The purpose of xG, or Expected Goals, is to showcase the scoreline that a team deserves. While I value its role and usage in the modern game, I can’t help but feel that it has not been used to its full potential. This was brought to my attention by my own father, who has been slightly reluctant to fully invest himself in analytics.

Our conversation went along these lines: “Are goals scored from low xG shots undervalued?”

I initially scoffed at the idea. Goals scored from these scenarios are not expected, so they shouldn’t be valued more or less compared to other shots, no matter the outcome. But, after some time meditating on the topic, I believe that this perception has some validity behind it.

xG can be looked like a percentage If a shot has 0.4 xG, you expect 40% of shots from that position/speed to be scored. But what would happen if a player consistently over performs the xG in that position? There aren’t many who can, but some notable examples can be found each season, although it is hard to find multiple cases of that lasting multiple years! (One of the exceptions is Messi…)

My point is that players can efficiently overperform their xG over a span of games. I feel like this phenomenon is sometimes thrown under the towel and can be under-represented in an analysis of a football match.

This slight issue can be fixed by adding a team’s xG and their goals scored together and then subtracting their xG on shots that entered the back of the net. This calculation is not a replacement of xG; it should be used as a complementary tool. If you win a game by a final score of 2-1 but have an xG of 0.7 while the other side has a 1.4 xG, people might consider you lucky and the other team unlucky, especially if they haven’t seen the game.

Now let’s use my calculation. Let’s assume both of your goals had a 0.06 and 0.2 xG respectively. Your entire xG for those two goals is 0.26, which is not much. The other team’s goal came from a 0.82 xG opportunity. What happens when we replace these numbers while also adding in the rest of the xG from the match? You would get a scoreline of 2.44-1.58, which resembles the actual scoreline.

Using this combination of the actual scoreline, xG, and this other equation, you can start to imagine how this game played out. For your team, you created some chances, but the difference came from two outstanding goals. Realistically, a third goal was not out of the realm of possibilities. For the other team, there was some unluckiness involved, but their players weren’t as clinical as yours.

What are some real-world applications of this calculation? The first situation which I could find use for this would be for the fans. One could easily find out the maximum number of goals you could have scored. If you scored six goals on low xG, you most likely shouldn’t have won by that margin, but your players were clinical that day, so they were deserving of the result. On the contrary, if you scored one goal with a low xG shot but then piled on 2.56 xG which didn’t result in goals, it is safe to say that your team could have realistically scored three to four goals that match at maximum. Expected Goals would have put your team in a lower range.

Do I think this thought experiment will become the norm? Probably not. I’m not the most analytical person in football, and maybe someone already came up with this idea, and I just haven’t seen it. What I do know is that I’ll be using this calculation to see how many goals a team could have possibly scored in a match, and I’ll keep using xG to see how many goals a team actually deserved to score.