Everton supporters endured a brutal hour and a half last week as the Toffees drew 0-0 away for the second time this season. However, the Blues have been scintillating at home in 2017, scoring 29 times in eight games. Their last six fixtures against Chelsea have produced an average of 4.5 goals per game, so there could be some fireworks at Goodison Park on Sunday.
Chelsea sit atop the Premier League table and have about an 88% chance of becoming champions. They won 13 straight matches over the course of October, November, and December, and have never really looked back.
At first glance they appear to have both an elite attack and defense. Only Tottenham Hotspur rival them in terms of goals scored per game, and they are third in goals conceded per game. If we dig into expected goals, we see that the narrative holds up on the defensive side but not so much on the attacking side.
Chelsea’s xG conceded is indeed quite good; 0.76 per game (according to Michael Caley’s model) means that only Manchester United have been stingier. That 13-game winning streak I mentioned included 10 clean sheets.
On the front end, though, it turns out that Chelsea are just 6th in terms of xG, just ahead of Everton’s. However, they are outperforming their xG at a ratio higher than any other team in the league. In layman’s terms, Chelsea are taking their chances like no one else. When you do that while keeping things tight at your end, good things happen. In this case, it looks to be delivering them the title.
For the first game of Chelsea’s unbeaten run, Antonio Conte switched the side from a 4-1-4-1 to a 3-4-3. He has stuck with a 3-man back line for every game since. Arguably it’s morphed a bit to what you might call a 3-4-2-1, but either way it looks something like this:
The fullbacks provide nearly all of the width, N’Golo Kanté is the key cog in the midfield, and Eden Hazard is the focal point of the attack.
Chelsea are somewhat different from other top sides in that they are not as possession-oriented. Their possession and passing accuracy numbers rank them 7th and 4th in the league, respectively. Part of this is because they are slightly more vertically-minded with the ball. Note the general lack of emphasis on strong lateral links in the passing map above; this has been a theme all season.
Interestingly, Chelsea’s passing accuracy jumps to #1 in the league once they move to the final third. In other words, they are clinical both with their passing and with their finishing in crucial areas. A huge part of this is simply having a player of Hazard’s caliber as your driving force, but Diego Costa is excellent at keeping the ball as well.
In summary, Chelsea are a well-drilled side that can beat you with pace and power at the same time. They tend not to dally on the ball and this season they have been clinical in front of the goal. At the moment they look set for the title.
Matching up with Everton
In my preview for the reverse fixture in November, I wrote the following:
Everton faced a similar dilemma a few weeks ago against Manchester City. They were away from home, facing a top team in a 3-4-3. Ronald Koeman’s answer was to go conservative, fielding a defensive midfield trio of Idrissa Gueye, Gareth Barry, and Tom Cleverley behind a counter-attacking trio of Gerard Deulofeu, Yannick Bolasie, and Romelu Lukaku. It...sort of worked? They drew the game, and while City definitely had more chances (and two saved penalties), only Southampton and Tottenham have done that well, holding them to just 1.1 expected goals.
Thing is, Chelsea’s 3-4-3 is a bit different than City’s...essentially City play with wingers and Chelsea play with more defensively-minded wing backs who will sit back and help combine with the central midfielders. Combine that with the fact that Gueye is suspended for this match, and you’ve got a recipe for absolute midfield domination. I’m not saying that Koeman should do anything drastically different here than what he did against City, just that he needs to worry extra hard now about being overrun in the midfield. Gueye’s ability to be the water-carrier will be sorely missed, and any poor touches in central areas will be punished mercilessly by Chelsea...
I’m sorry to bring up bad memories, but Everton went on to lose 5-0. It was far and away their worst match of the season. I was only partially right in my prediction; Everton got overrun not just in the midfield but everywhere. Notably, their fullbacks got sucked too far forward and Chelsea were able to overload the channels and isolate Everton’s center backs. For their part, the Toffees had 1 shot, which was not on target. Expected goals? Let’s ask Dean Wormer:
Not encouragingly, here’s what Everton came up with against West Ham United last weekend:
xG map for West Ham - Everton. Ugh. pic.twitter.com/axRR8iSmND— Caley Graphics (@Caley_graphics) April 23, 2017
Okay so Everton’s attack seems absent and they’re facing one of the best defensive sides in the league. If you’re looking for optimism, it’s in the form of the fact that Chelsea haven’t kept a clean sheet in the league in 3 months—a span of 11 matches. They’ve looked vulnerable at times.
The key for opponents has been isolating Chelsea’s back 3. Two things need to happen to get into that situation. One is that the wing backs, Marcos Alonso and Victor Moses, need to push forward and leave space in behind, much like Seamus Coleman and Bryan Oviedo did in the first matchup. Secondly, you need to get past Kanté. Unfortunately Kanté is really really good at not letting this happen. But he’s not perfect.
On the second chance, notice Chelsea’s positioning when Christian Benteke gets the ball. Cahill and Kanté in particular are away behind the play. David Luiz gets a little unlucky here but even so he doesn’t acquit himself particularly well. In the last clip, Kanté is again out of the picture as some good hold up play from Benteke allows for a 3-on-3 against Chelsea’s defensive trio. The result is a great chance for Palace.
Marcus Rashford’s goal from United ’s win a few days ago was another good example:
Again, Kanté is caught upfield and no one is protecting Chelsea’s back 3. A killer through ball puts the young striker through.
It’s worth noting that José Mourinho’s side did a great job at creating these sorts of situations all game, to good effect:
Everton aren’t the kind of team that can execute a really well-timed and well-coordinated press, so their best bet is probably to do something like what Palace did. In attack this means remaining conservative and going a bit more direct. The worry is that Romelu Lukaku becomes isolated, so it will be up to his supporting cast to make sure there is a second pass available following the long ball. Given Everton’s lack of true wingers, starting Ademola Lookman could be a shrewd move, with a mind towards exploiting the wide areas vacated by Chelsea’s wing backs.
In defense, the gameplan is less clear but Everton will probably want to clog the middle and force Chelsea’s attackers wide. Palace did this really well but still conceded a bunch of chances. In any case, avoiding the sorts of errors that allow Chelsea to gain possession in dangerous areas will be crucial.
With Arsenal looking like they are (barely) righting the ship, Everton need to keep pushing for points. The briefly-opened window for 6th place appears to be shutting, and a draw here would be a great result.