Chelsea left it late to beat Manchester City but were never in any real danger of losing the match to their closest contended for 3rd in the Premier League. Throughout the game they had total control of the central midfield despite there being ostensibly a three vs. two situation in the visitor's favour, and Edin Dzeko never troubled the defensive pairing of David Luiz and John Terry. Why? Let's look at the formation chart after the jump...
Chelsea played what will probably be called a flat 4-4-2 but to me looked more like a very wide 4-1-3-2/4-4-2 diamond. Whatever you want to call it, it featured Michael Essien dropping deep to pick up Yaya Toure, Frank Lampard advanced in the centre, Ramires as a hybrid right winger/right central midfielder, and Florent Malouda balanced between left forward and left midfielder. And if you're a little confused after reading that sentence, I don't blame you.
Against my better judgment, I'm going to call Lampard and Essien Chelsea's central pairing despite the two really playing on different bands. For Manchester City, the 4-2-3-1 mean that the central midfielders were ostensibly Toure, Nigel de Jong, and Gareth Barry, which would normally necessitate Ramires drifting inside to cancel out the numerical difference. Instead, neither of the two more defensive midfielders advanced high enough up the pitch to put any pressure on the middle, which ensured that they didn't get caught out defensively but also meant that Chelsea were able to keep hold of the ball far too easily.
Roberto Mancini is notoriously conservative, and while it's probably wise to be wary of Chelsea at Stamford Bridge his side would have been much better off if one of the two holding players in City's shape were actually capable of going forward. Nigel de Jong had a magnificent game as a defensive shield (34/35 passes, four tackles, six interceptions), and it's not like Gareth Barry played poorly, but City were acting much more like a 6-3-1 than anything else, and with the attacking talent available to them would have been better served asking Chelsea questions.
The use of James Milner to try to defend against Ashley Cole was also curious. Milner is a hard-working, versatile player, but he isn't a huge attacking threat as a winger and would often drift infield to try to disrupt the Chelsea midfield. Cole, meanwhile, was free to bomb up the flank to support Florent Malouda, which is where the hosts did most of their attacking from for the majority of the game.
If Adam Johnson had been fit, simply stationing him on the right flank would probably have been more effective in keeping Cole penned up - and Milner could have taken Barry's place and put his terrior instincts to good use in the centre of the pitch. As it was, Milner didn't do anything useful and the situation stopped Micah Richards from going forward until the very end of the match.
Edin Dzeko's magnificent incompetence on the attack is something that tactical analysis isn't really fit to describe - he was bad, and the Luiz/Terry pairing were very good (Luiz had eight interceptions, half of Chelsea's total), allowing Chelsea to leave Branislav Ivanovic covering the right side against David Silva knowing that Luiz would be free to clean up should Silva have escaped. Aleksandar Kolarov attacked more than Richards did down that flank, but with Ramires free to go chasing after him never really created and two on one situations with Silva against Ivanovic.
Of course, City's toothless attack came with a real bonus - the defence was virtually impenetrable. Salomon Kalou and Fernando Torres both worked extremely hard to open up space but even if they could get past the excellent Vincent Kompany or Joleon Lescott, Barry and de Jong were available to cover. Mancini came to defend, and his side defended excellently.
Carlo Ancelotti re-shuffled his side with twenty minutes to go, replacing each of his most advanced players and going to a 4-3-3(ish) shape. Of went Kalou, Torres, and Malouda, and on came Didier Drogba, Nicolas Anelka, and Yuri Zhirkov. Since Chelsea went on to score two goals, the impression is that the substitutions turned the game, but I think that's far from clear cut. Anelka, certainly, was totally uninvolved, getting something like one touch every five minutes, but even beyond that it's difficult to say what exactly changed.
The first goal was from a free kick that had been won by Luiz from another free kick (won by Anelka, incidentally), and while the assist was via Drogba, I have a difficult time ascribing what amounted to a simple set-piece goal to a tactical or personnel switch. After that, the game opened up and Mancini pushed players forward, introducing Mario Balotelli and Adam Johnson, and City's aggressiveness directly led to Ramires' beautiful second goal.
So, in short, one team dominated everything while the other side withdrew into a shell that lasted for nearly eighty minutes before being undone by a set piece. Prior to the goal, Chelsea's best chances from open play were created by the Kalou/Torres partnership, and it's unclear what Ancelotti's substitutions actually did to change the game. However, they'll be remembered as brilliant moves because the team then won.
Let's take a look at the passing:
Figure 2: Pass Completion and Frequency (15-min weighted averages), Manchester City vs. Chelsea, 3/20/11. Powered by Tableau.
Figure 3: Individual passing for Manchester City vs. Chelsea, 3/20/11. Powered by Tableau.
Yikes. City were totally and thoroughly destroyed here. Aside from maybe ten minutes at the beginning of the match when they did a reasonable job retaining the ball against Chelsea, they did virtually with the ball, which goes back to the defensive band of midfielders being used solely to disrupt the home side's attacks. The only Chelsea starter in the midfield or the defence who didn't outpass everyone on Manchester City was Ramires, but that's mainly because Chelsea were always looking to go up the middle or to the left side when conducting the attack.
All in all, I think Mancini made the game easy for the Blues. He came to Stamford Bridge hoping to park the bus, which meant Chelsea could hurl everything they had at his team. Without the threat of Carlos Tevez on the counterattack, they were never going to score a goal, which mean the Blues only need to sneak one to win it, which they managed on a free kick.
Speaking of set plays, let's look at the last few goals Chelsea have scored:
- David Luiz vs. Manchester United: Defenders still in the area following a corner.
- Frank Lampard vs. Manchester United: Penalty.
- John Terry vs. Blackpool: Corner.
- Frank Lampard vs. Blackpool: Penalty.
- Frank Lampard vs. Blackpool: Open play.
- David Luiz vs. Manchester City: Free kick at the corner flag.
- Ramires vs. Manchester City: Open play.
That's not a lot of open play goals, really. It would be quite nice to see us generate more chances for our strikers and less reliant on doing well from set plays (although it's also nice to be doing well on set plays too). Chelsea are playing much better than in November/December but they're still not really back to their free-flowing best.