The CA Data Viz Competition Participants

Here are the results for the competition we ran.

We decided to run a football data visualization competition, and we received 19 submissions for the competition. Big thanks to Dave Willoughby and Stratagem Technologies for being kind enough to provide the data to be visualized (and for powering most of our articles with comprehensive data, of course!). I’m also really happy with the sheer quality of the vizzes. All the participants have given us new ways to use data.

These visualizations aren’t arranged in any specific order. Take your time and a look at all the vizzes with their short descriptions. Note: you might have to open some of the images on a new tab and zoom in to get a good feel of what it’s all about.

Temma Hasegawa

Assist Heatmap

A chance creation heat map. This viz can roughly show how a team attacks. We can understand their tendency of attacking (from centre or wide). The viz can be modified to a chances created-against heatmap or to a certain player’s heatmap. This can be used for rough opposition analysis. Also, fans can see where attacking players create chances.

Nils Mackay

dataviz3

A corner profile visualization. It can be used as a starting point in opposition analysis when it comes to defending corners. Certain aspects of an opponent’s corner routine can easily be seen from this viz, such as the tendency to take inswinging/outswinging corners and the general aim of the corners. Also, the most dangerous players from the opponent are immediately clear. Video analysis and perhaps extra data analysis should be used by opposition analysts at clubs in addition to this visualization.

Freddie Wilson

Expected and Actual Chance Creation Chord Map

A chance creation chord diagram. It aims to communicate the attacking output of players in one team, quantified by xG + xA values of chances and visualised by arc length. It also shows the partnerships between players, displayed through chords (arrows) directed from the ‘chance assister’ to the ‘chance taker’. The arrows are coloured by ‘chance assister’ and go from purple to yellow in descending order of total expected assists.  Below the graphic for xG and xA is the graphic for actual goals and assists. These colours are from the colour blind-friendly Viridis package in R. Although this visualisation could be used in the first steps of opposition analysis to identify primary threats and trends in terms of chances assisted and taken, the main target audience will be stats-inclined fans.

Erdi Myftaraga

By Barca Numbers (Erdi)
Patterns in chance creation. The basic idea of the viz is to visualize different passes from which players create chances, trying to highlight their aspects such as “how vertical are the chances” etc. The main audience is the average fan.

David Quartey

David_Quartey_chance_dataviz_competetion_gif.gif
Highlights of the goals of the only Ghanaian footballer in the Chinese Super League, Frank Acheampong. The creator wanted to bring out the quality of his goals and the channels through which he’s typically fed with the ball, and wanted to capture this in an interesting way. The viz works for both the average fan as well as player/opposition scouts at clubs.

Yuriy Bakovych

primaryTypesVisSample
A heatmap comparing how different teams create chances. This viz is suited to the average fan.

Vito Dantona

competition

Assist and shot locations for the top 3 most prolific pairs of scorers and assisters. The bigger graphic shows the average locations of assists and shots and the smaller pictures detail the individual positions of shots and assists for each pair of players. The average fan is the target audience.

Kevin Shank

CSL Chance Quality vs Quantity.png

Chance quantity v chance quality. Aimed at the average fan, this data viz communicates the teams’ xG per game while also mapping the differences of chance creation in terms of quality and quantity across the league.  Ideally, a team would want to be in the upper right corner of the top right quadrant, creating good quality chances and lots of them and therefore more xG/game. In the bottom-left quadrant, teams create low-quantity, low quality chances, leaving them with the worst xG/game. While it is a simple scatter-plot, it is a powerful way to view a team’s expected goals per game since it offers a two-dimensional view of their chance creation.

Alexander Jobsis

Shandong Luneng

Team attacking and defensive tendencies. By clustering open play passes that led to a shot, this viz shows certain tactical trends at both ends. The thickness of each line indicates the xG of each chance, while the colour indicates the amount of goals scored (green=good; it means that a team is scoring more goals in attack and conceding less goals in defense). For instance, Shandong Luneng attacks dominantly via the flanks (thick green lines in the bottom half showing the dominant passes for high xG actions or actual goals) but concedes relatively a lot of goals when the ball is crossed from the right (red bar in the top half).

Vignesh Velu

dark-comp

A post-game graphic that shows a team’s chances, with the quality of each chance illustrated by the colour of their corresponding points. This viz is aimed at the average fan.

Abhinav Ralhan

Guangzhou_Evergrande (1).png

The purpose of this visualisation was to create to understand type of Chances (superb, great, good…) created by a team for every goal they scored. This can help validate your own xG model. The average fan can also see the positions from which their team is scoring.

Albert Edwards

final submission 2.png

A viz that shows that low crosses can come close to the goal and hence can be a good way of creating chances. The target audience is people working at clubs as they could use this information to improve performance by aiming to complete low crosses/cutbacks near the penalty area.

John-Paul Quinn

ChanceDVComp2

A chance map of Beijing Guoan, with the goals highlighted in blue. This viz allows you to see whether a team are creating good chances or not as well as displaying ‘where’ the chances/goals are coming from – perhaps having the benefit of demonstrating that most of the goals come from ‘good’ positions, whereas a lot of shots from outside the box are wasted. This could easily be adapted for chances conceded or for a particular player. The target audience could be analysts within a club that want to have an overall of view of their chance creation (or their opposition). It is also supposed to be something that is fun, nice to look at and easy to understand so is also accessible to fans.

Dan Clark

China Super League 2017 Chance Conversion.png

A graphic that shows the players with the best goal scoring conversion rates in the league. The colour indicates the number of goals a certain player has scored. The target audience would be the average football fan, showing them a stat in a simple and clear way.

Johannes Dusch

final graph (1).jpeg

The creator wrote:

“For this graph, I calculated a reference value of expected goals a player should be involved in (either as shooter, primary player or secondary player) given his club’s and position’s average and the amount of minutes played. I used my own model to calculate the xG of each shot in the dataset. I then randomly chose twelve players, three from each colour category, which are included in the attached plot.

The actual value of expected goals a player was involved in is displayed in the graph and compared to the reference value. The colours of the points are set according to the following criteria:
blue = above reference value, above 95% confidence interval
green = above reference value, within 95% confidence interval
yellow = below reference value, within 95% confidence interval
red = below reference value, below 95% confidence interval
I chose this method of visualization in order to get as much information about context into it while at the same time trying to keep it as easy to grasp as possible. It is mainly directed to scouts and other professionals as it is easy to detect outliers, which could e.g. be potential targets on the transfer market. But also fans/bettors might find this information useful.”

Julien Demeaux

Julien Demeaux

The average chance location and xG/chance of the top 10 chance creators in the 2017 CSL season. The viz is targeted at both fans as well as analysts at clubs.

Riley Wichmann

Riley's Steam Graph.jpg

A chance quality stream graph which displays the average chance quality and chances created by a team as the season progresses. It is aimed at the average fan, and the darker the colour shade, the higher the quality of the chances.

Kenneth Ang

chance_analytics_kenneth.png

The top 3 players who find themselves with chances under no defensive pressure in the 18-yard-box. While leaving these attackers alone may not necessarily mean conceding a goal, defenders playing against these three attackers could try and keep a closer eye on them by understanding the positions they take up from the viz.

Kyle Dijkstra

chance-analytics-viz.png

This viz displays the locations from which teams create and concede chances. This would be useful for clubs trying to scout where the opposition likes to attack, and where the opposition is most susceptible to being attacked.

 

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s