Women’s World Cup Semifinals

The semifinal diagram is here.

bPz9J5.md.png

The diagram definitely shows the advantage of using the blue and red spectrum colouring, it is much easier to see which are the biggest teams, and to see gradations of shade.

The national team closest to the centre is England, while the Huston Dash are the club team closes to the centre. The US is linked to others by Jodie Taylor who plays for Reign FC. The three other national teams form a pretty solid triangle.

Through this link to Jodie Taylor, Reign FC are one of the teams who will have someone in the final, no matter what the semifinal results are. The other two are Bayern Munich and Montpellier.

The club with the most players left is Manchester City with 8, followed by Chelsea with 7 and Arsenal with 6. I don’t know if this reflects the spending power of the Women’s Super League now that most clubs are attached to Premier League clubs, or if it’s pure chance.

Going to the community view, there are 4 teams left but 6 communities.

bPz2RD.md.png

Lyon and Wolfsburg, the two “extra” communities, link two national teams teams, with one player from each team, other clubs that are joining two teams have been put into the community of the nation who supplies the most players.

Of course none of this answers the most important question – who do I support when England play the US? The Lionesses or Megan Rapinoe’s team?!

Women’s World Cup Quarterfinals

We are now down to the last 8 teams, and I swear the schedule for the Women’s World Cup is shorter than the men’s because we’ve barely had time to breathe.

I’ve realised there was a question I meant to ask in the last post and didn’t. Does anyone know why so many of the Nigerian team played in the Norwegian league? Are they all following in a player’s footsteps (a la the Fijian rugby league players of Rochdale)? Or is it something else?

While the results of the second round have mostly gone to form (except Japan vs the Netherlands), I am starting to think Brazil enjoy making the Argentine men’s team look good at this knock out football thing. How can you be so good and still not win stuff?!

On to the diagrams:

Network diagram showing the links between the quarter final teams

The national team closest to the centre is England, and Manchester City are probably the club team closest but it’s hard to tell because of the way the teams are spread. The losses of Canada, Australia and Japan mean that the US is pretty much on it’s own with a heavy counterweight of European teams.

The club teams with the most players left in are Lyon (11), Bayern (10) and Chelsea (9). Barcelona lost most of its players when Spain lost.

The community view is less interesting because each country is its own community.

Same image as before, coloured by community

There seems to be far less overlap than there normally is in the men’s game and I don’t know if that’s because no “dominant” league has yet established itself. In the men’s you’d expect most of the finalists to play in one of the big 5 European leagues (Spain, England, Italy, Germany, France, and you can make the argument it’s really a big 2 of Spain and England), but that doesn’t seem to be the case here. Yet. Professional leagues still being a relatively recent thing in women’s football, and it will be interesting to see if this changes in the future.

Women’s World Cup 2019 Network Diagram

I had hoped to make a diagram like this for the Women’s Rugby League World Cup in 2017, but I couldn’t get enough information about which club team the players played for to make one.

I also wanted to have this ready for the start of the World Cup, but real life happened.

Without further ado, here is the team network diagram for the 2019 Women’s World Cup.

Unlabelled version – Network diagram of the teams at the Women's World Cup

Labelled version – Labelled version of the same diagram

I wanted to use both versions because the sheer overlap of some of the teams makes it hard to pick out the individual teams (looking at you so much United States and Canada).

Canada is the national team closest to the centre, while the Texas Longhorns are the club team closest to the centre.

The club team with the highest number of players involved is Barcelona, with 15 players, followed by Lyon on 14 and Chelsea and Manchester City with 12.

There are a lot more players who are down as unattached than I would expect, there’s a full 11 player. They’re not quite a full team because that 11 includes 2 goalkeepers but only 1 defender.

Jamaica are the only team where no-one plays in their national league, the US are the only team where everyone does. Interestingly, all the players for Sweden and the US play for club teams that have at least one other player from a different country also playing for them.

England and Scotland are a lot closer than I think the men’s team equivalents would be, possibly because a lot more Scottish women play in the English league than Scottish men do.

The national teams seem to be divided into those teams where most players play for a couple of club teams and those where their players play for many different club teams. It’s hard to see in the teams in the middle, but clear for those around the outside. I think Italy and Jamaica are probably the clearest example.
Figure showing Italy has players from very few teams while Jamaica has players from many

National teams with players from few clubs include Italy, Germany, Thailand, China and South Korea. Teams with players from many clubs include Argentina, South Africa, Chile, Jamaica and New Zealand. I have no idea if it means anything.

In the men’s World Cup, teams further away from the centre are more likely to go out early. If this is also true in the women’s, then it doesn’t look good for Thailand, South Africa, China and Japan. Because I know Japan are a strong team, I suspect they will be fine.

Community view of the diagram

As far as football is concerned, Australia is the 51st state. This is because a lot of the Australian players who play for non-Australian teams play for US teams that also have players on the US national team.

Also grouped together are England and Scotland (because lots of Scottish players play in England) and Sweden and Nigeria (because lots of Nigerian players play in Sweden).

Because they have players who play for the same teams and because they’re away from other teams, I was expecting Chile and Argentina to be grouped together but they’re not. I was also expecting Norway to be in the Nigeria/Sweden group, but they are not. This might be because Norway also has links to the Netherlands.

Further updates to come as the tournament progresses.

Film Locations

Once a year I like to look at where the films I’ve watched are set.

This is a deeply incomplete look at films up to August 2015.

Pie Chart

 

That’s all locations, including fictional ones, and, as you can see, the United States and the United Kingdom dominate.  They dominate even more if I only look at real places.

Pie Chart

 

If I look at the UK-set films

 

Pie Chart

 

They’re mostly set in England, and Wales and Northern Ireland have no representation at all. I really need to watch more films 🙂

2018 World Cup Final Diagram

What a match!

Probably the best final since 1994, although, admittedly, my memory only goes back to 1994.

Exciting, excruciating, and while the better team won, I feel terribly sorry for Croatia. Also, I told you that Lloris would drop a clanger.

No community view this time because once you’re down to two teams it doesn’t add any value.

Two team network

Five teams would have had a winner either way (Atletico Madrid, Barcelona, Juventus, Monaco, Real Madrid), this is a lot more than normal.

Possibly because very few Croats (2/23) play in their home league, and only 9/23 French players play in the French league.

At the last World Cup final, only two teams would have had a winner no matter who won, and those two teams were Real Madrid and Lazio.

Comparing it to the last two Euros, which I can do because this World Cup final was an all European final, in 2016 only Lyon would have had a winner whichever side won, and in 2012 there was only 1, Manchester City who had Mario Balotelli on one side and David Silva on the other.

Six clubs jointly had the most players in the final two teams, Atletico Madrid, Barcelona, Marseille, Monaco, Paris Saint Germain and Real Madrid.

Juventus are the club team nearest the centre.

2018 World Cup Semifinal Network Visualisations

Semi final team data viz

As Croatia are the team sticking out, I suspect that they are the least likely to progress.

France, Belgium and England are a giant smush still.

Spurs are the team with the most have most players left with 9, followed by Manchester City and Manchester United on 7. Manchester City lost several players when Brazil were knocked out.

Paris Saint Germain are the club team nearest the centre, while France are the national team nearest the centre.

The community view is interesting.

v0hA8p.png

There are 4 teams and 7 communities, one for each country and then 1 for Barcelona, Juventus and Liverpool. I don’t quite follow why Barcelona is separate but Monaco isn’t since they both connect 3 teams and have one player from 2 of the teams and 2 from the third. I presume Juventus are their own community because they have 1 player from Croatia and 1 from France and it’s more logical than giving Juventus to either country. However, following that logic, I’m not sure why Lovren is in the Croatia community rather than the Liverpool one, unless it’s because of the extra distance he is away from the others.

2018 World Cup Quarterfinal Network Diagrams

Quarterfinal network diagrams

Note how far Russia are from everyone else.

Manchester City remain the team with the most players (11) but they are now followed by Spurs (9) and Paris Saint Germain and Manchester United (8 each). Real Madrid and Barcelona’s numbers dropped rapidly and suddenly. I think Russia are responsible 😉

France or Croatia are the national side nearest the middle. It’s very hard to tell which is the closest given how far away all the teams are from the centre, mostly because Russia and Sweden are so far away from other teams that it distorts the graph.

Monaco are the club side nearest the middle.

When Shinji Okazi was removed from the network (because of Japan’s sad loss) it separated Belgium from England. I think that’s because the weight of Japan was pulling England down into Belgium.

Network diagram coloured by community

All 8 countries are their own communities. It’s interesting to see which teams have been linked to which countries. Barcelona, Real Madrid and Man City are all “Brazilian”, while Spurs, United and Liverpool are English. Chelsea are “Belgian” while Krasnodar are “Swedish”, which is probably news to them. Atletico Madrid are Uruguayan, but that at least makes sense.

2018 World Cup Second Round Network Diagrams

While most of the eliminated nations were, as expected, the ones around the edges of the diagram, the elimination of Germany has helped to reduce the crowding in what was the top of the diagram.

Network diagram of teams through to the second round of the 2018 World Cup

In the first round diagram Werder Bremen and Denmark (probably) were the club and national team closest to the centre of the diagram. I say probably because it was really hard to tell because even the closest national team was some way away from the centre.

In the second round diagram Genoa and Switzerland and the club and national team closest to the centre.

The club team with the most players still in hasn’t changed, it is still Manchester City, who have only lost one player, going down from 16 players to 15 players. The teams in second and third place haven’t changed but have changed order, because Barcelona are now in second place with 13, having lost only one player, while Real Madrid who had been in second place, having lost 3 players.

Community view of the teams in the second round of the 2018 World Cup

The community view has changed significantly. There is now only one multi-team community, England and Belgium, who remain a single community because of number of Belgian players who play in England. Several of the other multi-team groups have either had both members knocked out (Poland and Senegal and Tunisia and Egypt) or the removal of one team (Germany) has split the entire group.

World Cup 2018 Inter-Team Network Diagrams

Sorry that these are quite so late. The combo of the number of players, the relatively late date that the final squads were announced and a busy time at work meant that I couldn’t get the diagram anything like finished until the 12th, and I then went to Sweden for a week for work.

GO VISIT STOCKHOLM, IT’S LOVELY!

And then I wanted to tinker with the colours and force separation a little because part of the diagram was so overlapped.

Having now had the chance to tinker, please see the 2018 World Cup team interaction network diagram, with the players being attached to the team they last played for, with a cut off date of the 26th of May. The teams are as accurate as Wikipedia can make them. (You’re all lucky you missed the great “Inter” vs “Internazionale” revert fest.)

 

network vizualisation

As you can see, the top of the diagram is basically one giant smush, showing the international flavour of club football nowadays.

England are the only team where all of their players play in their home league, while Sweden and Senegal are the only teams where no-one plays in their home league. (A very simple bar chart pending on this topic some time soon.)

Iran, Panama and Peru are the least connected teams, while several others are so interconnected it’s hard to see them, and almost impossible to prevent them overlapping (looking at you, Croatia and Portugal).

This is one of the situations where the community view helps:

 

network diagram in community view

 

Using the community view makes it easier to see the team names. It also makes it easier to see which teams are tightly related to each other due to players either playing in each others leagues or players from both teams playing for teams from a third country.

There is a giant community of Croatia, France, Germany and Spain followed by a slightly smaller one of Brazil, Belgium, England (due to Manchester City). Then there are four two team groupings; Senegal and Poland, Australia and Saudi Arabia, Argentina and Colombia and Egypt and Tunisia. Everyone else is their own community.

It’ll be interesting to see how the inter-relations develop as teams are knocked out.

Yearly Book Location Data Viz

The yearly update on where the books I have read are set.

 

pie chart

 

There’s been a slight increase in non-UK set books, but in total, still more than half are set in the UK.

For UK-based books:

 

another pie chart

 

Still completely dominated by England. Theoretically, there should be one set in Scotland but I have had forgotten to write down the name of the book so it is not counting (until I can find it again).