What are xG, xA and xGI expected stats?

Used a lot in modern football analysis, what exactly are the xG, xA and xGI expected stats? Knowing this can aid Fantasy Premier League (FPL) managers.

• READ MORE: A beginner’s guide – What is FPL?

After years of counting basic events like possession and goal attempts, the last decade or so has brought a demand for some context-based statistics that can more accurately measure good and bad.

xG – Expected goals

This is a way of measuring the quality of shots, rather than quantity.

It calculates the probability that a shot will be scored, compared to many previous ones with similar characteristics.

Each individual attempt comes between 0 and 1, where the distance from goal naturally plays a big part, but it’s also based on other factors, like the angle, the body part and other players being nearby. On a simpler level, penalties are rated at 0.79, due to their 79% conversion rate.

For example, in 2024/25, Evanilson bagged 10 league goals, but he accumulated a higher xG of 12.46. It can be concluded that the striker was sometimes missing some ‘easy’ chances.

xA – Expected assists

Similarly, this calculates the likelihood that a completed pass will lead to its recipient scoring.

Another occasion where each is between 0 and 1, factors include player location, pass type, pass distance and its end point.

• READ MORE: Player and team statistics of FPL 2024/25

In 2024/25, Phil Foden ended with two league assists, but racked up a higher xA of 5.34. Unlike before, this is fairly critical of teammates, rather than the midfielder doing badly.

xGI – Expected goal involvement

Meanwhile, this is the straightforward addition of xG and xA.

In a match, if a player’s three attempts add up to 1.12 xG and they also set up 0.64 xA of chances for others, that day’s xGI is 1.76.

For instance, Mohamed Salah had a phenomenal 2024/25 of 29 goals and 18 assists. But his xG (25.37) and xA (9.06) combined for a 34.43 xGI tally, the league’s highest.

How FPL managers use these numbers

In FPL, having your team collect the highest-scoring players is a reaction. A good one, too.

But to achieve true greatness, a manager needs to also try predicting the next must-have asset. Studying xG and xGI data for underachievers and overachievers is one way of approaching this.

Although it’s not always true that these numbers eventually balance themselves out – as seen in the earlier examples – there’s usually some faith that a lucky or unlucky player’s points are about to change.