Seven unexpected things xG data can tell us about the Premier League season

Eleven games into the season and we're all jumping to conclusions. Crystal Palace are being written off, the title has been decided and is heading to Manchester, while Everton's transfer strategy is being declared fatally flawed. But could it be that there is more to this season so far than meets the eye.

Here, using Opta's expected goals metric, we have attempted to uncover some of the things we might have otherwise missed by delving a little deeper.

Expected goals (xG) is a system that gives a probability that any shot will be scored. That is, based on the quality of the chance - be it headed, right-footed, volleyed, a sitter - each shot will be assigned a value between 0 and 1 that shows how often it should be scored. 

Using xG we can more accurately assess a player or team's finishing ability, and also ascertain whether a team is playing better than their record suggests. That is, we can see if they are creating better chances than their goal record states and if, in fact, we should expect a team to improve and their luck to change. (If you want more of an explanation on xG, see here.)

Here's what we found:

Crystal Palace's attack is better than Chelsea's

Okay, this isn't quite accurate, but let me explain.

Crystal Palace, bottom of the league with barely a hope in hell of staying up, have struggled terribly in front of goal this season. Even before Christian Benteke sustained the injury that has kept him out for much of the season, Palace were missing too many chances that really should have been put away.

But maybe we should focus on the fact that they have been creating lots of good quality chances rather than their failure to finish them off. Crystal Palace may only have four goals, but they have created chances worthy of 13.81 expected goals. Chelsea, meanwhile, have scored 19 but should only have been expected to score 12.71.

When Benteke returns (to form as well as fitness), Palace should begin their assault on safety. The only worry is that that will come too late.

Burnley's form in front of goal might not be sustainable

Sean Dyche hardly has a set of strikers that other teams would crave, but only two teams - Manchester City and Chelsea - have finished chances more clinically than Burnley so far this season.

City have scored 10 more goals than expected; Chelsea have scored six more. But both boast far greater riches than Burnley, who have an expected goal tally of just 6.53 but have found the net on 10 occasions.

It isn't exactly world-beating form, but they are unlikely to keep this up for the entire season.

Liverpool and Arsenal perform exactly as expected

Fans of both Liverpool and Arsenal often feel they have a lot to complain about, usually with regards to refereeing decisions going against them or players that are not playing at their potential.

But expected goals metric in fact suggests that both teams are scoring as many chances as they should. Liverpool have scored 21 goals, but have 21.4 expected goals; Arsenal have scored 20 goals, but have 19.72 expected goals.

What this suggests is that both teams' attacks have consistent, yet rather average, finishing. When a shot should be scored, it is, but when difficult chances present themselves, they are often missed.

Southampton need a new striker

Manolo Gabbiadini settled into life in the Premier League faster than most, but has struggled to make much of an impact this season.

Southampton have scored nine goals this season, but have created enough good chances to score more than 13 - the second-worst such difference after Crystal Palace. A better striker would certainly help matters.

Burnley might have the Premier League's best goalkeepers

Nick Pope isn't a player many fans had heard of before this season, but he is proving a more than capable deputy for a more than capable Premier League goalkeeper in Tom Heaton.

Heaton has missed much of the campaign with a dislocated shoulder, and Pope has maintained the best save success rate (88.1 per cent) in his team-mate's absence.

Indeed, between them, the keepers have kept Burnley's goal against column down to nine, when they have allowed chances worthy of shipping more than 16. 

It may be that Lady Luck has been smiling on Pope and Heaton with them coming up against wasteful finishers week on week, and that might not last the season, but the positive difference here is the best in the whole Premier League, and certainly suggests they are doing something right.

City are being out-performed defensively by Bournemouth, Brighton and Newcastle

Manchester City's attack is rightly being lauded as one of the best this country has ever seen, but their defence needs more work than empirical evidence might suggest.

City are conceding more goals than their performances would suggest they should be - and the same cannot be said of Bournemouth, Brighton or Newcastle, all of whom have spent rather less on reinforcing the protection of their goal. Those three teams are all conceding fewer goals than Opta's data suggests we should expect of them.

Pep Guardiola is getting an awful lot right, but there is room for improvement at the back.

England's best goalkeepers are the Premier League's worst GKs

According to the 'expected goals against' data, West Ham's Joe Hart, Stoke's Jack Butland and Everton's Jordan Pickford should have conceded around 51 goals this season. They have in fact conceded 67.

None have been aided by the defences in front of them, but each of the three goalkeepers in Gareth Southgate's England squad last week has been under-performing woefully this season.

Of the 24 goalkeepers to feature in the Premier League this season, Hart, Butland and Pickford have the third-worst, second-worst and seventh-worst save percentages.

Hart is constantly battling doubters, but consistently retains the support of every England manager that comes and goes, while Butland and Pickford are praised as players with bright futures ahead. Both must improve significantly.