Football Manager 2021 - Out November 24th

[Red Horizon]

8 hours ago, s1111 said:

Someone correct me if I'm wrong but I believe every highlight of anything in FM starts from a dead ball as it's a natural place to start it from?

And there are 40-50 throw ins per game in real life, more than every other dead ball situation put together?

Making this possibly the most stupid complaint ever made in this forum, which is saying something.

It's more the fact that all the goals came from a set piece seconds before, rarely ever do you see a passage of play before a goal. 

[RandomGuy.]

9 hours ago, opgen22 said:

Although I love the addition of xG in the game, am I the only one thinking those values are too low? Every team seems to be absolutely clinical with their chances. I know that they created an own model so it is hard to compare with other ones, but they all represent the same idea. In the stream I haven't seen any team being wasteful with their chances, in the last game Milwall scored 3 with an xG of 0.55.  

Also, the value sometimes increased noticeably and there was no highlight to be seen. 

xG isn't a case of "i scored so it should be 1".

Depends on the system used (I don't know how SciSports works theirs) a chance of 0.25 or higher is one you should be expecting to see at least troubling the keeper. Penalties for example are worth ~0.76 usually.

I suspect if someone is managing Chelsea in the above, they'll be on this forum creating multiple topics about how flawed everything is.

Edited November 6, 2020 by RandomGuy.

[opgen22]

vor 6 Minuten schrieb RandomGuy.:

xG isn't a case of "i scored so it should be 1".

Depends on the system used (I don't know how SciSports works theirs) a chance of 0.25 or higher is one you should be expecting to see at least troubling the keeper. Penalties for example are worth ~0.76 usually.

I suspect if someone is managing Chelsea in the above, they'll be on this forum creating multiple topics about how flawed everything is.

I'm totally familiar with the concept. I was only wondering that at least in that small sample size there was no game in which a team had a higher xG value than goals actually scored (except for zero goals.) I was expecting xG values like this https://understat.com/match/14490 or this https://understat.com/match/14110.  I'm curious to see how the model works out in the lower league. Overall I'm very pleased, though. 

[Svenc]

38 minutes ago, RandomGuy. said:

xG isn't a case of "i scored so it should be 1".

Depends on the system used (I don't know how SciSports works theirs) a chance of 0.25 or higher is one you should be expecting to see at least troubling the keeper. Penalties for example are worth ~0.76 usually.

I suspect if someone is managing Chelsea in the above, they'll be on this forum creating multiple topics about how flawed everything is.

Isolated football matches are quite random. The biggest use of xG is actually over the longer term (end of video, article linked below). Does the data analys do some of that crunching?

https://www.youtube.com/watch?v=w7zPZsLGK18

https://www.espn.com/soccer/blog/name/67/post/2271195/headline (LFC's own analysis team came to similar conclusions, btw. which is one of the reasons they signed Klopp: https://www.nytimes.com/2019/05/22/magazine/soccer-data-liverpool.html)

11 hours ago, Gegenklaus said:

Imagine presenting a new game for the first time and being so clueless about just setting up a simple tactic. 4-2-4 tiki taka with no holding midfielder and wingbacks bombing forward on attacking mentality. Jesus.

They may have just wanted to provide an "authentic FM experience", because for a significant portion of their user base that may be what their tactics look like.  

Edited November 6, 2020 by Svenc

[Firehouse]

Early build, lol. Game is out in two weeks time. Thats not an early build, early build is what they have in April or something. Not saying they can't fix things in these two weeks.

I watched the stream until that goal against Cardiff. It was a nice finish 1v1 but 10 out of 10 defenders would've cleared the ball with the sliding tackle when the guy who scored the goal got the ball around center.

It does feel like there is a lot more clicking, which is not good, but I actually seem to like the new conferences and team talks.

[2feet]

STREAM FEEDBACK

Guys can you be a bit more structured in what you’re showing on the stream next time?

Like have a tick list of things you’re going to take us through, and prepare what you’re going to show in each menu item? EG training was hardly touched on.

If you could include useful tactical hints in your demonstrations too that would be cool.

And why not start the show when you are for example three games away from a transfer window, and you’ve already started your scouts on an assignment a few months previously, so we get a bit of scouting and transfers as well?

It was much more helpful when Miles came in, I picked up some good tactical tips too - more like this please, instead of blind-folded wanderings!

[Red Horizon]

Why does every goal start with a set piece? 

[Johnny Ace]

31 minutes ago, 2feet said:

STREAM FEEDBACK

Guys can you be a bit more structured in what you’re showing on the stream next time?

Like have a tick list of things you’re going to take us through, and prepare what you’re going to show in each menu item? EG training was hardly touched on.

If you could include useful tactical hints in your demonstrations too that would be cool.

And why not start the show when you are for example three games away from a transfer window, and you’ve already started your scouts on an assignment a few months previously, so we get a bit of scouting and transfers as well?

It was much more helpful when Miles came in, I picked up some good tactical tips too - more like this please, instead of blind-folded wanderings!

I wouldn't want tactic tips from Tom & Dom ( I think that was there names), was interesting seeing Miles in action though, but we all play the game differently. He's pro Opposition Instructions when I get the general opinion of them is to not use them & let your setup deal with the opposition

They were showing us new stuff remember so maybe nothing's changed with the training & transfers  

 

[Zemahh]

4 minutes ago, Johnny Ace said:

He's pro Opposition Instructions when I get the general opinion of them is to not use them & let your setup deal with the opposition

Indeed, that was interesting to hear. He repeated multiple times that Opposition Instructions should be done, which makes me wonder if they're more beneficial than we're led to believe. I've seen Rashidi and Knap both recommend you leave them empty, since they can obviously disrupt your shape quite a bit at times, but maybe Miles knows something we mere mortals don't. 

[Johnny Ace]

Just now, Zemahh said:

Indeed, that was interesting to hear. He repeated multiple times that Opposition Instructions should be done, which makes me wonder if they're more beneficial than we're led to believe. I've seen Rashidi and Knap both recommend you leave them empty, since they can obviously disrupt your shape quite a bit at times, but maybe Miles knows something we mere mortals don't. 

Maybe he's easy on the pressing, marking & tackling & stuff so uses OI's instead? He maybe just plays differently 

[Kingstontom88]

Was only able to catch bits and pieces: what did the ME look like? Improvement on last year or by and large the same?

[themadsheep2001]

8 minutes ago, Zemahh said:

Indeed, that was interesting to hear. He repeated multiple times that Opposition Instructions should be done, which makes me wonder if they're more beneficial than we're led to believe. I've seen Rashidi and Knap both recommend you leave them empty, since they can obviously disrupt your shape quite a bit at times, but maybe Miles knows something we mere mortals don't. 

The answer, and sorry you'll hate it is that it depends on what you're doing. Do it wrong and you can pull your team defensive shape wrong, do it right and you can really increase your potency. FWIW Rashidi's use is tactic dependent. 

I use a split press in my setup so I tend to use opposition instructions to press the entire back line.

[themadsheep2001]

10 hours ago, Mcfc1894 said:

Tactics clearly matter how we all seen when miles made a few changes 

I think not having some well made structured tactics ( in my opinion) from the start didn't quite help. Was flicking in abd and out with the election going on but it could have helped show off the ME more

[XaW]

1 hour ago, BusbyFergieJose said:

Why does every goal start with a set piece? 

You already posted that, and I posted a response. Why post it again without responding to your answers?

[steff91]

2 minutes ago, XaW said:

You already posted that, and I posted a response. Why post it again without responding to your answers?

i love that avatar may i borrow it

 

[XaW]

2 minutes ago, steff91 said:

i love that avatar may i borrow it

 

Be my guest! I've had it for about 20 years or so now whenever I can. Ever since I discovered the 8-bit theatre on NuklearPower (Just to send regards to the creators!).

[grade]

I saw the stream and I was quite impress with the game, but I mainly play FMT, so I'm curious of the new features are in it.

[steff91]

40 minutes ago, XaW said:

Be my guest! I've had it for about 20 years or so now whenever I can. Ever since I discovered the 8-bit theatre on NuklearPower (Just to send regards to the creators!).

thank you

[XaW]

6 minutes ago, steff91 said:

thank you

[Johnny Ace]

Oh, that's weird 

[JordanMilly]

57 minutes ago, steff91 said:

thank you

 

49 minutes ago, XaW said:

Sorry boys, but forum rules are that you have to fight now.

[XaW]

5 minutes ago, JordanMillward_1 said:

 

Sorry boys, but forum rules are that you have to fight now.

[Rashidi]

2 hours ago, themadsheep2001 said:

I think not having some well made structured tactics ( in my opinion) from the start didn't quite help. Was flicking in abd and out with the election going on but it could have helped show off the ME more

Agreed. I still think livestreams aren't exactly a great way to judge how the match engine plays. Ultimately each person has something specific they are looking for in the match engine. Quite unrealistic for people to make judgements on the engine, wait till i do my livestream :-)

[Defensive]

vor 6 Stunden schrieb BusbyFergieJose:

It's more the fact that all the goals came from a set piece seconds before, rarely ever do you see a passage of play before a goal. 

Not true. Only 8 of 22 goals in the stream came from set pieces if i counted correctly.

[Defensive]

vor 4 Stunden schrieb Firehouse:

I watched the stream until that goal against Cardiff. It was a nice finish 1v1 but 10 out of 10 defenders would've cleared the ball with the sliding tackle when the guy who scored the goal got the ball around center.

 

I see this kind of defender mistakes at least 20 times each weekend in Europes highest divisions.

 

If anything i found it strange that we saw only 1 of these mistakes leading to a goal in the whole stream.

Edited November 6, 2020 by Defensive

[Rashidi]

From what i can tell from yesterday's stream:

1. Recruitment meetings look pretty cool for people who want a more realistic experience and not want to use filters and such. How they execute this will be important. If it ties into transfer/budgets and club vision, it could be a great thing. Currently the DOF's recommendations sometimes ignores vision.

2. Icons - great idea, cos for too long people assume you need to play someone on 90%, but irl how do you know someone is 90%? Still though colourful icons could still be an issue for visually impaired people. 

3. Contracts - More options, definitely a win, but I want to see the AI be a lot more competitive with contracts, transfers

4. Save/load times - I noticed this on the stream yesterday, seemed lightning quick which also points towards match engine optimisation cos it would suggest more efficient code structure. Could suggest that people may be able to have larger saves. Will it run on older machines. I have 4 machines at home, 2 macs and 2 windows machines so it will be fun to find out.

5. Match Day experience, love the new look. No time bar. xG has been added to the game, which for me is a lifesaver cos now I don't need to visit match analysis. You still need to pay attention to spike changes in xG which would be indicative of a good shot

6. Streamers used a 433 wide on cautious, not exactly a great system in itself to reveal a lot about an engine, but match engine assessments are subjective. AMC was decent in FM20. People will need to get their hands on the game themselves to make their own assessments. Some may think its brilliant others will not. That's just normal.

 

 

[saiyaman]

I have to say after watching the Match Engine video, my stance of not pre-ordering has definitely changed to considering it. I like the changes.

However, am I the only one who thinks "xG" is just a glorified stat which is practically useless?

Scoring a goal depends on various factors: Mistake of the opposition, quality of the chance, the ability of the player, the weakness of the goalkeeper and a lot more factors which happen ONLY during the match. It's possibly the most useless and irrelevant stat I've seen in my life. I can guess all of this just by watching the match.

Edited November 6, 2020 by saiyaman

[jimbo22]

1 hour ago, Rashidi said:

1. Recruitment meetings look pretty cool for people who want a more realistic experience and not want to use filters and such. How they execute this will be important. If it ties into transfer/budgets and club vision, it could be a great thing. Currently the DOF's recommendations sometimes ignores vision.

 

 

One of my big frustrations in 20 (and before) is the head scout/scouts giving you reports and recommendations on players that are quite simply out of your reach. ...Scout recommends a player worth 20m, yeah mate, i've got 5 million in the bank, good job. 

 

Hopefully this works as we all hope it does. 

Edited November 6, 2020 by jimbo22

[Svenc]

27 minutes ago, saiyaman said:

However, am I the only one who thinks "xG" is just a glorified stat which is practically useless?

Scoring a goal depends on various factors: Mistake of the opposition, quality of the chance, the ability of the player, the weakness of the goalkeeper and a lot more factors which happen ONLY during the match. It's possibly the most useless and irrelevant stat I've seen in my life. I can guess all of this just by watching the match.

For some it will be less useful, for some more so. A lot of FMers from my experience are as "bad" at that guessing as these guys. 

https://statsbomb.com/2016/10/xcommentary/

The xg plot (good that they've included this) actually should be useful for anyone, in particular considering the dynamic changes made by AI managers. Like when they take an early lead and then focus on defending, or vice versa switch to something more attacking late. Spikes there can be indicative of a big switch to watch out for. 

Edited November 6, 2020 by Svenc

[Rashidi]

44 minutes ago, saiyaman said:

However, am I the only one who thinks "xG" is just a glorified stat which is practically useless?

IN FM 20 you don't really have a dynamic way of telling you whether you are generating quality chances. All you have are clear cut chances, which don't tell you much. xG allows us to assess all shots that are taken, so you have a dynamic reference that keeps changing indicating whether your chances are good quality chances. Most people reckon they know what a good chance is, but xG is a different thing altogether because it factors in several conditions that need to be met before a chance is recognised as a good chance. So a game kicks off you had one shot and its 0.01, three shots later its 0.10, then you take another shot and this time it goes up to 0.70. That final shot is a good chance. So you can always go back to see what the chance looks like.

And it takes into account all those things you mentioned when calculating what a good xG score is. Where the shot is taken from, defensive positioning of keeper, whether its with the head/not, deflected, etc

The same thing doesn't happen with current match stats. So yes I do think xG will be a helpful addition for those who understand it.

Try the quiz here to see how you do with assessing what a good chance is https://www.bbc.com/sport/football/40699431

 

[2feet]

So if your XG is high but you still haven’t scored then that means sub your striker?

[Rashidi]

11 minutes ago, 2feet said:

So if your XG is high but you still haven’t scored then that means sub your striker?

Yes that suggests you have a striker who is performing poorly.

[Mr U Rosler]

XG is very much flavour of the month. 

Not sure how useful it can be in a video game. 

In this implementation it's a performance metric extrapolated from a computer simulation that has historically had difficulty quantifying half and clear cut chances never mind having to measure shot angle, distance from goal, whether it was a header, weaker or stronger foot, type of attack, type of pass, were defenders in position etc, etc. 

This can be done in the real world, not sure how it could be pulled out of code in a meaningful way. 

Will probably end up winding loads of people up when their XG is 3 per game and they're only averaging 1.

Edited November 6, 2020 by Mr U Rosler

[Johnny Ace]

7 minutes ago, Mr U Rosler said:

Will probably end up winding loads of people up when their XG is 3 per game and they're only averaging 1.

Yeah, I can see XG being the reason for many complaints this year 

[HUNT3R]

28 minutes ago, 2feet said:

So if your XG is high but you still haven’t scored then that means sub your striker?

That's assuming he's the one fluffing chances though. Context, as always, will be important.

[crarfc]

I'm hoping Michail Antonio is a STRIKER in the game, and a huge improvement ! 

[2feet]

Is there a way to hover the mouse over the xg chart at half time / full time to see who it was who fluffed the various chances?

[Red Horizon]

1 hour ago, Rashidi said:

IN FM 20 you don't really have a dynamic way of telling you whether you are generating quality chances. All you have are clear cut chances, which don't tell you much. xG allows us to assess all shots that are taken, so you have a dynamic reference that keeps changing indicating whether your chances are good quality chances. Most people reckon they know what a good chance is, but xG is a different thing altogether because it factors in several conditions that need to be met before a chance is recognised as a good chance. So a game kicks off you had one shot and its 0.01, three shots later its 0.10, then you take another shot and this time it goes up to 0.70. That final shot is a good chance. So you can always go back to see what the chance looks like.

And it takes into account all those things you mentioned when calculating what a good xG score is. Where the shot is taken from, defensive positioning of keeper, whether its with the head/not, deflected, etc

The same thing doesn't happen with current match stats. So yes I do think xG will be a helpful addition for those who understand it.

Try the quiz here to see how you do with assessing what a good chance is https://www.bbc.com/sport/football/40699431

 

What are your thoughts on set pieces on the game? It seems like most goals come from them to me. 

 

I realise a lot of goals are scored from set pieces in real life but it's seems way too much on FM. 

[paddypower]

58 minutes ago, Mr U Rosler said:

XG is very much flavour of the month. 

 

It's been in mainstream football for 10 years. If anything it's outdated.

[ajw10]

Does anyone have an opinion on set piece goals? I've not seen any mention of it...

[Rashidi]

28 minutes ago, BusbyFergieJose said:

What are your thoughts on set pieces on the game? It seems like most goals come from them to me. 

 

I realise a lot of goals are scored from set pieces in real life but it's seems way too much on FM. 

Can't base much from one livestream. You need to play the long game, finish a whole season to assess it.

[DP]

21 minutes ago, Rashidi said:

Can't base much from one livestream. You need to play the long game, finish a whole season to assess it.

I don’t think this is true. If it wasn’t previously a problem, fine. But because we’ve seen it so much in previous iterations seeing it again in the first few games is a good indication it’s still the same. 
 

It’s not a deal breaker but it’s something that needs looking at IMO. 

Edited November 6, 2020 by DP

[Red Horizon]

13 minutes ago, Rashidi said:

Can't base much from one livestream. You need to play the long game, finish a whole season to assess it.

I agree, but it's a problem that has been on previous versions, we noticed on our network game that the marking on set pieces is terrible, on long throws and short throws, the amount of goals coming from them was ruining the save. 

And watching the stream yesterday didn't fill me with confidence that it was fixed (goals were coming from throws) 

[Zemahh]

1 hour ago, 2feet said:

So if your XG is high but you still haven’t scored then that means sub your striker?

If your xG is higher than the number of goals you scored, it means you created good quality chances, but failed to finish them. If your striker is consistently getting high xG without actually scoring, that could indeed indicate that it's him and not your tactics. 

If your xG is lower than the number of goals you scored, it means you created no good quality chances, but still managed to finish them through difficult shots. That could indicate your tactic isn't good enough, but you're getting bailed out by your players (i.e. once your luck runs out, you're toast).

However, keep in mind that it's best to judge these things over the course of a few games. Not every individual game will go exactly as the xG indicates it should, but if there's a pattern, then you might be onto something.

[themadsheep2001]

Xg is more than 10 years old, and clubs use it and far more sophisticated models of the concept on an everyday use

[-Jef-]

What clubs use now will be available to the public in about 15 years. The amount of data they have is astonishing.

Edited November 6, 2020 by rosque

[RTHerringbone]

A bunch of comments earlier were bemoaning the fact that loads of highlights start from set pieces. There's a reason which @Jack Joyce explained in another thread:

 

[samdiatmh]

12 hours ago, Johnny Ace said:

Yeah, I can see XG being the reason for many complaints this year 

I'm literally just waiting for the rages when the xG is something ridiculous like 3.4-0.3, and they still lose 1-0

Particularly on something like a twitch stream, that post-match is going to be amazing - particularly if the press conferences dive into those stats too

 

Personally, I'd like to see a *slight* tweak to the xG, where it actually displays goals when they're made as an overlay (yeah, it might end 4-1, but if we're only scoring from shots where we can't miss, I want to see that)

[Svenc]

4 hours ago, samdiatmh said:

I'm literally just waiting for the rages when the xG is something ridiculous like 3.4-0.3, and they still lose 1-0

Particularly on something like a twitch stream, that post-match is going to be amazing - particularly if the press conferences dive into those stats too

I think this is going to happen less than with the ill-fated CCC, no less as not all chances are going to be treated equal anymore. "I just had 4 CCCs and didn't score!" is something else than having four "Big chances" with an xG of ~0.25 each and not scoring. There may be discussions too how a certain chance would have such a "low" xG though.

Still, RNG is going to be RNG. And football, in-game or otherwise, can still be quite random. For the recently released Baldur's Gate 3 Early Access there is rage all over the shop how people could miss with a 95% chance to hit four times in a row... . The devs are actually considering to "stabilize" their internal dice, which as a (D&D) purist, I don't like at all. Nothing cheaper than knowing that the dice are rigged in your favor and you only won because of that, too.  


 

13 hours ago, themadsheep2001 said:

Xg is more than 10 years old, and clubs use it and far more sophisticated models of the concept on an everyday use

Yup: https://footballbh.net/2019/05/28/brentford-championship-tactical-analysis-statistics-analysis/

In the end, this is really about having models that accurately measure the quality of chances -- for and against. How well are we actually playing? Do we deserve the results we have? And as such ruling out that what you see on the point tables isn't just a "fluke". Teams who consistently have better chances have better chances to win football matches. It's not only clubs though, but everybody who earns their living with assessing football teams. Bettors, bookmakers, et all. They all use such stuff because the league table in such a low scoring sports in particular isn't very reliable -- see also the Newcastle-under-Pardew example from the article above under the header "The league table always lies", basically the opposite of the Dortmund-under-Klopp scenario in his rotten last season there. Klopp resigned (and Dortmund remained a top club in the Bundesliga), Newcastle meanwhile gave Pardew a lifetime contract and came down crashing hard.   
 

Edited November 7, 2020 by Svenc

[Johnny Ace]

13 hours ago, RTHerringbone said:

A bunch of comments earlier were bemoaning the fact that loads of highlights start from set pieces. There's a reason which @Jack Joyce explained in another thread:

Makes perfect sense, I think I commented at the time it looked like the set piece was being used as a starting point for the highlight, some of the highlights were a lot longer than usual