If there was one thing that was completely different last season in Fantasy Premier League, it was the rise of conversations and arguments around the concept of xG – Expected Goals. And while xG has been a remarkably consistent statistic in the sense that it has, to a large degree been accurate in terms of absolute goals, critics continue to slander the statistic, often citing the ‘eye-test’ as we call it in FPL lingua franca.

What is xG and Why is FPL Obsessed with it?

The answer is that xG simply pertains to the chance of a goal being scored. The easier the shot, the higher the xG and vice versa. The second question that crops up is the metric used to ascertain what is easy and difficult, and that can be ascertained broadly from:

  1. Distance from goal (Farther the goal, lower the xG)
  2. The angle from which shot is hit (Wider the angle, lower the xG
  3. Body part used to hit the ball (Headers leading to a lower xG)
  4. The number of defenders acting as an obstruction (Higher the number, lower the xG)

In an ideal world, xG would also consider the calibre and the quality of a player but at the moment that seems like a far cry.

One of the main reasons why xG is of utmost relevance and importance to the game is FPL’s tendency to have a points mechanism that is heavily tilted towards attack – especially goals scored. If managers focus and view the metric over a period of 4-5 Gameweeks, there is a lot to discover both on the team side and individual players.

Over and underperforming players

A quick look at the top 10 goal scorers of the 2018-19 season, brings out a reasonably good accuracy of 80% within a deviation range of ± 2. The key outliers who have overperformed their xG massively are Eden Hazard and Sadio Mane.

A more detailed look at the numbers can be seen in the visualization below:

What makes the two of them stand out from the rest of the pack? Some players do not conform to the metric because they take fewer shots, which are often “tougher” by xG standards, especially in the case of Eden Hazard.

The Belgian is also a good finisher, something xG data does not consider. To give you some perspective, Messi had an xG of 26 but scored 36 goals (truly a freak of nature).

 

Source: Understat

 

Key takeaways

Here is my five-point mantra to effectively understand and use xG in your favour and build a strong fantasy team using the stat:

  1. Do not view xG in isolation and always take the figures over a duration of at least five Gameweeks. Viewing xG numbers for a single Gameweek is bound to give you incorrect insight that might end up frustrating you. Also, never view the number of fixtures but also see the difficulty of these fixtures as they could tilt the number one way or the other.
  2. Check up on the teams xG as well, a higher figure reveals that the team has a tendency to create a lot of chances which directly benefits your chosen player. Once you have the teams xG, look how much of the total is contributed by the player that is on your radar. The higher the better as it shows high dependency on the player, which translates to him being the most probable goal scorer.
  3. Look at other data. In particular the number of shots that the player takes per match/per 90 minutes. A large number of shots taken will only pad up the xG and inflate the figure. As a thumb rule, anything more than 5 shots/90 minutes is effectively not showing you the true goal-scoring ability of the player.
  4. Do not use xG as a measure on deciding who will win the game. Football is a game of chances, and if the last season is anything to go by, results may not always show the true nature of the match going on.
  5. Continue to trust the ‘eye-test’. Do not solely use the xG numbers or the ‘eye-test’ in isolation. Instead, combine them for the best insights. If that is not possible, use the abundant resources online whether in terms of match summaries, FPL subreddit, twitter accounts to get an idea of how the match transpired.

I have prepared a scatter plot that immediately gives you an idea of how the top 25 goal scorers fared in the 18-19 season. Preferably the top right quadrant is where you want your player to be, followed by top left, bottom right and at the very end, bottom left.

Happy analysing and best of luck to everyone for the next season.

15% off hub membership

If you’re not already, you can join thousands of fantasy managers and access this and all our membership content for less than £1.50 per month. Simply enter coupon code: REBANT at checkout for any of our yearly plans You’ll get:

  • OPTA data (NEW).
  • How to Win at FPL Guides (NEW).
  • Ultimate pre-season guide (NEW).
  • Entry into £3,000 member’s prize league (NEW)
  • Slack channel.
  • Ben Crellin’s FPL planning sheet.
  • How To Win at FPL guides.
  • Player comparison tool.
  • Rate my team tool.
  • Advanced fixture ticker.
  • FPL Heisenberg team reveals
  • FPL watchlist tool
  • Key FPL stats tool.
  • Minimum 50 extra articles each year.
  • A forecasting model for FPL.
  • A lot more!