Expected Goals Analysis (xG)

What on earth is xG? I hear you ask.

xG started cropping up more frequently last season and stands for Expected Goals. Somehow analytical-data-types determined when people should score more often than not. This is how the boffins describe the factors playing in to determining xG: “Expected goals uses a whole bunch of indicators based on Opta’s on-ball event data – where on the pitch the shot had been taken from, what part of the body was used, the type of pass that had set up the chance, how quickly the move progressed down the pitch before the shot, the proximity of the opposition players, and so on – to determine exactly how likely it is that a particular opportunity will result in a goal.”

Using xG who is underperforming?

So with that in mind I decided to have a look at who didn’t score as many goals as they should have according to xG last season. Why? My thinking was that these guys are getting in the right places to score, but for whatever reason, they aren’t hitting the back of the net as much as they should do, or could do.
Looking at the Top 5 European Leagues, I have selected those players with a large difference between the goals they scored last season and how many goals they were expected to score (xG), rounding the numbers to the nearest whole number and coming up with an xG Difference. I then moved on to sofascore, they’ve got a stat on their player pages called ‘Big chances missed’. Using the selected players I had a look on sofascore to see how many big chances they’d missed, which should tie together with their xG difference and indicate who could come good this season.
Finally, I took a look at Noirx4’s data to see what dividends they returned last year and their PB High. Obviously the fact these guys didn’t score as many goals as they should’ve done means they didn’t get much in terms of PB last year but it gives you an idea if they’re also scoring points for their other contributions to the game.


So with all this in mind, I think it’s fair to say there is potential for one or more of the players below to ‘click’ and start banging the goals in. 


Christian Benteke (Crystal Palace) £0.89 
A huge 20 ‘big missed chances’! Probably part of the reason he only mustered 3 league goals last term. His xG of 11 suggests he has potential score more often and if Palace can keep hold of Zaha, he will have more chance of a decent service too. He got 42 in 89 games while at Villa but has struggled since his move to Liverpool. Is this the season he comes good again?


Karim Benzema (Real Madrid) £1.65
So many ifs, buts and maybes surround this guy and he seemingly gets linked away from the Bernebau every transfer window. With Ronaldo moving on, will Benzema come back to the fore? 5 league goals last season isn’t what you’d expect from a guy with 127 goals at Real Madrid. He has an xG Difference of -9, the highest in this data, plus you’d expect Real to go deep in the Champions League too.


Thorgan Hazard (Borussia M. Gladbach) £1.58
One of the most intriguing ones for me. He registered double figures in the league last season and reached a Max PB of 211*, the highest on this data. However, he should have had more goals with an xG of 14 and 12 big chances missed. Potential for more goals there and an all-round contribution with age on his side too.


Lorenzo Insigne (Napoli) £3.19
At the price above, he is the most expensive player on this list with a PB return to date of 19p. It may give an indicator as to where others could get to, although there are obviously many other factors playing into this. An xG Difference of -7 shows there is room for more from Insigne, playing in a very good Napoli team who are also in Europe.


Richarlison (Everton) £2.22
Potential to do more at Everton this year having had experience of English football for the first time. As a £50m signing, he will be the main man at Goodison and he will be expected to deliver more goals than his 5 last season. With xG Difference of -6, there is scope for that to happen.


If you’re chasing Performance Buzz dividends these players maybe aren’t for you. However, if you are looking for someone’s value to rise (as they inevitably do when people score regardless of dividend returns), there may be potential in some of these players detailed above, or in the data selected. 


So what about the flip side? Who is banging in more than they should?


Using the stat-tastic understat.com I grabbed a few of the highest scorers in the Top 5 European Leagues from last season and went back a few seasons so I had 3 seasons worth of data, noting down their goals and xG.
To me, in real terms, there is a simple way to look at this data. xG is expected goals remember, so someone with an xG Difference of zero, scores when they are ‘supposed’ to score. If you are less than zero, you’re not scoring as often as you ‘should’, and the higher above zero you are, the more goals your scoring than you are ‘expected’ to.
My interpretation of the last bit means, the higher above zero you are, the more worldies you’re scoring, because the xG is calculated on the difficult of the scoring opportunity.

Let’s use a working example, Lionel Messi:


His xG Difference = his actual goals – his expected goals.
Messi xG Difference = 34 – 29 (+5) (for 2017/18 season)
His Average xG Difference = his xG Difference over 3 seasons divided by 3
Messi Average xG Difference = +5, +10, -1 (14) divided by 3 = 4.67


Based on xG, Messi has scored, on average almost 5 goals more per season that he ‘should’. Unsurprising, given that he is one of the best players to ever walk the planet, but that says to me that he’s smashing in some great finishes. Risky business? Maybe not for Messi, given how good he is, but is anyone else performing above expectations who may not be able to keep it up?


Wildcards or just top class?
Icardi, Kane, Greizmann, Messi and Dybala all average 4 or more goals per season higher than their xG. They are scoring more than they are expected to, based on all those factors in the previous blog, that’s some going.
In the last blog we looked at under achieving strikers, and using the data this way you’d find Benzema (-9), Benteke (-8) and Insigne (-7). So are these players such as Icardi, Kane, Greizmann etc over achieving or are they just that good?
Maybe their luck will run out and they’ll go through a dry patch? How will that affect their price? Does that make them a more risky purchase, especially at the prices they are at now? Maybe it just wont click and they wont find the back of the net as often, or maybe they will just banging them in when – according to the data – they have no right to. That’s for you to decide.
As regular as clockwork?
Maybe a better approach is to go for those who just do the simple things, slot the chances away when they are presented, do nothing flash but do what they need to? Less exciting, but maybe less risky.
If this is your choice, you’re looking for someone with an Average xG Difference of zero. He scores when he should, rarely misses a sitter and rarely notches a worldie. A rare miss of a simple chance might be off-set by a slightly more difficult chance, but the general thought is as close to zero as possible.
Cristiano Ronaldo, Fabio Quagliarella (-0.67) and Edinson Cavani (0.67) come closest to this mark, while Neymar, Raheem Sterling (-1.00), Robert Lewandowski, Sergio Aguero and Dries Mertens (1.00) are just 1.00 either side of the magic zero.


Five focus
Robert Lewandowski (Bayern Munich) £3.44
With a goal record that reads 29, 30 and 30 over the last 3 seasons, there isn’t much wrong with his goalscoring prowess. He also does what he does with an Average xG Difference of 1.00. He’s the highest goalscorer with the closest Average xG Difference to zero. He also brought in 20p PB and 7p MB since August last year.
Mo Salah (Liverpool) £10.13
Can Mo keep up his record goal scoring record of last season? His Average xG Difference was 7 last season with 32 league goals, so that cut in from the left and curl it into the top bin must be considered a difficult goal to score, despite him doing it regularly. His past two seasons yielded 15 and 6 goals, so was this a freak season or a sign of things to come?
Fabio Quagliarella (Sampdoria) £0.75
19 league goals last season in Serie A, with an Average xG Difference of just 0.67, Quaglirarella presents the cheapest option of the closest to zero club. His previous two seasons tallied just 12 and 5 goals, so this may have been a one off. A cheap option but probably justified given he is 35 years old and there’s no European football for Sampdoria.


Dries Mertens (Napoli) £1.84
If Quagliarella presented an uninspiring choice, Dries Mertens is the next cab off the rank and may suit you more. An Average xG Difference of 1.00, his 18 goals were down on his 28 in the previous season but he scores when he should and at £1.84 comes with the bonus of European football next season.
Radamel Falcao (Monaco) £1.42
So long as his knee doesn’t explode again and he doesn’t come back to the Premier League, Falcao may present good value at Monaco. Since being back in Ligue 1 he’s netted 18 and 21 league goals respectively. His Average xG Difference of 3.00 suggests he nets the odd worldie too, while being consistent with his scoring…in France.


*Divs and PB stats based on IndexGain and Noirx4 data (Noirx4 does not include a full season).
All data was correct at the time of writing.
Share This Post
Written by Nick
A football index user for over 18 months now, I've turned my hand to something I love about football, the tactical battle. Creating a mash of Football Index related info and stats from IndexGain with what's happening tactically, hopefully gives an insight into another way to play the Index.
Have your say!
2 0
NEW: Introducing Spread Alerts & Notifications on Slack. Find out more here.
Holler Box