Chelsea finally fielded their (probable) first choice wingers in the 2-1 victory over Sunderland, but despite the Blues being in total control of the match until after ninety minutes had passed, the win at the Stadium of Light still raised plenty more questions than it answered. Still, a win's a win, right?
Although there was some question about the shape of the Chelsea midfield without John Obi Mikel, Andre Villas-Boas stuck to his guns and ran the Blues out in what was more or less their usual 4-3-3. The defensive line was as expected, with Ashley Cole and Jose Bosingwa flanking John Terry and Branislav Ivanovic, and so too were the wide forwards, where Juan Mata and Danny Sturridge both received their first Premier League starts.
The centre was where the tweaks were made - Raul Meireles came is as the holding midfielder due to Mikel's international travels and Nicolas Anelka got the start ahead of Fernando Torres as the centre forward. These two changes combined to force both Frank Lampard and Ramires to sit far deeper than they usually do - Meireles is not a natural defensive player and needs more protection than does Mikel, and Anelka was making a habit of dropping back into a false nine role, which cut off some of the space in which the forward pivot would normally operate. The two experiments met with limited success - we'll look at that in more depth a little later.
Sunderland were set up in a rather confused 4-5-1. It wasn't particularly clear who was being played where, but eventually a loose 4-1-4-1 crystalised out of the mess, which Lee Cattermole holding behind Stephane Sessegnon and Craig Gardner, who were flanked by Jack Colback and Sebastian Larsson. On-loan Arsenal striker Nicklas Bendtner led the line.
Steve Bruce altered his team's shape with two substitutions late on in the match, introducing Connor Wickham as a second striker for the ineffective Lee Cattermole to make a 4-4-2 in the 66th minute and then handing a debut to South Korean forward Ji Dong-Won for Stephane Sessegnon with ten minutes to go, leaving the Black Cats in a lopsided 4-3-3 with Ji as a left winger. Both changes served to stretch a hitherto unchallenged Chelsea defence, although it was still a surprise when Sunderland did manage to beat Petr Cech as injury time began.
Chelsea's first formation change was more subtle. Juan Mata was removed for Fernando Torres, with Anelka pushed out to the right side. Following Anelka's removal from the centre forward position, Lampard and Ramires were able to get further up the pitch and were a little more influential in the middle, although Torres' terrible play rather reduced Chelsea's efficacy after he came on.
Oriol Romeu's introduction marked Chelsea moving to a 4-2-3-1 with Florent Malouda and Ramires as wide players. It's easy to forget just how well-suited Ramires is to playing on the flanks, but he looked totally at home out there (unlike Torres). The shape worked perfectly well anyway, giving Meireles and Romeu a solid foundation from which to hold possession.
Anyway, although Sunderland's formation change(s) allowed them to put more pressure on Chelsea in the late stages of the match, it would be difficult to claim that the game was decided based entirely on shape (even more so than normal, I mean). The difference between the teams lay elsewhere.
Shutting Sunderland Down
Figure 2: Sunderland vs. Chelsea team/individual passing, 9/10/2011. Powered by Tableau.
Chelsea were comfortable throughout the match, but it's important to note that this was as much a function of Sunderland playing poorly as Chelsea playing well. After you've noted that, you can also go ahead and note that when one team plays badly, their opposition has usually had a hand in it. So it was here.
As a team, Sunderland don't look too bad, but their averages are being held up by Kieran RIchardson and Jack Colback on the left flank. The centre and right side were both miserable. Sunderland's attacking threat, such as it is, currently requires Nicklas Bendtner and Craig Gardner to score goals, and Stephane Sessegnonto create them. With Sebastian Larsson in Ashley Cole's pocket all evening, Bendtner was constantly drifting to the right to relieve his teammate, while Gardner was expertly marshalled by Lampard and Meireles in the midfield triangle. Until Bruce added Wickham and Ji at the end of the match, there was simply nobody capable of putting the ball in the back of the net for the hosts. Small wonder Chelsea were so comfortable.
Things could have been very different had Sunderland had a more threatening player than Jack Colback on the left wing or paired Bendtner with another striker (should Wickham have started?). Colback had an excellent game, of course - he was probably Sunderland's best player outside of West Brown - but he was never any sort of threat to do more than retain possession in the midfield. A true left winger would have given Chelsea problems here, one suspects, although Jose Bosingwa would probably have seen much more of the ball had Sunderland tried to push forward on the left with any real seriousness.
One of the ways Chelsea kept Sunderland quiet was with a very high press, with the midfielders and the defenders both working extremely hard to keep their hosts making passes through the middle third of the pitch:
Figure 3: Chelsea at Sunderland location data, 9/10/2011. Powered by Tableau.
Chelsea had as many interceptions in Sunderlands half (seven) as Sunderland did. Meireles and Lampard each did an excellent job of closing down their opposite number, and the interceptions totals don't account for the number of times the Chelsea midfield harried their opponents into sending the ball into touch. The five midfielders that started for the Black Cats on Saturday - including Cattermole and Colback who weren't under much pressure at all, managed a 69% pass completion percentage, lower than every Chelsea player save Petr Cech. Ouch.
Raul Meireles vs. John Obi Mikel
One of the surprised Andre Villas-Boas sprung on the world with his lineups was the inclusion of Raul Meireles as the holding midfielder at the base of Chelsea's central triangle, with John Obi Mikel the odd man out. Thanks to his performances with Liverpool, Meireles has been labelled as a player who's miscast as anything but an attacking midfielder - his play with Porto and the Portuguese national team put that conclusion in question, however, and Meireles was actually used as the holding player for Portugal in their 1-0 win over Norway this sumemr for European qualification, playing reasonably well and showing superb vision in pinging the ball to his fullbacks.
How did he do here? We've already seen him atop the pass completed and accuracy charts for the match, and I doubt anyone's forgotten his superb assist on Danny Sturridge's goal, but let's take a look at the radials for the match to see if we can get a better read on how Meireles compares to Mikel:
Figure 4: Chelsea radial passing maps vs. Sunderland, 9/10/11. Powered by Tableau.
That's a lot of forward passes, eh? 31 of 84 attempted passes by Meireles went forward, which is 37%. Only 13 went backwards. Surely that's an improvement on the usually-stodgy Mikel? How's his distribution looked this year?
Figure 5: Passing distribution comparison between John Obi Mikel and Raul Meireles.
It's fascinating that Meireles doesn't really turn out much better than Mikel by this measure, considering how influential the Portuguese was on Saturday and Mikel's reputation for never passing forwards. Considering that Meireles was singled out for special praise by Villas-Boas after the match, he can't have put in a performance equivalent to an average Mikel day, can he?
No, he can't. The types of passes played by both players are similar, but the frequency is not. This is because Meireles had the ball significantly more often than Mikel has over the course of the season - he received a pass once every 80 seconds while Mikel's season average is 104. The manager's praise for Meireles was very specific, too - the £12M man was lauded for making himself available as an outlet to his teammates. This is something Mikel doesn't do very well (although against West Bromwich Albion, his best match, the Nigerian did receive the ball 73 times), and that means that a midfield with Meireles is likely to be better at moving the ball around than one without.
Remember, it's receive, pass and offer. The last bit is just as important as the first two.
Of course, the role of a holding midfielder is to shield the back line, and here Meireles falls a little short. Mikel can tackle and can read plays well enough to break them up before they become much of a threat, making him a perfectly capable holding player when the opposition has possession. Meireles has the vision to interfere with plays - he had four interceptions in Saturday's match, a very good figure - but his tackling is awful, to the point where it's a liability.
In other words, for toothless teams like Sunderland where the goal is to score and score often without worrying much about conceding, Meireles is probably a good choice as a holding midfielder. Against actual threats, things look very different. So despite an assured performance from the Portuguese in Sunderland, Mikel probably shouldn't be worried about his place for the Bayer Leverkusen or Manchester United matches.
Nicolas Anelka, Centre Forward
There's something about the Stadium of Light which makes managers use Nicolas Anelka in strange ways. Last season, Carlo Ancelotti played Anelka as a trequartista behind a strike pairing of Didier Drogba and Salomon Kalou. This time around, Villas-Boas opted to use Anelka as a lone centre forward with a mandate to drop back and create. Ancelotti's unorthadox approach worked a charm (until it ran into the buzzsaw of Liverpool in a 3-6-1 the next time out). Anelka-as-Messi was, unsurprisingly, rather less successful).
We haven't seen Anelka play well as a lone striker since he helped Chelsea dominate the first half of that 2-1 win at Old Trafford in April 2010. Anelka's an odd case as a central attacker, because he likes to stay deep for two reasons. Firstly, he's quick enough to outpace most defenders playing a high line, so playing shallow in order to lure out the centre backs helps him. Secondly, he likes to drop into the midfield in order to influence play. That's all well and good, but there are a number of repercussions for the team when Anelka is used like that.
Firstly, the emphasis shifts away from the centre of the pitch and towards the flanks for chance creation, which left Daniel Sturridge and Juan Mata responsible for most of the heavily lifting in terms of breaking down Sunderland's defence. The latter did an admirable job, but the ball rarely ended up with Sturridge - he received the ball just 23 times duing his hour on the pitch.
Figure 6: Nicolas Anelka dropping deep changes the emphasis of attack as well as impacting the midfield.
Anelka as a centre forward also changes Frank Lampard's job, to a certain extent, since more than Torres or Drogba Anelka likes to occupy the spaces that Lampard would normally be looking to exploit. This, of course, has a synergy with playing Meireles as the holding midfielder, since Lampard had to drop deeper to protect the Portuguese player.
None of these are bad things per se, but they rely on Anelka to actually create if he's going to go all false nine on us. He did not. Anelka's main contributions to the game came when he neglected the passing aspect of his game to act as a more traditional centre forward, coming close to scoring a few times when played through by Mata and winning the free kick that resulted in Chelsea's opening goal, as pointed out astutely by Carefree Chronicles.
It's also worth pointing out that Lampard, who was impressive against Norwich, didn't play particularly well here. How much of that is attributable to Anelka's positioning is anyone's guess. Chelsea fans have done a lot of assigning blame to the midfield for Torres' lack of scoring, so it may well be that Lampard's relatively weak game was causing Anelka problems, but it seems likelier to me that it was the other way around in this case.
Considering Torres' awful cameo in the late stages and Salomon Kalou being himself, it's difficult to see this performance as something that will prevent Anelka from making a start against Bayer Leverkusen in midweek, but it sure didn't do him any favours.
Danny Sturridge, Man Of The Match?
Figure 7: Chalkboard comparison: Daniel Sturridge at Sunderland vs. Salomon Kalou at Stoke.
In the poll attached to We Ain't Got No History's match report, Danny Sturridge received 23% of your votes for Chelsea's man of the match, behind Juan Mata with 45% and ahead of the hugely influential figures of Ashley Cole and Raul Meireles. From a certain point of view, this is difficult to understand - his raw numbers suggest that he was barely involved compared to Mata. In fact, take a look at the chalkboards to the right and see what you make of them. One is Sturridge against Sunderland with the pass to John Terry excluded, and the other is a mystery Chelsea right winger who we'll call Salomon Kalou in order to protect his anonymity.
Manage to work out who's who there? For the record, Sturridge is A and Kalou is B, but the similarity is a little bit disturbing, at least to me. Of course, there's something deeply unfair about ignoring the fact that the 22-year-old was involved in the two most important plays of the game - he fed John Terry for the opener and of course scored the second in absolutely sublime fashion when he fastened onto a Meireles through ball and backheeled past Simon Mignolet, and any analysis that does so while pretending to be objective is deeply flawed.
Back when Didier Drogba was scoring against Arsenal pretty much whenever he got out of bed, Arsene Wenger described him as a player who does nothing all game except score twice. I don't think anyone's worked out whether or not that was a compliment, but it's a comment that can be extended to Sturridge's game on Saturday. He was clearly a vital component in the win (hence the man of the match voting), but equally he didn't do much of anything except win the game for Chelsea.
This isn't really an argument about whether or not Sturridge deserves man of the match for his display, nor is it a critique of his performance - considering how often he got the ball* he did spectacularly well. What it is is a request to the rest of the team: This is what he did when you played him 23 passes. Next time, try more of them.
*Although we should bear in mind that as with Meireles, at least some of the responsibility for getting passed to lies on the potential receiver**.
**Yeah, Torres. I'm looking at you.
Verticals & Horizontals
One of the things that Villas-Boas has been emphasising with the club has been the need for fast-paced, vertical football. Chelsea have been having real problems getting the ball to the forward three this year, and it was to everyone's relief that the midfielders were able to play the ball forwards far quicker than they have done in previous games.
But despite the emphasis on vertical passing, there were very few true chances created by a Chelsea side that should have run rampant over a dismal Sunderland. This is striking, as the speed of buildup play has been cited as the major issue causing Chelsea to be somewhat less robust in attack than they'd have liked. Was quicker play not the answer after all?
Well, not quite. Chelsea might have been better on the vertical, but it was only on one axis - the left. Cole was sublime, and Mata saw more of the ball than any other Blues winger has managed this season, but over on the other side Sturridge was barely involved and Bosingwa had a forgettable match (at least when you compare it to his scintilating start to the season. There was also zero thrust from the centre, where Anelka played withdrawn rather than forcing the issue.
Figure 8: Chelsea attacking patterns vs. Sunderland, 9/10/11. Via whoscored.com.
The only times Chelsea really created a threat to Simon Mignolet's goal was when a winger cut inside (Mata's pass to Anelka for the free kick which led to the first goal, Sturridge's run for the second, Malouda's volley) and for a team that relies on those wide players for creativity, that's not ideal. Worse, the team was almost completely unable to switch the play from side to side, which meant that Bosingwa virtually never received the ball despite Sunderland cheating on our left - 44% of Chelsea's attacking play came in that third of the pitch.
It was Frank Lampard who served as the left-right link against Norwich City in the 3-1 win two weeks ago, pinging long balls to both fullbacks, but here he did nothing of the sort. Seven long (25 yard+) passes from Lampard found Bosingwa when Stamford Bridge hosted the Canaries; but in the Stadium of Light Lampard didn't hit a single cross-field ball to the right back. Small wonder that Sturridge, ahead of Bosingwa, was starved of possession.
Chelsea need to move the ball forward quickly, yes, but they can't let their desire to do that let them fall into the trap of only attacking in one area. In order to be successful, the club needs to play the ball horizontally as well as vertically - switching the point of attack more often (i.e.at all) would allow for deeper penetration into relatively unreinforced zones.
Fortunately, there doesn't seem to be a reason that Chelsea couldn't do both in a match - it's a matter of tasking Lampard or Meireles with that role rather than being either a defensive midfielder or sitting around idly doing not much of anything. There's some hope, then, that the Blues can put everything together and turn into a truly formidable offence rather than the flawed-with-flashes-of-genius incarnation we're seeing now.
The general narrative is that this match was Chelsea's first good performance of the season. It wasn't. Despite the presence of Juan Mata in the starting eleven and a Sunderland midfield which did approximately nothing all game, the Blues managed just five shots on Simon Mignolet's goal after ninety minutes of football - that's less than they managed against Stoke City (and they hit the woodwork twice against the Potters as well).
We're still waiting on a Chelsea goal that comes from short passes in buildup play. So far this year we've had a goal that came from an accidental pass that was deflected in (Anelka, West Brom), a converted cross (Malouda, West Brom), a phenomenal long-range shot (Bosingwa, Norwich), a penalty earned via a long pass forward (Lampard, Norwich) amd a major defensive error (Mata, Norwich). Neither of the goals scored by John Terry and Daniel Sturridge against Sunderland were the result of careful buildup play, and there were very few good chances in the match either.
In other words, it was a win that does little to show that the main concerns about Chelsea have been addressed. That's not really being negative - I'm pretty confident that this team has it in them to come together properly in the near future, and right now they're winning games without being anything like near their best, but as it stands they haven't managed to gel quite yet, and the Sunderland game shouldn't be taken as evidence that the attack has turned the corner.