Football itself is a game largely based around luck. Although skill and ability play a large part, it’s possible to achieve results in football that simply were not deserved. As football fans, I am sure many of us have found ourselves in the moments after a disappointing result moaning and groaning claiming we deserved more from the game. Expected goals (xG) is a statistical analysis method that aims to quantify what was truly deserved from the game. Football is random and unpredictable, otherwise we would have all made our fortunes from betting, xG aims to represent what truly occurred in the game and determine whether or not the performance deserved a better result or if the team was ‘lucky’ so to speak.

What is xG?

Expected goals (xG) tells us the quantity of chances and the quality of these chances that a team creates each match. For example, the game against Burnley at home where Villa drew 0-0. The narrative following this game was that Villa dominated and should have won, which they should have, but by how many? The main statistic used were the 27 shots Dean Smith’s side had. Indicating a bombardment on Burnley’s goal. However, Villa’s actual expected goals for the game were 1.81 to Burnley’s 0.32. Showing Villa should have won, but it wasn’t the domination we all interpreted from the shot’s statistic. Expected goals paints a true and unbiased picture of the way each game played out. Another useful example would be Villa’s historic win against Liverpool at home. The score line 7-2 is difficult to forget for anyone but attracting this level of publicity also attracted the comments about the goals scored and how undeserved they were. Whilst McGinn’s, Grealish’s first and Barkley’s goals may all be considered lucky due to deflections, xG states Villa had 3.08 expected goals and Liverpool had 1.66. Villa may not have deserved to win by the margin they did and some of the goals possibly were lucky, but they did outplay Liverpool on this occasion and undeniably did deserve all three points.

How is it calculated?

Expected goals analyses the likelihood of a shot resulting in a goal following the analysis of over 100,000 shots from the position the player takes a shot or similar to it. The data used is sourced from understat.com. Their algorithm also considers 10 parameters of each shot taken such as players in-between the shot and goal to determine the shot probability. For example, a shot on a player’s weak foot from 25 yards out may only have a 1% chance of going in, therefore giving it an xG value of 0.01xG. Whereas penalties due to their common characteristics will always have an xG value of 0.76xG.

Expected goals being used as a statistic is becoming more and more common, with Sky Sports using it in the pre-match analysis for the game against United on New Year’s Day. The ability to quantify the true reflection of the game is slowly becoming a more favourable measure within the modern game. 

By Patrick Rowe


Facebook and Twitter readers, click the respective icon below to follow our utvpodcast social media accounts.

Subscribe to utvpodcast.co.uk to receive instant notifications on all podcast and blog releases.

Help us grow and subscribe to our YouTube channel.