A passing network is the application of social network analysis and the network theory put into football. Each player is essentially a node and the connections between them are the passes.
In most cases for scouts and data analysts, they will only have time to analyse the last 3 matches of their opponents. By using passing networks, it gives quick information in an easily digestible format and also saved them an extra 10+ hours of watching videos.
Of course, there are limitations to passing networks. It is a tool which aids analysing games, but it is important to understand its weakness.
Firstly, each node of the average location of a player’s touch. If they are switching sides frequently, their average will look central even if they never touch the ball in central areas. This is a clear downside and emphasises the need to also use video. On the other hand, you could also use a heat map or a dot touch map. This allows you to stay data based and will be more accurate than videos.
Another limitation is that it is an estimation of what happened. Despite the picture showing it, the right back didn’t pass to the wing 25 times along the exact same path. That information is available, just not via the passing network.
The third limitation is that by themselves, these don’t explain that much. They’re a collection of snapshots of the game and combine to make a bigger picture. It’s like watching a third of a film; sometimes you get a clear idea of the plot, whilst other times you will be surprised when your friends are talking about a surprise twist at the end.
They are useful; however, they must always be paired with other analysis to complete the picture.
Some are done vertically but it makes more sense when it is horizontal. This is because most people are used to watching football horizontally (as this is how it is shown when broadcasted). Although you can get footage from a camera behind goals, but the vast majority of football viewers don’t see this video as it is mainly used for tactics analysis.
One danger of passing networks is adding too much information on to the visual. If it has too much information then it becomes difficult to interpret. By adding the xG chain onto it, it has enough information to be useful, but it is still easy to digest.
Firstly, you need to understand what the xG chain metric (xGC) is – What is an xG Chain? – Football Opinions (sport.blog)
Every time a player is involved in a pass within the possession, they get xG chain credit. The sum of their involvement of the match and their colour is based off that.
This allowed analysts to take the network beyond just looking at stats and helps to examine the value of a player’s contribution to the match. Red is the higher end of the scale whilst green is at the lower end. This gives a sense of how non-attacking players contribute to valuable build-up play.
The side of the node = number of touches
Thickness of the line = number of passes between the two nodes
Colour of the node = scale of the xGC, green is the lower end and red is the higher end
Colour of the line = the total xGC of possession featuring a pass of A to B and vice versa.
Here you can see Liverpool pushed far forward. They had a lot of possession and created a few chances with a lot of players involved.
Whilst for Swansea, Fernando Llorente was the only player who had an average position in the opposition’s half. The Spaniard and Wayne Routledge both put up big numbers as the Swans came away with all three points.
Just one plot from Liverpool’s trip to Bournemouth. Roberto Firmino is seen out wide instead of centrally and had little impact on creating chances for his teammates. He is normally a large red circle but in this game he was an ineffective green. It also allows you to isolate one player across a number of games and positions and see what it does to their performances.
In Man City’s match against Middlesbrough, both team’s maps are extraordinary but for different reasons. The Citizens’ front three average touch position was around the edge of the penalty box and every outfield player is red or orange. Meanwhile Aitor Karanka’s side had almost no possession and created hardly anything either. Somehow the match ended 1-1 with an equaliser in the 91st minute, the xG was 2.50 to 0.21 in favour of Pep Guardiola’s side.
How do sides use them?
Typically, sides will take passing networks for the last 10 sides from their next opponents, dividing them into home and away games. This allows them to look for patterns such as:
Do their fullbacks get forward in possession? Do they leave space in behind that can be attacked?
Which player(s) provide the engine for their attacks?
Which players should be pressing triggers?
Which players have the most valuable touches?
To conclude, provided you understand the positives and negatives, they can provide a huge boost in analysing any teams. The addition of the xG chain metric adds a layer to the visuals within analysis that previously only was contained within the numbers.