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

The Statistics Thread

Discussions about anything in general about Bayern, such as tactics, finances, kits, merchandise etc.
 

Re: The Statistics Thread

Postby ottackon » Mon Mar 20, 2017 4:24 am

Thiago and Vidal are one of the best midfielders without possession in comparison to Europe’s elite

Can't see tweet? Click here!

In the Bundesliga, Thiago shows again that he is an exceptional footballer. Vidal also has great stats regarding interceptions. Looking ahead, Rudy is another player of interest and the stats are here to confirm it.

Image
Image
User avatar
ottackon
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membership
 
Posts: 6130
Joined: Sat Jul 20, 2013 12:05 am
Location: Casablanca
National Flag:
Morocco
Has thanked: 1830 times
Been thanked: 3188 times
Gender: Male

Re: The Statistics Thread

Postby JANCKER » Thu Mar 23, 2017 5:50 pm

.........................................Image.........................................
User avatar
JANCKER
I tattooed BayernForum.com on my ass
 
Years of membershipYears of membershipYears of membership
 
Posts: 17582
Joined: Thu Jun 19, 2014 2:03 am
Location: 710 miles SE from Allianz Arena
National Flag:
Basque Country
Has thanked: 9162 times
Been thanked: 8468 times
Gender: Male

Re: The Statistics Thread

Postby Manchu » Thu Mar 23, 2017 10:55 pm

I would like to introduce you all to a very interesting and potentially very useful stat known as goalimpact. You see, most “comprehensive” stats like whoscored’s ratings attempt to rank players based on their actions on the field like shots, key passes, and dribbles. This has two primary problems. One, it is very hard to determine the true value of an action because the value of an action is highly situational. Dribbling someone close to your own goal is probably less valuable than doing it close to the goal of the other team. Two, players spend 95% of their time without the ball, yet nearly as statistics are only collected for actions related to the ball. Blocking a passing lane through your presence is worth a lot, yet it does not appear in any statistic I’m aware of.

Goalimpactis an attempt to get around this limitation through looking at the only thing that really matters in football: results. It’s a measure of your team’s goal difference when you are on the field vs when you are not on the field. As substitutions are rare and goals are highly random, the algorithm used needs a very large number of minutes played(and not played) in order to get a good idea about how good you are. Thus, it is predictive measurement based on your entire career(with results modified in accordance to an aging curve because 18 year old players are much worse than 26 year old players), not a measurement of how well you played last season.

I must point out that it does take into account how good your teammates are, how good the opposition is, and home field advantage. Thus, when Bayern Munich beat Barcelona 3-0 away, it almost certainly helped Thomas Muller’s goalimpact more than when Bayern Munich beat Werder Bremen 6-0 at home. And if Werder Bremen were to beat Barcelona 3-0, it would the goalimpacts of the players on their team an awful lot.

Now that I’ve described the metric to you all, it is time to examine the Top 50 list of the best players according to goalimpact published in December 2016. Bayern Munich has 9 players on that list, more than any other club. According to the goalimpact algorithm, the best players in the world are
1. Mario Gotze
2. Thomas Muller
3. Sergio Busquets
4. Messi
5. Pique
6. Pedro
7. Manuel Neuer
8. Toni Kroos
9. Benzema
10. Ozil
A lot of you are probably skeptical of the high placement of Mario Gotze, but I’m actually interested. You see, this algorithm only sees the results of games and substitutions within. It doesn’t know that Messi scores 30+ goals a season and has 5 ballon d’Ors. It doesn’t know that Neuer is considered the best keeper in the world. It doesn’t know that Kroos is considered to be the best set-piece taker in the world. Yet it picks out those players as some of the best. In fact, of the 10 players there, 8 of them are probably not a big surprise to the knowledgeable football fan.

That leaves us with the two surprises: Pedro and Gotze. However, if the algorirthm is so good at picking out good players(it was actually designed for youth scouting), the question is as follows: are we the ones mistaken or is the algorithm mistaken? Are these players doing things that help their team succeed that we don’t notice?
I don’t know the answer to these questions. I have never really watched Pedro play, although the little I’ve seen makes me wonder if he is a Thomas Muller like player. I don’t know what is up with Gotze, although I have my suspicions(pressing, defensive work, beginning attacks, changing styles of play, and too high expectations). The diagnoses of myopathy only muddies the waters further.

However, I still believe goalimpact to be the single most objective measurement of player performance(and for those annoyed at not seeing Lewandowski and Thiago within the top 10, they are both rated high on the top 50 list and Thiago hasn’t even peaked yet).

On the other hand, goalimpact has some flaws. One of them is underrating very old players with access to Bayern’s patented anti-aging formula whose performance is expected to decrease with age but hasn’t yet. Another is that it doesn’t account for managers misusing players as this will likely cause the goalimpact of the player to drop even if they are still just as good in the right system. Coincidently, Thomas Muller’s goalimpact increased under van Gaal, peaked under Heynckes, decreased under Guardiola, and is now increasing under Ancelotti.
User avatar
Manchu
I'm a chatterbox!
 
Posts: 320
Joined: Wed Mar 15, 2017 6:34 am
Has thanked: 243 times
Been thanked: 255 times
Gender: Male

Re: The Statistics Thread

Postby PunkCapitalist » Fri Mar 24, 2017 12:47 am

Manchu wrote:I would like to introduce you all to a very interesting and potentially very useful stat known as goalimpact. You see, most “comprehensive” stats like whoscored’s ratings attempt to rank players based on their actions on the field like shots, key passes, and dribbles. This has two primary problems. One, it is very hard to determine the true value of an action because the value of an action is highly situational. Dribbling someone close to your own goal is probably less valuable than doing it close to the goal of the other team. Two, players spend 95% of their time without the ball, yet nearly as statistics are only collected for actions related to the ball. Blocking a passing lane through your presence is worth a lot, yet it does not appear in any statistic I’m aware of.

Goalimpactis an attempt to get around this limitation through looking at the only thing that really matters in football: results. It’s a measure of your team’s goal difference when you are on the field vs when you are not on the field. As substitutions are rare and goals are highly random, the algorithm used needs a very large number of minutes played(and not played) in order to get a good idea about how good you are. Thus, it is predictive measurement based on your entire career(with results modified in accordance to an aging curve because 18 year old players are much worse than 26 year old players), not a measurement of how well you played last season.

I must point out that it does take into account how good your teammates are, how good the opposition is, and home field advantage. Thus, when Bayern Munich beat Barcelona 3-0 away, it almost certainly helped Thomas Muller’s goalimpact more than when Bayern Munich beat Werder Bremen 6-0 at home. And if Werder Bremen were to beat Barcelona 3-0, it would the goalimpacts of the players on their team an awful lot.

Now that I’ve described the metric to you all, it is time to examine the Top 50 list of the best players according to goalimpact published in December 2016. Bayern Munich has 9 players on that list, more than any other club. According to the goalimpact algorithm, the best players in the world are
1. Mario Gotze
2. Thomas Muller
3. Sergio Busquets
4. Messi
5. Pique
6. Pedro
7. Manuel Neuer
8. Toni Kroos
9. Benzema
10. Ozil
A lot of you are probably skeptical of the high placement of Mario Gotze, but I’m actually interested. You see, this algorithm only sees the results of games and substitutions within. It doesn’t know that Messi scores 30+ goals a season and has 5 ballon d’Ors. It doesn’t know that Neuer is considered the best keeper in the world. It doesn’t know that Kroos is considered to be the best set-piece taker in the world. Yet it picks out those players as some of the best. In fact, of the 10 players there, 8 of them are probably not a big surprise to the knowledgeable football fan.

That leaves us with the two surprises: Pedro and Gotze. However, if the algorirthm is so good at picking out good players(it was actually designed for youth scouting), the question is as follows: are we the ones mistaken or is the algorithm mistaken? Are these players doing things that help their team succeed that we don’t notice?
I don’t know the answer to these questions. I have never really watched Pedro play, although the little I’ve seen makes me wonder if he is a Thomas Muller like player. I don’t know what is up with Gotze, although I have my suspicions(pressing, defensive work, beginning attacks, changing styles of play, and too high expectations). The diagnoses of myopathy only muddies the waters further.

However, I still believe goalimpact to be the single most objective measurement of player performance(and for those annoyed at not seeing Lewandowski and Thiago within the top 10, they are both rated high on the top 50 list and Thiago hasn’t even peaked yet).

On the other hand, goalimpact has some flaws. One of them is underrating very old players with access to Bayern’s patented anti-aging formula whose performance is expected to decrease with age but hasn’t yet. Another is that it doesn’t account for managers misusing players as this will likely cause the goalimpact of the player to drop even if they are still just as good in the right system. Coincidently, Thomas Muller’s goalimpact increased under van Gaal, peaked under Heynckes, decreased under Guardiola, and is now increasing under Ancelotti.

I've been a huge fan of goalimpact for a while. Götze's score is indeed strange.

I haven't looked at the fine details to know if this suspicion is true but I think that goalimpact might be less precise for players who don't miss games (either because of injuries and/or rotation). If there is not enough data for games in which the player didn't partake, it should be harder to evaluate his impact.

Also, a second thing that goalimpact might have problems dealing with is bench quality. For instance; imagine there are two teams with the same squad in which Messi starts but; on one team, Messi's sub is Hazard and in the other team it's Oscar. Because the difference between Messi and Oscar is wider than that of Messi and Hazard, you would expect that Messi's goal impact in the Messi-Hazard team would be lower than in the Messi-Oscar team.

I don't mean this as a comment on Götze, These are general observations.

I'm guessing this report is for jan16 to dec16? I would expect Thiago's goal impact to rise as he accumulates appearances under Carlo.

Sent from my XT1064 using Tapatalk
User avatar
PunkCapitalist
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membership
 
Posts: 5512
Joined: Sun Dec 22, 2013 6:15 pm
Has thanked: 2381 times
Been thanked: 2268 times
Gender: Male

Re: The Statistics Thread

Postby Manchu » Fri Mar 24, 2017 3:40 am

PunkCapitalist wrote:I've been a huge fan of goalimpact for a while. Götze's score is indeed strange.

Yes, it does seem strange. On the other hand, looking at a relatively recent graph of Gotze's goalimpact over time(the dark red line represents his calculated goalimpact, the light red line represents his calculated peak(at age 26), and the blue line represents his goalimpact solely based on his performance over the last two years for anyone unfamiliar with this metric), there doesn't seem to be anything weird going on. He improves at about the same rate no matter what team he's playing for. Evidently he's doing something well. In fact, all of this would suggest that he actually played well as a utility player and is far more versatile than people give him credit for. Or he's just insanely lucky and teams should just play him anyways in order to tap into his luck.

To be perfectly honest, I think at this point we have to accept that there's an excellent chance that Gotze's one of the best players in the world when healthy but Guardiola screwed over his career because he expected Gotze to play like Messi, as in the same style as Messi. I also find it quite interesting that another player Guardiola rejected, Xherden Shaqiri, is also in the top 50 list, although I wasn't paying a lot of attention to Bayern when that all went down.
I haven't looked at the fine details to know if this suspicion is true but I think that goalimpact might be less precise for players who don't miss games (either because of injuries and/or rotation). If there is not enough data for games in which the player didn't partake, it should be harder to evaluate his impact.

Yeah, that's a problem, although it's more with players always playing together than with a player always playing, because if one player always plays but the rest of the squad rotates frequently the algorithm should be able to get a good idea about how good he is, but if two players always play together it has no idea who is the better one. However, that situation almost never happens in the long run due to injuries and transfers if nothing else, so algorithm basically just needs more games in order to get a good idea about how good players who usually play together are.
Also, a second thing that goalimpact might have problems dealing with is bench quality. For instance; imagine there are two teams with the same squad in which Messi starts but; on one team, Messi's sub is Hazard and in the other team it's Oscar. Because the difference between Messi and Oscar is wider than that of Messi and Hazard, you would expect that Messi's goal impact in the Messi-Hazard team would be lower than in the Messi-Oscar team.

I don't mean this as a comment on Götze, These are general observations.

That's not actually really that much of a problem. Let's say two players-let's call them Gotze and Muller-are both equally good and are both excellent, but their coach never plays them together. Thus, the team preforms just as well when Muller is playing as when Gotze is playing, and their goalimpacts are calculated to be equal.

Now, let's say that for some reason, possibly having to do with injuries from a fight over who's better or said coach being an idiot, neither Gotze or Muller play for several games. The team performs much worse without them, and both their goalimpacts rise a lot but remain approximately equal. Thus, the fact that your bench is made up of very good players doesn't hurt the goalimpacts of the players being subbed off because the team is expected to perform better with the very good players from the bench on the pitch than it would if bad players were coming on.

I'm guessing this report is for jan16 to dec16? I would expect Thiago's goal impact to rise as he accumulates appearances under Carlo.

Sent from my XT1064 using Tapatalk

No, it's of the standings of the players goalimpact scores as they appeared on December 6th 2016. Like almost all things involving goalimpact, the estimates are based on the players' entire careers because just basing them on one season or one year would be very noisy and heavily benefit players who had gotten lucky.

That being said, Thiago does seem to have been preforming extremely well recently, judging by his graph from December.
User avatar
Manchu
I'm a chatterbox!
 
Posts: 320
Joined: Wed Mar 15, 2017 6:34 am
Has thanked: 243 times
Been thanked: 255 times
Gender: Male

Re: The Statistics Thread

Postby JANCKER » Mon Mar 27, 2017 3:15 pm

Bayern has conceded 0 headed goals this season?
.........................................Image.........................................
User avatar
JANCKER
I tattooed BayernForum.com on my ass
 
Years of membershipYears of membershipYears of membership
 
Posts: 17582
Joined: Thu Jun 19, 2014 2:03 am
Location: 710 miles SE from Allianz Arena
National Flag:
Basque Country
Has thanked: 9162 times
Been thanked: 8468 times
Gender: Male

Re: The Statistics Thread

Postby Manchu » Wed Mar 29, 2017 3:43 am

I found this absolutely must read article about statistics and how they relate to an undervalued player. It's about basketball, but many of the same principles apply to football, and it is particularly notable how basketball players maximize visible stats. I expect that very much the same thing goes on in regards to football, and perhaps we shouldn't give young players who score a lot or have a high pass completion rate all the attention we do.

Battier’s game is a weird combination of obvious weaknesses and nearly
invisible strengths. When he is on the court, his teammates get better, often a lot
better, and his opponents get worse — often a lot worse. He may not grab huge
numbers of rebounds, but he has an uncanny ability to improve his teammates’
rebounding. He doesn’t shoot much, but when he does, he takes only the most
efficient shots. He also has a knack for getting the ball to teammates who are in a
position to do the same, and he commits few turnovers. On defense, although he
routinely guards the N.B.A.’s most prolific scorers, he significantly reduces their
shooting percentages. At the same time he somehow improves the defensive
efficiency of his teammates — probably, Morey surmises, by helping them out in all
sorts of subtle ways. “I call him Lego,” Morey says. “When he’s on the court, all the
pieces start to fit together. And everything that leads to winning that you can get to
through intellect instead of innate ability, Shane excels in. I’ll bet he’s in the
hundredth percentile of every category.”

----

"When I ask Morey if he can think of any basketball statistic that can’t benefit a
player at the expense of his team, he has to think hard. “Offensive rebounding,” he
says, then reverses himself. “But even that can be counterproductive to the team if
your job is to get back on defense.” It turns out there is no statistic that a basketball
player accumulates that cannot be amassed selfishly. “We think about this deeply
whenever we’re talking about contractual incentives,” he says. “We don’t want to
incent a guy to do things that hurt the team” — and the amazing thing about
basketball is how easy this is to do. “They all maximize what they think they’re being
paid for,” he says. He laughs. “It’s a tough environment for a player now because you
have a lot of teams starting to think differently. They’ve got to rethink how they’re
getting paid.”
Having watched Battier play for the past two and a half years, Morey has come
to think of him as an exception: the most abnormally unselfish basketball player he
has ever seen. Or rather, the player who seems one step ahead of the analysts,
helping the team in all sorts of subtle, hard­to­measure ways that appear to violate
his own personal interests. “Our last coach dragged him into a meeting and told him
he needed to shoot more,” Morey says. “I’m not sure that that ever happened.” Last
season when the Rockets played the San Antonio Spurs Battier was assigned to
guard their most dangerous scorer, Manu Ginóbili. Ginóbili comes off the bench,
however, and his minutes are not in sync with the minutes of a starter like Battier.
Battier privately went to Coach Rick Adelman and told him to bench him and bring
him in when Ginóbili entered the game. “No one in the N.B.A. does that,” Morey
says. “No one says put me on the bench so I can guard their best scorer all the time.” [source]



This does make me wonder a lot about Gotze and Muller. Judging from their goalimpacts, they seem to be the football equivalents to this Battier guy, although probably more offensively than defensively. And Muller at time rakes up goals almost incidentally to what he brings to the team.

Maybe we should pay attention to the words of Badstuber on the subject of Muller:
“He never stood out in the youth teams, but… the team that picked him usually always won” [source]
User avatar
Manchu
I'm a chatterbox!
 
Posts: 320
Joined: Wed Mar 15, 2017 6:34 am
Has thanked: 243 times
Been thanked: 255 times
Gender: Male

Re: The Statistics Thread

Postby Manchu » Wed Mar 29, 2017 7:31 am

PunkCapitalist wrote: Götze's score is indeed strange.


I rewatched the highlights of all Bundesliga games Gotze played last season, and I think I have a couple ideas about what he was doing to help the team.

One of the issues is that Gotze is/was insanely involved in the offense. He touched the ball on many(perhaps even the majority) of plays where Bayern scored, and this contact usually came relatively shortly before Bayern scores. Often times, Gotze's pass isn't the assist, but instead the pass before the assist, or the pass before the pass before the assist. Still, it's obvious that's he really good at ball distribution.

Second, he seems to have good positional awareness, although not as good as Muller's awareness. He sometimes manages to open space for other Bayern players* and often manages to get unmarked, although he doesn't end up scoring many goals of that.

Obviously, this isn't a perfect method and doesn't give me a good idea of the defensive aspect of the game, but it's a start.

Moreover, I would like to point out that when Muller, Gotze, and Lewandowski played together, they were often just sublime. This only happened for 1 and 1/2 games last season, but the results were excellent as Bayern scored 10 and conceded 1 against Wolfsburg(before they were completely terrible) and Dortmund(although admitably Bayern was at home). In goalimpact terms, that was the best attacking trio in the world at the time, and I don't think goalimpact is necessarily wrong. Of course, the way that Bayern played and scored when they were together didn't fit well with Guardiola's "send it outside and cross it" plans for the season, but I think it's entirely possible that Gotze could have nailed down a starter spot if he hadn't been injured so badly(due to his myopathy) after the 5-1 against Dortmund. I much also confess that Douglas Costa was also on the field with the three during these two games, but he was much less offensively involved.

*This is interesting considering how Joachem Low attempted to solve Germany's striker problem by playing Gotze as a "pseudo 9" in the away game against Norway, with idea that Gotze's job was not drop into the midfield like a traditional false 9(and like he's usually played upfront for Germany), but instead keep his position upfront and draw the attention of the defense so that Muller, Ozil, Kroos, and company could play around him to maximum efficiency. Judging for Gotze's 8 touches on the ball and the 3:0 scoreline in favor of Germany with two Muller goals, it seemed to work, but it is probably not a tactic Low would use if he had an actually world class striker.
User avatar
Manchu
I'm a chatterbox!
 
Posts: 320
Joined: Wed Mar 15, 2017 6:34 am
Has thanked: 243 times
Been thanked: 255 times
Gender: Male

Re: The Statistics Thread

Postby ottackon » Wed Mar 29, 2017 12:44 pm

Image
User avatar
ottackon
I live on BayernForum.com
 
Years of membershipYears of membershipYears of membershipYears of membership
 
Posts: 6130
Joined: Sat Jul 20, 2013 12:05 am
Location: Casablanca
National Flag:
Morocco
Has thanked: 1830 times
Been thanked: 3188 times
Gender: Male

Re: The Statistics Thread

Postby Dumbledore7 » Mon Apr 10, 2017 12:22 pm

Manchu wrote:Moreover, here's an articlewhich touches on both the follies of applying "traditional" statistics, and the advantages of applying "scientific statistics to basketball. Football is a much more complex and chaotic game with a smaller sample size than basketball, so we'll probably never see statistics applied to that degree, but the article is still interesting.

This is interesting. Couldn't follow it fully because I don't follow basketball at all but I understand how it could work.

Manchu wrote:You're right(although I think too many people only think of Muller as a good goal poacher.). A lot of people noticed it before I did. You have to understand, until about a year ago, I paid very little attention to professional football outside the World Cup. When I did, I wasn't watching for the right things. The Muller experience helped me learn what I should pay attention to.

Moreover, what I use goalimpact is an indication of which players I should scout. That is what my example was. That's also how I found about Brandt and started watching him and how I realized that Paulo Dybula is highly overhyped(I can never forget watching a 10 minute highlight video of him and realizing that the creators had stuck every single example of him being effective across multiple seasons in the first five minutes.)

It's a useful tool because, again, I can't watch everyone to the required level of attention.

Fair enough. In fact it's quite good that you can learn and embrace Müller's profile within just one year, especially when this last year was probably the worst of his career.

Manchu wrote:I'm curious, what exactly do you think of Mario Gotze's ball distribution? He has a very unusual goalimpact graph in that he continued to improve constantly during his time at Bayern despite everyone thinking he was a flop. Nearly all other major flops who started out promising (Torres, Mario Balotelli, Freddy Adu) have rapidly decreasing goalimpacts and clearly fail to live up their potential, but Gotze seems quite unusual, despite having relatively few goals and assists. I watched a lot of his play, and I think what is going on is that he has very good ball distribution in the attack, but I'm not sure.

A lot of factors. First, I always think Götze is a supreme player, maybe even a bit biased for him than most of the people here. He has the "bottler" tag on him, never being able to fulfil his potential. I think this is because he never gets a good run of games, he dribbles a lot near the box which really highlights the moment where he lost the ball to a lot of people, most of his better games are just average performances plus a couple of moments of magic. He also played in a slow system where he is required to contribute in deeper positions, whereas he's an end-product player (the alternative is being a winger, which is not his profile). Also note that he has this terrible luck of always getting injured just as the team started to perform better under Pep. From that alone I think the impression on him was not accurate in the first place. When it comes to it though, I think whenever he didn't fail to dribble (he did seem to fail a lot) and the ball left his feet on his own accord, he nearly always makes the right decision. The technique in his locker is better than most Bayern players if he can execute it.

The other reason would have to be the Pep system. Remember this hinged on players making all the right passes at all the right times, and Pep is outstanding at coaching that. How much can be attributed to Pep or Götze I don't know, but surely the high-line possession system helps bring every forward better numbers in the final third.

Manchu wrote:We can't. We can't prove anything beyond doubt, and we don't want to. We're trying to identify as set of players that we should scout, not make decisions just based off of statistics. By using the criteria that I selected, we're going to miss some talented players who are the victims of bad crosses and get some players that are the beneficiaries of good crosses. However, actual skilled target forwards are more likely to be included on our list, and are thus more likely to be scouted, and hopefully we can find one. In any case, it's better than sending scouts to random games and hope that we get lucky.

Maybe not in the way that we're talking about cross conversions specifically, but clubs must send scouts based on something, which are likely to be impressive statistics. So fair enough.

Manchu wrote:For example, I just ran a calculation tonight and found that Bayern Munich conceded many more goals and scored many fewer goals when Xabi Alonso is on the field than when he isn't, both home and away. The difference is so large(literally a difference in gd of 80 goals per season) that it almost certainly isn't coincidental despite the fact that Xabi sometimes plays against tougher opponents(in the Bundesliga, Bayern have conceded a goal every 140 minutes when Xabi Alonso plays and a goal every 278 minutes when Xabi Alonso doesn't play), but the issue still needs to be approached carefully. I won't advocate benching XA until I understand what's going on. My suspicion is that Thiago is better at deep defense than XA while Muller is better at counterpressing than Thiago.

As you pointed out, still not evidential enough to conclude the correlation. You'd have to watch all the goals conceded to even subjectively judge whether he cost us those goals. On the other side, it's clearly evident that we nullified both Leipzig and Dortmund with him on the pitch. Vidal and Thiago are as mobile as they come to prevent most counter-attacks and Alonso was always there in the centre to recycle the ball and slow down the play. The best Bundesliga teams at attacking, and they were hopeless with Alonso on the pitch and Müller on the bench. Of course there's definitely a stigma of being a slow old player with him, but this season is the one throughout his Bayern career where this was the least visible.

Let's say, for the sake of argument, that his slowness in one way or another does weaken our defense. Still does not mean that Müller should start in his place. For one, I think Thiago is a better high presser than Müller this season and he should be well forward. I think Thiago-Alonso-Vidal decreases the probability of conceding a counter-attack, even if someone wants to argue that Alonso is weaker when we concede the counter-attack. We saw this on Saturday - Dortmund had absolutely no chance.
User avatar
Dumbledore7
I'm a post king!
 
Years of membershipYears of membershipYears of membership
 
Posts: 2930
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 2033 times
Been thanked: 1857 times
Gender: Male
BayernForum.com fan club: Active member

Re: The Statistics Thread

Postby Manchu » Tue Apr 11, 2017 1:32 am

Dumbledore7 wrote:Fair enough. In fact it's quite good that you can learn and embrace Müller's profile within just one year, especially when this last year was probably the worst of his career.

You shouldn't give me too much credit. The main time that I was figuring out this stuff was over the summer, when Muller wasn't playing very much, so I was mostly watching old videos and games, and I also took advantage of the fact that other people had already thought about this before and written about it.

However, the turning point was when I watched Gotze's goal in the world cup final and realized that Muller was crucially involved even though he never touched the ball.
A lot of factors. First, I always think Götze is a supreme player, maybe even a bit biased for him than most of the people here. He has the "bottler" tag on him, never being able to fulfil his potential. I think this is because he never gets a good run of games, he dribbles a lot near the box which really highlights the moment where he lost the ball to a lot of people, most of his better games are just average performances plus a couple of moments of magic. He also played in a slow system where he is required to contribute in deeper positions, whereas he's an end-product player (the alternative is being a winger, which is not his profile). Also note that he has this terrible luck of always getting injured just as the team started to perform better under Pep. From that alone I think the impression on him was not accurate in the first place. When it comes to it though, I think whenever he didn't fail to dribble (he did seem to fail a lot) and the ball left his feet on his own accord, he nearly always makes the right decision. The technique in his locker is better than most Bayern players if he can execute it.

The other reason would have to be the Pep system. Remember this hinged on players making all the right passes at all the right times, and Pep is outstanding at coaching that. How much can be attributed to Pep or Götze I don't know, but surely the high-line possession system helps bring every forward better numbers in the final third.


Again, Gotze is an exceptionally interesting case. By the end of his time with Bayern, he was being rated by goalimpact as one of the best attacking players in the world alongside Messi and Muller. Moreover, we now understand why he was having so much trouble dribbling, why he was constantly getting injured, and so on.

Moreover, Gotze’s rate of improvement remained relatively constant as he went from Dortmund to Bayern to Dortmund again, so he didn’t seem to particular depend on a system or even playing a particular position.

As for Gotze getting injured at inopportune moments, I wasn’t watching Bayern at the time, but just judging from old highlight videos and results, he was on the verge of forcing himself into the first choice starting lineup, having played a massive role in both Lewandowski’s 5 in 9 and Bayern’s 5-1 victory over Dortmund, when he got injured on international break last season.

Of course, goalimpact would also suggest that any Bayern optimal lineup should have contained Muller, Gotze, and Lewandowski, and those two are the only games that Bayern played like that.

I do find the question of Gotze playing on the outside quite interesting. It’s obvious that he can’t act as a traditional winger there(although, like Muller, if he ends up in some space there, his crosses actually seem to be very accurate). Moreover, he has had trouble dribbling in the last few years(although hopefully that will change if he gets successfully treated for his disorder), so having him cut in from the left and shoot wasn’t a good tactic to rely on when he was at Bayern.

However, I’m not sure if he was worse playing from the wings. If, as I hypothesized(and I need to go back and watch a lot of games focusing only on Gotze), one of the most important aspects of his game was ball distribution(pass before the pass before the goal, so to speak) in the final third, it’s likely that he could do just as well from the wings like Julian Brandt does as he could in the center.

Finally, when I was researching Gotze and trying to figure out what was going on with him, I realized that Thomas Muller might actually have one of the best understandings about Gotze’s play, being an intelligent player that’s played a lot with and without Gotze. He doesn’t appear to have ever said anything explicit about Gotze’s style of play, but he did say this:
‘I’ve talked to Mario, of course we’ve talked in the dressing room and I fully appreciate that he might not be satisfied at Bayern Munich,’ said Muller.

‘From a purely sporting point of view, of course you want to be on the pitch, you want to be showing the public what you can do.

‘But he jokes around, he is full of spirit, and in training we really dovetail on the pitch and in the club and this won’t go away in a hurry. [source]

Given that the entire club hierarchy was trying to force Gotze out by that point(March of 2016), it's somewhat interesting Muller appears to really want Gotze to stay and appears to value something that Gotze brings to the team.



Maybe not in the way that we're talking about cross conversions specifically, but clubs must send scouts based on something, which are likely to be impressive statistics. So fair enough.

Exactly. The entire idea of correctly applied sports statistics is to shift the situation ever so slightly in your favor.

As you pointed out, still not evidential enough to conclude the correlation. You'd have to watch all the goals conceded to even subjectively judge whether he cost us those goals. On the other side, it's clearly evident that we nullified both Leipzig and Dortmund with him on the pitch. Vidal and Thiago are as mobile as they come to prevent most counter-attacks and Alonso was always there in the centre to recycle the ball and slow down the play. The best Bundesliga teams at attacking, and they were hopeless with Alonso on the pitch and Müller on the bench. Of course there's definitely a stigma of being a slow old player with him, but this season is the one throughout his Bayern career where this was the least visible.

Let's say, for the sake of argument, that his slowness in one way or another does weaken our defense. Still does not mean that Müller should start in his place. For one, I think Thiago is a better high presser than Müller this season and he should be well forward. I think Thiago-Alonso-Vidal decreases the probability of conceding a counter-attack, even if someone wants to argue that Alonso is weaker when we concede the counter-attack. We saw this on Saturday - Dortmund had absolutely no chance.


Yeah, it’s hard to draw definite conclusions. That being said, the level of correlation is extreme, or I wouldn’t have even brought it up. I really do need to rewatch every single goal we’ve conceded this season in order to figure out if anything is happening, but I should probably give you the raw numbers that led to my suspicion that Xabi Alonso weakens the team first:

In the Bundesliga:
Home with XA: 25 goals scored, 6 goals conceded, for GD of +19, in 893 minutes of play, for 2.52 goals scored and 0.64 conceded every 90 minutes.

Home without XA: 23 goals scored, 0 goals conceded, for GD of +23, in 367 minutes of play, for 5.64 goals scored and 0 goals conceded every 90 minutes.

Away with XA: 11 goals scored, 6 goals conceded, for GD of +5, in 791 minutes of play, for 1.25 goals scored and 0.68 conceded every 90 minutes.

Away without XA: 12 goals scored, 3 goals conceded, for GD of +9, in 469 minutes of play, for 2.30 goals scored and 0.58 goals conceded every 90 minutes.

Of course, if I had been less lazy and included Champions League data, the data would probably favor XA more because of the one game in Rostov, but at the very least, this data suggests that Alonso should not be a regular Bundesliga starter anymore, and he isn’t. I’m not that surprised by this, because Xabi Alonso is of the age where your mind knows where to go, but your body can’t take you there anymore, which is why he’s retiring at the end of this season; he did remarkably well to make it this far at the highest level.

Of course, there have been some games where we’ve played very well with Xabi Alonso on the pitch, so it’s possible that he still has what it takes to play a few more big matches. On the other hand, for the Dortmund game, the rest of Bayern’s lineup was extremely talented and in their primes, while the Dortmund lineup included many young and inexperienced players, and we were playing at home, so a big win was not entirely unexpected regardless of Xabi Alonso.
User avatar
Manchu
I'm a chatterbox!
 
Posts: 320
Joined: Wed Mar 15, 2017 6:34 am
Has thanked: 243 times
Been thanked: 255 times
Gender: Male

Re: The Statistics Thread

Postby Dumbledore7 » Tue Apr 11, 2017 2:01 am

Manchu wrote:Of course, if I had been less lazy and included Champions League data, the data would probably favor XA more because of the one game in Rostov, but at the very least, this data suggests that Alonso should not be a regular Bundesliga starter anymore, and he isn’t. I’m not that surprised by this, because Xabi Alonso is of the age where your mind knows where to go, but your body can’t take you there anymore, which is why he’s retiring at the end of this season; he did remarkably well to make it this far at the highest level.

The data suggested that we concede a certain number of goals in some games. Didn't suggest that Alonso shouldn't be a starter. That would also imply that Alonso not starting would... have a more positive impact on our league position somehow? If Müller starts and we play 4-2-3-1 with Thiago and Vidal as double pivot, how would we benefit? That we would concede less goals? For the sake of argument, let's suppose that. I however am also inclined to believe that we would have scored considerably less goals, given how absolutely bloody awful Müller was from September to December.

Manchu wrote:Of course, there have been some games where we’ve played very well with Xabi Alonso on the pitch, so it’s possible that he still has what it takes to play a few more big matches. On the other hand, for the Dortmund game, the rest of Bayern’s lineup was extremely talented and in their primes, while the Dortmund lineup included many young and inexperienced players, and we were playing at home, so a big win was not entirely unexpected regardless of Xabi Alonso.

Watch the game and note what I said - I didn't highlight the fact that we won big, I tend to not take much away from goal differences. Beyond two goals, too many chaotic factors come into place, impossible to assess what the third, fourth, fifth, sixth goals actually pertain to. No, I said Dortmund had no chance. They are a pressing team, and they couldn't press. Thiago was pressing aggressively in the final third, Vidal mopped up stray balls, and Alonso dictated play, even forayed to the right side to receive the ball in the right place and drove it elsewhere harmless. As detrimental as Alonso's pace is to defending a counter, it's a significantly different issue with preventing the opponent's press and I think it's a huge misunderstanding that he's susceptible to pressing. Leipzig tried, Arsenal tried, Dortmund tried, didn't work. Let Real Madrid try.
User avatar
Dumbledore7
I'm a post king!
 
Years of membershipYears of membershipYears of membership
 
Posts: 2930
Joined: Thu Jan 02, 2014 3:02 pm
Location: London
National Flag:
Indonesia
Has thanked: 2033 times
Been thanked: 1857 times
Gender: Male
BayernForum.com fan club: Active member

Re: The Statistics Thread

Postby JANCKER » Mon May 01, 2017 8:21 pm

.........................................Image.........................................
User avatar
JANCKER
I tattooed BayernForum.com on my ass
 
Years of membershipYears of membershipYears of membership
 
Posts: 17582
Joined: Thu Jun 19, 2014 2:03 am
Location: 710 miles SE from Allianz Arena
National Flag:
Basque Country
Has thanked: 9162 times
Been thanked: 8468 times
Gender: Male

Re: The Statistics Thread

Postby JANCKER » Mon May 01, 2017 8:58 pm

.........................................Image.........................................
User avatar
JANCKER
I tattooed BayernForum.com on my ass
 
Years of membershipYears of membershipYears of membership
 
Posts: 17582
Joined: Thu Jun 19, 2014 2:03 am
Location: 710 miles SE from Allianz Arena
National Flag:
Basque Country
Has thanked: 9162 times
Been thanked: 8468 times
Gender: Male

Re: The Statistics Thread

Postby JANCKER » Tue May 02, 2017 3:18 pm

.........................................Image.........................................
User avatar
JANCKER
I tattooed BayernForum.com on my ass
 
Years of membershipYears of membershipYears of membership
 
Posts: 17582
Joined: Thu Jun 19, 2014 2:03 am
Location: 710 miles SE from Allianz Arena
National Flag:
Basque Country
Has thanked: 9162 times
Been thanked: 8468 times
Gender: Male

PreviousNext

Return to General Bayern Discussion

Who is online

Users browsing this forum: Yandex [Bot] and 1 guest