Editor’s Note: This is a featured post from the fine folks at Infogol, with the underlying message of positivity, as Everton have solid underlying numbers and should be much higher in the table.
We hope to continue this working relationship with Infogol as the season continues, and thank Jake & Glenn for their contribution.
Everything is doom and gloom around Goodison Park heading into the international break, with Marco Silva the favourite to be the next Premier League manager to leave his post following Everton’s slow start to the season.
Everton sit in the bottom three after eight matches, with just two wins to their name, both of which have come at home, but there is growing unrest among fans who feel like the team should be doing much, much better.
The Toffees have splashed the cash in recent windows and look to have the squad to compete in the race for a top-six place, but so far have fallen well short.
So, they stick with Silva? And is there reason to be upbeat?
Infogol Analyst Jake Osgathorpe uses expected goals to answer these questions.
What is Expected Goals (xG)?
Infogol is a revolutionary new football product, which harnesses Opta data to power an Expected Goals (xG) model.
Expected Goals quantifies the quality of a scoring opportunity, allocating each chance a probability of it finding the back of the net.
The xG metric can be used to assess teams and their performances, and it also helps to give an insight into future prospects.
Tricky questions such as whether a team is over-performing or under-performing, or whether a side are likely to maintain their current form, can be answered with xG.
Not all doom and gloom
Everton’s defeat at Burnley last weekend was their fourth defeat on the spin in the Premier League, following defeats against Bournemouth, Sheffield United and Manchester City.
That result left Everton in the relegation zone.
However, while Everton have been really poor on the face of it, according to expected goals the Toffees rate as the fifth best team in the Premier League after eight games.
It is fair to say that Silva’s side have been very unfortunate in matches, with a case in point being their defeat at Burnley at the weekend (xG: BUR 0.54 – 1.48 EVE), where they created the better chances and limited Burnley to low-probability efforts, but lost 1-0.
Infogol re-simulated the game using chances created by both sides, and calculated that Everton would win that game 60% of the time, and lose only 13% of the matches.
It was the same story against Sheffield United (xG: EVE 1.50 – 0.80 SHU) as the Toffees limited the Blades to just two shots, and had a 53% chance of winning the game compared to United’s 20%.
Against Aston Villa it was the same (xG: AST 0.81 – 1.56 EVE) – spotting the pattern yet?
Everton have averaged 1.55 Expected Goals For (xGF) and 1.32 Expected Goals Against (xGA) per game this season, a steady process, and one that we would expect from a top-half team. They may be sitting in the bottom three but their performances have been far removed from that of a relegation-threatened team.
Sooner or later Everton’s results will start reflecting their performances, and if they maintain their current level they will start climbing the table.
Non-penalty Big Chances created
Silva’s side rank joint-third in terms of non-penalty big chances (35% or greater) created this season, racking up 13, placing them behind only Chelsea (14) and Manchester City (24). They have created three more than rivals Liverpool (10) and seven more than Manchester United (6).
Keep doing that and the goals will come.
Need to be more Clinical
The main issue for Everton has been converting their chances. So far this season they have netted just six times, but, based on the quality of chances they have created, they should have scored closer to 12 (12.46 xGF).
Only Watford (4 goals, 11.67 xGF) have underperformed more in attack, and it appears this is the reason why Everton are where they are in the table.
Player assessment
So far this season, according to expected goals, Gylfi Sigurðsson has been Everton’s standout player.
He has contributed 3.70 non-pen Expected Goal Involvement (xGI), the most in Everton’s team, and has the highest non-pen xGI/90 of any Toffees player (0.49), so has to be one of the first names on the team sheet.
Richarlison (3.24 non-pen xGI) and Dominic Calvert-Lewin (2.87 non-pen xGI) follow, and have both impressed based on underlying numbers, but Everton fans will still be wanting more.
Everton xGI leaders 19/20
Full-backs Lucas Digne and Seamus Coleman have offered a decent attacking threat again this season, and will continue to be major outlets for the Toffees.
The early indications from new signings Alex Iwobi and Moise Kean have been positive, and, although they have had limited game time, what we have seen so far suggests both should be in the starting XI - Iwobi is averaging 0.36 non-pen xGI/90, while Kean is averaging 0.48 non-pen xGI/90.
The major disappointment this season has been Bernard, with the Brazilian playmaker not doing too much playmaking, averaging a measly 0.12 non-pen xGI per 90, so he should perhaps be left on the sidelines for a game or two.
Stick with Silva
There is no question that Silva should be doing better in terms of results, given the money spent and the players he has at his disposal.
He perhaps isn’t quite getting the best out of his players, and could maybe do with a change in shape or system to freshen things up a bit, but the bottom line is that the underlying numbers of his team have actually been impressive.
If they maintain the same level of process for the rest of the season they will likely finish in the top half at the worst - it is just a matter of time before the results fall in line with their xG performances.
Whether Silva gets the chance to see his project through is the big question.
Games against West Ham and Brighton are next up after the international break, and they could be make or break for Silva. Although the process has been good, a few more bad results would likely spell the end for the Portuguese coach.