How Technology Helps Measure Soccer Statistics and Tactics

Over the last 5-6 months, I have had different conversations with people who are working on technologies related to soccer.  And I recall writing about Tactical Analysis in Soccer for SportTechie some years back and thought I will reshare it here plus add a bit of update as some things have changed since then.

What are some of the Football Tactical Analysis Websites?

Squawka is a tactical analysis web application that provides a platform where one can view real-time and post-match statistics of (almost) everything that goes on in a football match. This includes time of possession, number of passes, number of shots, shot accuracy, chances created, tackles (or duels), blocks (or defensive actions), player stats etc. Match analysis data is presented in a field diagram with coloured dots and lines and heat maps. Those statistics can then be filtered by team and player for various types of analysis – whether it’s comparing player performance or looking at shots from one team or overall team events. A useful feature is the timeline scrolling which allows one to look at specific 5 min blocks of match activity.

Screen Shot 2014-09-01 at 9.46.42 PM

Player Performance Comparison

Screen Shot 2014-09-01 at 10.04.14 PM

Shots – where they were taken and where they landed

Screen Shot 2014-09-01 at 9.59.22 PM

Events Heat Maps

The comparison matrix is another interesting tool on the site that looks at stats over an entire season. For example, one can select five different teams in the 2013/2014 season of the Australian League and compare stats that they are interested in. The stats displayed can be filtered by ‘total for the season’, ‘average per game’ or ‘per 90 metrics’. One can also compare teams from different leagues and different seasons.

A special metric of this website is the Squawka Player Performance Score which is calculated using a large amount of data. This player performance score is broken down into “attack”, “defence” and “possession” statistics.

Screen Shot 2014-09-01 at 8.11.33 PM

Comparison Matrix Example

Four Four Two also provides match analysis data on their website in their Stats Zone section. It pretty much provides the same data available in Squawka, except the information is presented in a slightly different way and they don’t have a player performance score. The  Stats Zone allows all the match activities of a player to be viewed together in the Overall-player dashboard, instead of having to select the individual events in the Squawka dashboard. {Update: unfortunately FourFourTwo has discontinued Stats Zone due to limited resources}

Screen Shot 2014-09-01 at 10.56.06 PM.png

Stats Zone and what their Summary Statistics used to look like

Another site called Outside of the Boot writes commentaries and analysis on selected matches and supports their analysis using statistical data from the above two websites. Other than breaking down the statistics and what was going on in each match, they give reviews on the general style and tactics of a team, player or coach.

Where and how are the data collected?

Regardless of how the data is presented, what’s important is the reliability of the data and where their source is. Interestingly, both sites get their data from Opta, a sports data company that collect, package, analyse and distribute live data. Opta briefly explains on their website that their data collection process is labour intensive with three performance analysts assigned to each match; with one collecting all of the home team actions, one doing the away team and a third analyst checking the data for consistency and adding additional layers of data. They then run a full post-match check within 48 hours to ensure that the database is as accurate as possible.

But what exactly does each analyst do? By using their proprietary software, each Opta analyst puts in the live video feed of a match, then by using hotkeys, every activity that involves the ball is “tagged” – this “tagging” or tracking will record the time each activity started and ended and the X-Y coordinates of the start and end positions. For those who have used video analysis software like SportsCode (now owned by Hudl) or Dartfish, this will sound familiar. But what Opta has done is standardize their activity definition and tracking methods, so every analyst is trained to tag or code the exact same way. This means consistency in the data, allowing every match and every player to be compared using the same standards. The cool thing is, by feeding in historical matches (like all the past world cup matches), they can compare the performance of players from different decades. Check out this video that talks a bit more about what Opta Sports do:

Is there alternative (automated) technologies?

There are a number of athlete tracking technology out there that are either based on wearable technology (Catapult Sports, Tracktics, Polar Team, STATSports & SPT etc) or camera and image processing technology (Stats SportVU and TRACAB). The advantage of wearable sensors is that they can accurately track each athlete’s acceleration and impacts (in three axes) and some even track the players’ heart rate – something that is not possible with any of the current camera or video technology. But data from wearable sensors typically belong to the teams and not shared unless there is an arrangement with broadcasters. With Stats, they claim to not only track real-time 2D (X-Y) positioning data of the ball and the players, but its complex algorithms can also analyse and work out information like speeds, distances, possessions, passings, defence stats and turnovers. So technically, automating tactical analysis to some extent is possible but how much information can be made publicly available is another question.

Some final thoughts

The technology and methods used in Tactical Analysis in football have become more widespread over the years (and its still growing). The statistics that are made available can not only give punters additional information for betting, it can add new dimensions to watching each game. It provides viewers with a better understanding of what the players are actually doing (individually and as a team) and how they have been performing over a season with an unbiased quantified evaluation. For coaches and team managers, it means their decisions (in terms of training, strategising or even talent identification) don’t have to rely too much on gut feel but can be supported with numbers. How much they want to trust those numbers is another thing altogether.

If you know an exceptional app or technology in tactical analysis that is not mentioned here or maybe its still in development, feel free to leave a comment about it, and finally, thanks for reading!


Here are a few other related articles and blogs for those who like geek out a bit more on the topic:

  • Different Game: https://differentgame.wordpress.com/
  • Paper on big data and tactical analysis in elite soccer: link
  • Paper on tactical analysis using pattern recognition: link
  • Paper on a new tactical metric that looks at effective play: link
Advertisements

Leave a comment, share your thoughts. If your a SPAMMER or building back links to unrelated sites - this is not the site you are looking for!

Please log in using one of these methods to post your comment:

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