Calculating expected goals

The expected goals (xG) formula is a soccer statistic that shows how many goals a team or striker is meant to score. If your child is into soccer, ask them to predict who'll win the premiership league and see if they can justify their prediction using this simple mathematical model. 

10 minutes

7-8

27 November 2020

Things you need

  • Device with internet access
  • Notepad
  • Calculator
 
 
 

The challenge

Is your teenager a fan of Messi, Ronaldo, Mo Salah or Harry Kane? If they’re getting up early to watch European games, it’s time to put that obsession to good use. They can calculate their favourite player's expected goal (xG) value for their next match by using this simple mathematical model. 

When all shots are treated equally, the expected number of goals is about 10% of the player’s attempted shots at the goal in their previous match. This can be represented using the formula: 

xG = 0.1 x shots

So, if a player attempts 10 shots at the goal in their previous match, it is expected that they'll score 1 goal in the next match. This is because:

xG = 0.1 x 10

xG = 1

 
 
 

The conversation

The xG formula is a simple model that you can use to fuel some good discussion during the game.

  • "Did the formula accurately predict the number of goals scored, or who the goal scorers would be?"
  • "Could you use the formula to compare the two teams to predict who is more likely to win?"
  • "If a player took 15 shots, they’d get xG = 1.5. This means they’d be expected to get 1 or 2 goals. What would you expect if a player took 5 shots? What about if a team took 25 shots?"
  • "The xG model can change when it takes other factors into account. What are some other factors that might help build a more accurate formula?" (Hint: distance and angle to the goal.)
  • "Do you think the xG model will always be able to predict the number of goals scored? What are its limitations?"
 
 
 

The career

While technology and artificial intelligence can calculate statistics faster than humans, there’s still a need for humans to interpret it. This is especially true when there are a high number of variables, such as in a football match. Any kind of career in sport - from sports scientists, coaches and talent scouts to sports journalists and the professional athletes themselves - will come with a need to interpret data. In fact, pretty much any business, from running a cafe through to managing a mining facility, requires you to use data effectively to make critical business decisions on a daily basis.