Premier League xG Stats Review: Reds hot, Foxes not

HOT trends?

Home comforts for Gunners

Much has been made of Arsenal’s weak away form over the past decade, even during the final few years of Arsene Wenger’s time in charge, and while they went to Wolves and won in their last league game, the Gunners remain a below average travelling team.

The average expected goal difference (xGD) per game when teams play away in the league is -0.34. Arsenal’s is -0.51 this term.

Fortunately, then, Mikel Arteta’s side have made the Emirates a fortress, winning seven of 11 games at home this season.

Only Liverpool (2.43) and Manchester City (2.38) have averaged more xPoints per home game than Arsenal (1.96) this season, with the Gunners boasting an xG process of

2.00 xGF and 1.04 xGA per game.

Their only defeats at home have come against Chelsea and Manchester City, and coincidentally, those are the only two home games in which Arsenal have lost the xG battle.

Brentford beware, as the Emirates is now a scary place to visit.

Record highs for Liverpool

Liverpool’s strength is undoubtedly in their attack.

The Terrific Trio of Mohamed Salah, Sadio Mane and Roberto Firmino became the Fantastic Four with the addition of Diogo Jota – who is the second top scorer in the Premier League this season.

After January though, we can add another body to that list to make it a Fabulous Five, with Luis Diaz a legitimate star.

The fact the Reds now have star quality to ‘back-up’ the original triumvirate should stand them in good stead to challenge on all fronts, as does the return of key players from injury.

But the reason Klopp’s men feature in this column is due to the attacking numbers they are putting up this season.

Liverpool are averaging a mind-boggling 2.68 xGF per game, which currently has them on pace to become the best attacking team Infogol has ever seen, based on xG.

Since 2014, when Infogol started collecting data, Manchester City’s 19/20 team hold the record for the best attacking season, averaging 2.67 xGF per game. If the Reds continue in the same vein of form, they are set to edge ahead in that ranking.

The same goes for xGF per game in home matches. The Reds are running at an extraordinary 2.86 xGF per home game this term, which if maintained, would eclipse another Manchester City team of the past decade (18/19 – 2.81).

That all spells real trouble for relegation candidates Norwich, who after taking a beating at home to City last weekend – allowing 4.55 xGA – must visit Anfield to take on one of the best Premier League attacking teams xG models have ever seen.

Can we play at home every week?

As I mentioned on last weeks Football…Only Bettor, Crystal Palace are one of the more extreme teams this season in terms of home v away processes.

The Eagles boast the fifth best xP per home game (1.90) as well as the fifth best xG process (1.87 xGF, 0.98 xGA per home game) in the entire Premier League.

On their travels they sit 14th in both categories.

Their xGD per game stands at +0.89 in front of their own fans, a really impressive figure that suggests that, on average, they create two and a half more big chances (0.35 xG+) than their opponents in every home game.

Away from home their xGD per game is -0.74. Such a stark contrast is strange, with the Vieira’s side playing like a top four team at Selhurst and a relegation candidate on their travels.

Fortunately for the Eagles, they are at home this weekend, albeit against one of the league’s better sides in Chelsea. If they hit the same levels as they have done at home this season, they can cause problems for the Blues.

Sá-nsational

Wolves’ win over Tottenham last weekend moved them up to seventh and just six points behind fourth placed Manchester United with two games in hand. All that despite sitting just 11th based on xP/game.

They have overperformed by a wide margin defensively, conceding just 17 goals from chances equating to 31.1 xGA – allowing more than 14 fewer than would be expected.

Goalkeeper José Sá has been a major contributor to that.

Based on post shot xG (psxG), which is a measurement of the quality of shots a goalkeeper faces, taking into account power, placement, swerve, deflections etc, Sá has saved over seven goals more than would be expected (+7.4).

According to the on target shots he has face this season, the Portuguese keeper would have been expected to concede over 24 goals (24.4 psxG).

Only David de Gea (+9.5) has performed better in the league this term, and if Sá continues in the same vein of form, he could be the difference between Wolves finishing in a European place or not.

COLD trends?

Hodgson’s Hornets ineffective

To make the best assumptions using your expected goals model, you ideally need a decent sample size.

With Roy Hodgson at Watford, we unfortunately don’t have that just yet, but what we have seen so far looks very similar to what we saw from him in his last few seasons at Crystal Palace.

The Hornets have lost two and drawn one (with bottom side Burnley) under Hodgson’s guidance, failing to score in all three.

Their xG process (small sample alert) since his arrival reads
0.63 xGF and 1.31 xGA per game.

Ultimately, the early signs are that he has tightened them up at the back – with Watford allowing 1.85 xGA per game in the first 20 games of the campaign.

But, their attacking process has more than halved (small sample alert), averaging 1.30 xGF per game prior to his arrival.

It doesn’t look good for Watford then, especially when we factor in Crystal Palace’s underlying process in Hodgson’s final two season’s in charge.

In 19/20 they were running at 1.03 xGF and 1.65 xGA per game. The season after, things got worst still – 1.02 xGF and 1.77 xGA per game.

The xGD per game (-0.62 & -0.75) in both season’s is very similar to the early signs in his tenure at Watford (-0.68), so perhaps it wasn’t a great appointment to take on a manager who had been fortunate to finish 14th in back-to-back season’s?

All of this is working with a small sample of games, but you get the working theory. A visit to Aston Villa will be the latest test for Hodgson’s men.

Away day Blues

Newly crowned Club World Cup champions Chelsea return with a potentially tricky away game at Crystal Palace (see above if you don’t believe me), but Thomas Tuchel’s side haven’t been the best travellers in the Premier League of late.

They have won just two of their last six away games in the league, winning at Watford and Aston Villa while losing to West Ham and Manchester City (draws with Wolves and Brighton).

Across those six away games, Chelsea have averaged just 1.17 xGF and 0.88 xGA per game, so it is no surprise that they have struggled to get wins.

That attacking process isn’t going to help them win a Premier League title, and that is the main area of concern.

Not so cunning Foxes

Many are surprised to see how badly Leicester are struggling this season, with the Foxes languishing in 11th place in the league, some 16 points off a top four spot and 12 off the top six.

Brendan Rodgers’s side have major issues, mainly defensively, with the Foxes on track to be one of, if not the, worst defensive teams Infogol has ever seen.

The KP outfit are allowing a whopping 2.04 xGA per game this season, an incredible increase on last season’s figures (1.36).

Currently, the worst defensive team Infogol has on record is Aston Villa in 19/20, with them also allowing 2.04 xGA per game.

The only reason Leicester aren’t embroiled in a relegation battle is because of their attack, which has also declined on last season, but is still that of a mid to top half team.

Last season they averaged 1.63 xGF per game, this season it’s down to 1.43.

So, in total, Leicester are on average 0.88 xG worst per game than they were last season. That is some fall from grace.

The bigger picture of course is that the Foxes are Premier League constants, something they couldn’t say a decade ago.

A trip to Molineux at the weekend presents them with a tricky task, especially if they continue to defend the way they have.

Set-pieces could be an angle for a bet against Wolves, with around 27% of the xGA Leicester allow coming from dead-ball situations, while the Foxes have conceded eight goals in their last 14 games to centre backs alone…

Conor Coady (two goals already this season) at 12/1 to score anytime anybody, for a team who generate 29% of their xGF from set-pieces?

You can download the FREE Infogol app on the IOS and Android stores.