You are not logged in or registered. Please login or register to use the full functionality of BayernForum.com

Robert Lewandowski

Discussions about the former / present players & staff, and their personal life. Was a player fined for speeding? Did a former player become a coach? Find it all here.
 

Re: Robert Lewandowski

Postby #12 » Thu Oct 24, 2019 6:50 pm

So you’re rather relying on around the worlds and backpasses?

And here I was thinking the point of the game was to score goals...
Summer 2020: Hasan, show us your weakness!!
User avatar
#12
Chief Critic
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 21007
Joined: Sat May 16, 2015 1:31 pm
National Flag:
Germany
Has thanked: 1083 times
Been thanked: 4783 times
Gender: Male

Re: Robert Lewandowski

Postby MUTU » Thu Oct 24, 2019 8:39 pm

Borusse wrote:You're arguing against math at this point. Expected pretty much means the chance of an event happening and how many times you can expect it to happen when it's played out multiple times. Like a toin coss, expected heads or tails are 0.5 but there can easily be a few heads or tails in a row before it evens out long term.

There is a 1/6 EX on getting a specific number when throwing a dice. It's very important to know all of this before you start gambling.

xG in football gives a much better understanding of how real the threat the team created in a given game. Looking just at the number of shots is too vague and can lead to wrong conclusions.

Let's say you take 20 shots during a game against opponent's 5 shots. Looks like you dominated the game and created a huge number of chances, right?
But what if those 20 shots were taken from 40meters out and had no real chance of being effective? The stat becomes very misleading. This is where xG comes in handy because it estimates the probability of the shot going in and it would probably give all of those shots something like 2% chance of being effective. So the team that took 20 shots had actually only 0.4 xG. The other team took 5 shots but all of them were 1on1s with the keeper with a 60% chance each. That gives us 3.0 xG. Suddenly it makes you take a completely different look at the game. This is of course an extreme example for the discussion's purposes.

We can argue about the model and how it could be more accurate in some cases and that it's not optimal yet, sure. But saying it's "shit" is like denying that probability exists. And sorry to break it to you but probability is always going to be a factor no matter what kind of competition you take part in.

There are three types of people in this world.

1) those who understand science and maths and trust it
2) those who don't understand science and maths but are wise enough to realise it's not bullshit
3) those who don't understand science and maths and are unaware of their ignorance in the subject and think they know better (and they never do).

I try my best to avoid people who fall in the 3rd category but end up struggling to hold myself from answering them back. That's why I still argue with zozon.
30GB free cloud storage. Click here for the referral.
User avatar
MUTU
Site Admin & EURO 2016, 2017/18 and World Cup 2018 Prediction Game Winner
Site Admin & EURO 2016, 2017/18 and World Cup 2018 Prediction Game Winner
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 43981
Joined: Mon Aug 14, 2006 11:00 pm
Location: L-Imqabba, Malta
National Flag:
Malta
Has thanked: 8046 times
Been thanked: 12502 times
Gender: Male
BayernForum.com fan club: Active member
BayernForum.com donator: Yes

Re: Robert Lewandowski

Postby Ziklitschli » Thu Oct 24, 2019 10:02 pm



And finally, the goddamn choker is exposed for good! I particularly love the methodology, ignoring the flat track bully's CL goals in the Dortmund shirt, those were only used to cover the truth anyway.

Count them in though, who cares, he still has only 19 KO goals, a mere third of his total of 57. Just compare that to the god of football Messi, a 40 of 113, such a true gulf between the small big man and a big little Lewa man, easily tripling on the pathetic attempts to take into account details like Messi always having at least 300 million worth of attacking players running around him.

Choke on, halberd man.
It's time to deut the Raum
Ziklitschli
Yeah, now I'm talking!
 
Years of membershipYears of membershipYears of membershipYears of membership
 
Posts: 148
Joined: Wed May 11, 2016 10:56 am
Has thanked: 20 times
Been thanked: 80 times
Gender: Male

Re: Robert Lewandowski

Postby Dumbledore7 » Thu Oct 24, 2019 10:17 pm

#12 wrote:So you’re rather relying on around the worlds and backpasses?

And here I was thinking the point of the game was to score goals...

I rely on watching the matches because I have yet to figure out a way to test the scientific method for player abilities in football, so I’d rather not consume half-way numbers that can’t possibly be correctly interpreted without a caveat, which makes them meaningless numbers. And I *am* a rocket scientist.

MUTU wrote:There are three types of people in this world.

1) those who understand science and maths and trust it
2) those who don't understand science and maths but are wise enough to realise it's not bullshit
3) those who don't understand science and maths and are unaware of their ignorance in the subject and think they know better (and they never do).

I try my best to avoid people who fall in the 3rd category but end up struggling to hold myself from answering them back. That's why I still argue with zozon.


You and zozon aren’t far off. It’s one thing to completely distrust any metric, but it’s just as bad to take everything at face value. Trump/Brexit mongering, hello?

These kinds of things are written by those who have limited understanding of science or maths, but are arrogant enough to think that using numbers implies an understanding of either. Science is built around the scientific method, and maths is a study of conjectures that merely use numbers as a tool. Accounting use arithmetics, but isn’t maths, and I’ve seen more nuanced figures in accounting books than G+A/90. Statistical modelling, on the other hand, fits perfectly in this realm as it is a way to test hypotheses against real results - xG is a good footballing example, and even that should only be used to make any assessment within the boundaries it was modelled for (i.e. assessing a singular game).

People who actually understand science and maths would always be the first to question any numbers and not trust it for granted. For any form of comparison to be even discussable, you have to first do a large quantity of tests, with ideally perfectly controlled variables other than the one you’re testing. Football is a low-scoring game, and no two minutes of football are the same, yet you use a /90 metric because it looks “smart” to standardise the comparison between two players without recognising whether the two players had the exact same volume (for lack of better word - and this very lack of a better word is also a problem) of opportunities to score goals within those arbitrary 90 minutes.

It’s sometimes wise to recognise that there are things we can’t fully explain.

You seemed perfectly happy to use WhoScored ratings to compare Khedira and Can, yet I’ve strangely missed any reference to that for Coutinho and Müller - last time I checked a few weeks ago, Coutinho’s was higher. What was that about understanding and trusting science and maths?
User avatar
Dumbledore7
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 5777
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 4203 times
Been thanked: 3478 times
Gender: Male
BayernForum.com fan club: Active member

Re: Robert Lewandowski

Postby MUTU » Thu Oct 24, 2019 10:32 pm

It was already explained multiple times on this forum (myself included) why WhoScored makes little sense to compare between two players whose average time on the pitch is not comparable.

As for xG+xA/90 vs G+A/90 both have benefits and are useful together. You can't say one is good and the other useless. Imagine a hypothetical player who scores literally every shot he takes, but he's marked so well he only scores a hat trick each match, from like the halfway line sometimes. His xG would be poor but his G/90 is going to be high.

On the other hand if someone has terrible finishing he might be positioning himself really well and get passes from teammates and have great xG, but if he shoots them all wide what's the point.

xG+xA is a measurement of the positioning of a player.
G+A is a measurement of the produce of the player
The comparison of one against the other is an indication of the efficiency of a player.

Also, need to be condescending. I'm also a scientist myself.
30GB free cloud storage. Click here for the referral.
User avatar
MUTU
Site Admin & EURO 2016, 2017/18 and World Cup 2018 Prediction Game Winner
Site Admin & EURO 2016, 2017/18 and World Cup 2018 Prediction Game Winner
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 43981
Joined: Mon Aug 14, 2006 11:00 pm
Location: L-Imqabba, Malta
National Flag:
Malta
Has thanked: 8046 times
Been thanked: 12502 times
Gender: Male
BayernForum.com fan club: Active member
BayernForum.com donator: Yes

Re: Robert Lewandowski

Postby DRvad14 » Thu Oct 24, 2019 10:40 pm

^ so it whats the point for g/90 now ? " or its all bs can be made to fit ? swear thought it my golden weekend
and no it aint for 1 ,for the 2 i agree as in what player could /may produce
3? so u saying its useless to for efficiency
hats off to u for being one(ig or as u claim)
Last edited by DRvad14 on Fri Oct 25, 2019 12:22 am, edited 1 time in total.
User avatar
DRvad14
I post all the time
 
Years of membershipYears of membershipYears of membershipYears of membership
 
Posts: 2395
Joined: Sat Oct 24, 2015 6:40 pm
Location: NY
National Flag:
India
Has thanked: 1772 times
Been thanked: 1376 times
Gender: Male
BayernForum.com fan club: Active member
BayernForum.com donator: Yes

Re: Robert Lewandowski

Postby Dumbledore7 » Thu Oct 24, 2019 11:08 pm

MUTU wrote:It was already explained multiple times on this forum (myself included) why WhoScored makes little sense to compare between two players whose average time on the pitch is not comparable.

So, a caveat. That’s one more metric to add to the meaningless list, good. Let’s not see WhoScored on this forum ever again.

MUTU wrote: As for xG+xA/90 vs G+A/90 both have benefits and are useful together. You can't say one is good and the other useless. Imagine a hypothetical player who scores literally every shot he takes, but he's marked so well he only scores a hat trick each match, from like the halfway line sometimes. His xG would be poor but his G/90 is going to be high.

If you put together all the expressions I’ve made in posts around this topic (difficult though I know), you’ll find that I respect xG but would also think xG/90 is as useless as G/90 individually, because doing the /90 broke the boundaries of the intention in which xG is modelled. xG is generated per match, so it is only as useful as its description, i.e. expected goals in the match that was just played. So it can be be used to assess what level of chance the team generated in that match, and should be used as such, e.g. as a measurement for a coach to study his success/mistakes from that one match.

Also, what you really meant by being “useful together” came down mathematically to cancelling out the arbitrary, and highly misleading, /90. So we’re heading in a good direction.
MUTU wrote: On the other hand if someone has terrible finishing he might be positioning himself really well and get passes from teammates and have great xG, but if he shoots them all wide what's the point.

Good, this is the first time you talk like someone who understands the nuances of numbers.
MUTU wrote: xG+xA is a measurement of the positioning of a player.
G+A is a measurement of the produce of the player

The comparison of one against the other is an indication of the efficiency of a player.

Now this is your best one so far. Throughout this post you’ve managed to logically conclude that the /90 stats are completely misleading without more controlled processing, and you’ve come to a far more interesting and intelligent set of posits - that maybe it’s the xG and G that are more meaningful because what they imply are far simpler to define, and your starting points are clearly baselined. That being said, a combination of these two should still only be seriously talked about when it comes to one specific match and that match only. /90 is highly complex because you would have to weight and normalise every minute, if not second, against the likelihood of scoring a goal amongst many things. It makes me wonder why you bothered with G+A/90 in the first place, other than visual effects.
MUTU wrote: Also, need to be condescending. I'm also a scientist myself.

You weren’t showing it, and you were being pitifully arrogant AND condescending. A scientist isn’t somebody who knows the most things, they’re somebody who doesn’t know the most things.
Last edited by Dumbledore7 on Thu Oct 24, 2019 11:36 pm, edited 1 time in total.
User avatar
Dumbledore7
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 5777
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 4203 times
Been thanked: 3478 times
Gender: Male
BayernForum.com fan club: Active member

Re: Robert Lewandowski

Postby #12 » Thu Oct 24, 2019 11:13 pm

Rocket science or not, even if you DO claim they do not "get" the same chances (being a striker myself I strongly believe in players working on and FOR their own chances as well...) - doesn’t negate the assists...

Even in overall stats Müller's numbers are better than hype boy Coutinho... And I think even Müller's are bad...
Summer 2020: Hasan, show us your weakness!!
User avatar
#12
Chief Critic
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 21007
Joined: Sat May 16, 2015 1:31 pm
National Flag:
Germany
Has thanked: 1083 times
Been thanked: 4783 times
Gender: Male

Re: Robert Lewandowski

Postby DRvad14 » Thu Oct 24, 2019 11:20 pm

#12 wrote:Rocket science or not, even if you DO claim they do not "get" the same chances (being a striker myself I strongly believe in players working on and FOR their own chances as well...) - doesn’t negate the assists...

Even in overall stats Müller's numbers are better than hype boy Coutinho... And I think even Müller's are bad...


so the reality is different from probability yea ?
User avatar
DRvad14
I post all the time
 
Years of membershipYears of membershipYears of membershipYears of membership
 
Posts: 2395
Joined: Sat Oct 24, 2015 6:40 pm
Location: NY
National Flag:
India
Has thanked: 1772 times
Been thanked: 1376 times
Gender: Male
BayernForum.com fan club: Active member
BayernForum.com donator: Yes

Re: Robert Lewandowski

Postby Dumbledore7 » Thu Oct 24, 2019 11:30 pm

#12 wrote:Rocket science or not, even if you DO claim they do not "get" the same chances (being a striker myself I strongly believe in players working on and FOR their own chances as well...) - doesn’t negate the assists...

Even in overall stats Müller's numbers are better than hype boy Coutinho... And I think even Müller's are bad...

Assists are equally arbitrary. How much of an assist is the assister’s great positioning, and how much of it is the brilliance of the player that made the pass before that?

To skip the potential long-winded, despite interesting discussion, you’ll find that if you think about it, the shot of the player whose name will go on the goal is literally not any more important than a seemingly random pass made by the left back in the sequence leading to the goal. A goal is a team stat, not a player stat.

Goalimpact does attempt to normalise the above, which is why I respect it, and Müller does have that on his side - I would be far less anal about this if you argue for Müller using Goalimpact because I see it as an acceptable supporting metric. It’s still flawed (e.g. as simple as the fact that it’s centred around goal difference, which may or may not be the most important thing), but to its credit it has taken into account far more nuances than kindergarten maths. It just forms one of the many points when you assess the ability of a player though.
User avatar
Dumbledore7
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 5777
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 4203 times
Been thanked: 3478 times
Gender: Male
BayernForum.com fan club: Active member

Re: Robert Lewandowski

Postby zozon » Fri Oct 25, 2019 12:39 am

MUTU wrote:1) those who understand science and maths and trust it
2) those who don't understand science and maths but are wise enough to realise it's not bullshit
3) those who don't understand science and maths and are unaware of their ignorance in the subject and think they know better (and they never do).

I try my best to avoid people who fall in the 3rd category but end up struggling to hold myself from answering them back. That's why I still argue with zozon.


Yeah, except that I'm an engineer, so tell me more about science and maths, please. Had like 10+ math-based subjects in my 4-year bachelor degree, so guess that makes me a... journalist or a gardener?

Good on you for judging someone's knowledge of science and maths just because they think that xG is a shitty tool to measure anything. A tool based on some formula. A formula that can be modified by whoelse wants to change it, it will give different results, but in the end - its some statistical probability. At the end of the day its a "woulda - shoulda" tool.

And you still havent answered how that Coman blocked shot (after 1m (or less) of mere travel - the ball was blocked by a defender) had a higher xG than Gnabry's open shot to the post. I know you argued it was 0.01 higher, but now when you are clinging to your educational titles, you should know its STILL higher. And I couldnt care less if it was 0.00001.
User avatar
zozon
I post all the time
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 2407
Joined: Wed May 08, 2013 10:04 am
Has thanked: 326 times
Been thanked: 1084 times
Gender: Male

Re: Robert Lewandowski

Postby #12 » Fri Oct 25, 2019 7:19 am

Dumbledore7 wrote:
#12 wrote:Rocket science or not, even if you DO claim they do not "get" the same chances (being a striker myself I strongly believe in players working on and FOR their own chances as well...) - doesn’t negate the assists...

Even in overall stats Müller's numbers are better than hype boy Coutinho... And I think even Müller's are bad...

Assists are equally arbitrary. How much of an assist is the assister’s great positioning, and how much of it is the brilliance of the player that made the pass before that?

To skip the potential long-winded, despite interesting discussion, you’ll find that if you think about it, the shot of the player whose name will go on the goal is literally not any more important than a seemingly random pass made by the left back in the sequence leading to the goal. A goal is a team stat, not a player stat.

Goalimpact does attempt to normalise the above, which is why I respect it, and Müller does have that on his side - I would be far less anal about this if you argue for Müller using Goalimpact because I see it as an acceptable supporting metric. It’s still flawed (e.g. as simple as the fact that it’s centred around goal difference, which may or may not be the most important thing), but to its credit it has taken into account far more nuances than kindergarten maths. It just forms one of the many points when you assess the ability of a player though.


Of course, but still it would be an indicator fir me that the respective player is a better fit in said team than another one whose numbers are worse...
And we both know that this game is about more than -just- math...
Summer 2020: Hasan, show us your weakness!!
User avatar
#12
Chief Critic
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 21007
Joined: Sat May 16, 2015 1:31 pm
National Flag:
Germany
Has thanked: 1083 times
Been thanked: 4783 times
Gender: Male

Re: Robert Lewandowski

Postby #12 » Fri Oct 25, 2019 7:22 am

zozon wrote:
MUTU wrote:1) those who understand science and maths and trust it
2) those who don't understand science and maths but are wise enough to realise it's not bullshit
3) those who don't understand science and maths and are unaware of their ignorance in the subject and think they know better (and they never do).

I try my best to avoid people who fall in the 3rd category but end up struggling to hold myself from answering them back. That's why I still argue with zozon.


Yeah, except that I'm an engineer, so tell me more about science and maths, please. Had like 10+ math-based subjects in my 4-year bachelor degree, so guess that makes me a... journalist or a gardener?

Good on you for judging someone's knowledge of science and maths just because they think that xG is a shitty tool to measure anything. A tool based on some formula. A formula that can be modified by whoelse wants to change it, it will give different results, but in the end - its some statistical probability. At the end of the day its a "woulda - shoulda" tool.

And you still havent answered how that Coman blocked shot (after 1m (or less) of mere travel - the ball was blocked by a defender) had a higher xG than Gnabry's open shot to the post. I know you argued it was 0.01 higher, but now when you are clinging to your educational titles, you should know its STILL higher. And I couldnt care less if it was 0.00001.


1.) they gave YOU a degree? Did you insult everyone in the oral so they just wanted it over with?

2.) pretty easy, because it had a higher chance of ending up in goal - if someone sith a regular finishing took it that is
Summer 2020: Hasan, show us your weakness!!
User avatar
#12
Chief Critic
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 21007
Joined: Sat May 16, 2015 1:31 pm
National Flag:
Germany
Has thanked: 1083 times
Been thanked: 4783 times
Gender: Male

Re: Robert Lewandowski

Postby Dumbledore7 » Fri Oct 25, 2019 10:13 am

#12 wrote:Of course, but still it would be an indicator fir me that the respective player is a better fit in said team than another one whose numbers are worse...
And we both know that this game is about more than -just- math...

Benzema (Real), Firmino (Liverpool), Mandzukic (Bayern 12/13), Giroud (France) are examples of strikers who are without a doubt perfect fits for their respective teams, leading to colossal success without having that many goals to show for it.
User avatar
Dumbledore7
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 5777
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 4203 times
Been thanked: 3478 times
Gender: Male
BayernForum.com fan club: Active member

Re: Robert Lewandowski

Postby #12 » Fri Oct 25, 2019 12:01 pm

ONE indicator...

I agree it shouldn’t be the be all and end all but dismissing it as garbage is way too mich for me...
Summer 2020: Hasan, show us your weakness!!
User avatar
#12
Chief Critic
 
Years of membershipYears of membershipYears of membershipYears of membershipYears of membership
 
Posts: 21007
Joined: Sat May 16, 2015 1:31 pm
National Flag:
Germany
Has thanked: 1083 times
Been thanked: 4783 times
Gender: Male

PreviousNext

Return to Bayern Player & Staff Discussion

Who is online

Users browsing this forum: No registered users and 2 guests