening in game but just not showing up on analytical data? Or are they not happening at all?

This is what I think is happening. I've seen it several times where my winger goes past a defender, it's acknowledged in commentary and nothing pops up on the stats. I have seen a few posts mentioning that stats and post-match analysis don't match up as well.

Edited by jdmk88
Link to post
Share on other sites

On 18/01/2021 at 12:36, Jack Joyce said:

Sorry but you're completely misrepresenting what I said:

"Stats discrepencies between FME and QME - We will look at ways to improve this to bring them closer together, but in reality there will always be some difference between the two as a fully played match will always be at least slightly different to a quick ME simulated match. If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail (obviously this does use extra processing so be aware of that). This will ensure that the selected matches are fully played as opposed to being simulated."

I clearly stated that we are looking at ways to bring the stats closer together, so there's no need for exaggerated (and inaccurate) accusations like "you mean to tell me you're satisfied with this end product?" or "to dismissively consider these differences as mere "statistic discrepancies" is an insult to your customer base". It's needlessly antagonistic which doesn't help anybody.

"however, you send people on here to explain away all these legitimate issues as if there's something we the players are doing wrong 'If you want stats that are 100% directly comparable you can use the detail level menu from within your save game to set your desired leagues to be played in full detail'"

Again, we've not dismissed these complaints at all and have already said we're looking into it. But regardless of any potential changes, my statement is true. Even if we were to use FME generated data to simulate the QME stats, those stats are still being 'spoofed' and if you're comparing those stats directly against FME generated data, then it's always going to be less reliable than directly comparing two sets of data generated in the FME. That was the point I was making.

Thanks,

Jack


Everyone understands that QME stats and FME stats will never be 100% directly comparable, it's always been like that on FM. I think the majority just wanted acknowledgment of how whacky some of these FME stats are.

Edited by jdmk88
Link to post
Share on other sites

5 hours ago, yorkie87 said:

Are the things like key tackles, dribbles etc happening in game but just not showing up on analytical data? Or are they not happening at all?

The match engine is fine, its just how the statistics are calculated could do some work. E.g. the definition of what a key tackle is means key tackles basically never happen, if the definition was more lenient then more key tackles would be recorded.

I will say though the avg dribbles do seem to be too low but I havent played enough of FM21 to know if players are actually dribbling more than the statistics are reporting.

Link to post
Share on other sites

Am I the only one experiencing the opposite about key passes since the last patch? They were definitely much more inflated compared to the previous years before the patch, but since the patch I observe other strange things about key passes that make me think that stats are just made up regardless of what happens during the match.

Recently an attacking midfielder of mine made his debut for my team, and he had lots of key passes that I have actually seen during the match. Once the match was over, I went to my squad to check the key pass/90 stat that I always give a lot of importance to, and it was 0. How is that possible? Is it possible that I have a wrong idea about the key pass stat in FM? Isn't it the same with Opta where it is the final pass before a shot?

Link to post
Share on other sites

7 hours ago, mhaffy said:

Re implementation of xG I wonder why they needed to re-define the stats at all. Couldn’t they have kept FME and QME as is and simply apply appropriate xG  to appropriate events....

Yeah they could have of course. Id guess that theoretically it makes more sense for some of the stats to be based round xG. Of course theory and real life can often be different, and once they saw the reality (for this years implementation at least) during development they should have gone back to the old way of counting stats.

Edited by francis#17
Link to post
Share on other sites

16 hours ago, themadsheep2001 said:

Guys if you're not actively uploading data and examples here, and a fair few of you are which is super helpful, please do not clog up the thread. The bugs forum is explicitly for uploading and adding examples

the problem is that sometimes there is no activity from SI is there is no activity in a topic...

Link to post
Share on other sites

6 hours ago, 4football said:

the problem is that sometimes there is no activity from SI is there is no activity in a topic...

It's marked as a known issue and Jack Joyce responded on Monday. They cannot spend all their time here because that takes time away from working on development/issues. These are their working forums, so they will check them even if they don't always have time to respond. Which is why it's important not to spam them with non relevant information. Thanks

Link to post
Share on other sites

19 hours ago, Mikal said:

Below are two videos from my save showing the individual player detailed stats (couldn't be bothered for individual screenshots):

1. Half a season on no detail (QME)

2. Half a season on full detail (FME)

Belgian Pro League 23/24: 1 - 18 games no detail, qme

 

Belgian Pro League - 23/24: 19-36 games full detail, fme

The stats I mentioned potentially being an issue in my previous post also seem to be highlighted here as well, e.g.:

- Players getting 1 or 2 clear cut chances attributed to them in the second half of the season compared to 35+ in the first half

- Players attempting 200% more passes in the second half of the season

- Not a single key tackle happened in the second half of the season

- Clearances through the roof second half of season

- Key passes up up up on the fme.

Cheers Mikal, flagging for @Jack Joyce

Link to post
Share on other sites

On 25/01/2021 at 18:31, themadsheep2001 said:

Just been chatting to @Davidog and actually worth sticking what I said to him in here:  "I believe QA have another internal match engine update coming up shortly, and they are going to review the stats again in that. Can't give you a timeframe i'm afraid because not even they would know that yet, In the mean time, they are keeping an eye on any new information coming into the thread, so please keep it coming. They are very stretched and aware they are not able to respond as much as they'd like"

Just to follow up, if you are seeing any examples that are NOT already covered by anything people have raised above, please get them in, they are still actively looking at this

Link to post
Share on other sites

In this screenshot it is showing the number of clearances per team after 12 games in a FME league.

There are 90 minutes per game so after 12 games thats 1080 minutes. That means Concord Rangers, Eastbourne Borough and Tonbridge Angels (they have played 11 games) all have at least 1 clearance in every minute of every game with the other teams not far off at all. This is not including the opposition teams clearances too so really it means there are at least 2 clearances in every minute of every march. When you  watch matches you can see that this isn't true and matches would be unwatchable if this was actually happening so there must be an issue with the counting of clearances.

image.png

Link to post
Share on other sites

Apologies if this has already been raised but the number of Crosses Attempted seems incredibly high in FME versus QME.

Just checked my paused save, after 22 matches in Premier League.

Salah tops Crosses attempted with 350.......The 20th highest is Allan Saint Maximim with 172

Looking at the stats in a few leagues in my save that use QME for crosses Attempted:

Spain 215 Highest, 67 for 20th highest

Italy  174 Highest, 77 for 20th highest

Sky Bet Championship 124 Highest, 85 20th highest.

Cross attempts in FME seem way...way too high.

Until sensible stats in FME and FME/QME more closely aligned it is a total waste of time using "stats" to scout players.........

 

Edited by mhaffy
Link to post
Share on other sites

5 hours ago, kingking said:

Does incorrect stats make a players average rating wrong?

because now i can only analyse players using average rating as stats are incorrect due to this bug

Which, in my opinion, is wrong. Tackles are tackles, key passes are key passes, average rating is...nothing.

Edited by yanloup64
Link to post
Share on other sites

On 01/02/2021 at 18:59, mhaffy said:

Apologies if this has already been raised but the number of Crosses Attempted seems incredibly high in FME versus QME.

Just checked my paused save, after 22 matches in Premier League.

Salah tops Crosses attempted with 350.......The 20th highest is Allan Saint Maximim with 172

Looking at the stats in a few leagues in my save that use QME for crosses Attempted:

Spain 215 Highest, 67 for 20th highest

Italy  174 Highest, 77 for 20th highest

Sky Bet Championship 124 Highest, 85 20th highest.

Cross attempts in FME seem way...way too high.

Until sensible stats in FME and FME/QME more closely aligned it is a total waste of time using "stats" to scout players.........

 

Yep, crosses are one of many stats that differentiate massively between the FME and QME. I think this is one of the reasons clearances are also incredibly high in the FME.

Reminds me of that game Man Utd had under Moyes against Fulham. :lol:

Stats are useless unless every league is running the FME and nobody has time for that...

My main worry is that the source of all these odd stats is the actual match engine. Which I'm sure has implications on player ratings e.g. Practically every corner going down as a key pass which massively inflates your set piece takers rating. 

Or your defenders being incapable of performing a key tackle which is the equivalent of a striker being unable to score a goal.

Link to post
Share on other sites

  • 2 weeks later...
1 hour ago, mhaffy said:

I understand that FM definitions will not align exactly with their real world equivalents but they should be at least broadly in line for both the FME and QME. It is important that FME and QME give similar results so that scouting/search comparisons are not rendered useless. I ran a full season Premier League using Full Match Engine and again using the Quick Match Engine. I then extracted full season stats for each team and compared the FME results with the QME and 19/20 Premier League "equivalent" stats:

image.jpeg.6cbf2a254d27f065f43a28e2299da2b7.jpeg

A few summary observations from the same FME Vs QME saves used above regarding player stats:

image.jpeg.a8c66b67f1af973d813fef613c1a19b5.jpeg

The above stats were only readily available for the top 20 players and not available within detailed team stats.


With the exception of Distance/90 mins, I believe that all the other "red font" items have been well flagged in the bugs forum for SI, some in this thread, and are "Known Issues" or “Under Review”. 


The game is not broken and unplayable but for players who like the immersion and use stats when comparing/scouting players etc there are clearly issues that have surfaced in this years FM that render such comparisons/scouting useless. Hopefully these issues will be largely resolved in the next major update so that SI have a firm foundation from which to develop FM22.

LEGEND! 

Link to post
Share on other sites

On 15/02/2021 at 09:29, mhaffy said:

The game is not broken and unplayable but for players who like the immersion and use stats when comparing/scouting players etc there are clearly issues that have surfaced in this years FM that render such comparisons/scouting useless. Hopefully these issues will be largely resolved in the next major update so that SI have a firm foundation from which to develop FM22.

Here here! Stellar work. 

Link to post
Share on other sites

On 15/02/2021 at 17:29, mhaffy said:

I understand that FM definitions will not align exactly with their real world equivalents but they should be at least broadly in line for both the FME and QME. It is important that FME and QME give similar results so that scouting/search comparisons are not rendered useless. I ran a full season Premier League using Full Match Engine and again using the Quick Match Engine. I then extracted full season stats for each team and compared the FME results with the QME and 19/20 Premier League "equivalent" stats:

image.jpeg.6cbf2a254d27f065f43a28e2299da2b7.jpeg

A few summary observations from the same FME Vs QME saves used above regarding player stats:

image.jpeg.a8c66b67f1af973d813fef613c1a19b5.jpeg

The above stats were only readily available for the top 20 players and not available within detailed team stats.


With the exception of Distance/90 mins, I believe that all the other "red font" items have been well flagged in the bugs forum for SI, some in this thread, and are "Known Issues" or “Under Review”. 

PKM files for my FME and QME saves on which the above data is based are available if required but I am sure SI can replicate the above readily.

The game is not broken and unplayable but for players who like the immersion and use stats when comparing/scouting players etc there are clearly issues that have surfaced in this years FM that render such comparisons/scouting useless. Hopefully these issues will be largely resolved in the next major update so that SI have a firm foundation from which to develop FM22.

Funny how SI never respond to posts like this. They love to ask us to post bug reports, and then when we do, IGNORED. what a useless company.

Link to post
Share on other sites

1 hour ago, ARustyFirePlace said:

Funny how SI never respond to posts like this. They love to ask us to post bug reports, and then when we do, IGNORED. what a useless company.

Because they haven't responded doesn't mean it's ignored they said on a post they read everything just haven't got the time to respond all the time

Link to post
Share on other sites

16 hours ago, ARustyFirePlace said:

Funny how SI never respond to posts like this. They love to ask us to post bug reports, and then when we do, IGNORED. what a useless company.

Harsh.

 

From the games I play there is not many with the same level of engagement we get from SI.

Link to post
Share on other sites

On 25/02/2021 at 14:32, namlas said:

After the update the issue has become even worse. Dribbling numbers have gone even lower(max is 1.5) However this time its uniform across all leagues including the one I'm managing in . 

Im seeing dribbling numbers higher than that

Link to post
Share on other sites

Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...