Does Chelsea’s Midfield Miss Luiz? – A Look at Big Games

Updated: March 16, 2015

I don’t know if you heard, but Chelsea drew a game a few days ago and lost a European tie. The events and nature of this performance led to hyperbolic hand-wringing from all quarters, and as I’ve mentioned before, losses are a great time to start making noise about your favorite pet issue.

Chelsea were bad all over the field against PSG… Sort of. There’s a lot of questioning desire and heart and tactics regarding the performance with a man advantage (and in so-called “big games”, but it’s also true that Chelsea took the lead to win the tie with ten minutes remaining, and took the lead in extra time. That these leads were followed by equalizers simply changes the perspective a bit. In a way, this makes analyzing the game more difficult. After games like the Tottenham match it’s easy to start ripping into people; the team conceded five goals and was shell-shocked. Pretty much anyone could be criticized, from midfielder to defenders to keeper (though perhaps not the attackers).

What I don’t get about the PSG game is that the same Wild-West style criticism seems to still apply. The manager? Heavily criticized for not pushing hard enough. The defenders? Criticized for not marking corners. The forwards? Criticized for not scoring from open play. The midfielders? Criticized for not controlling the game well enough.

It’s not that these criticisms aren’t valid, I just don’t think they really exist if Chelsea don’t have ten seconds of horrific set piece defending.

Therein lies my issue with the post-PSG analysis. There are a host of reasons to be upset about the game for both sets of fans, ranging from refereeing decisions (sure, Zlatan shouldn’t have been sent off, but Luiz definitely should have and he scored the game-saving goal, but I digress) to striker performance (hey what’s up Cavani and don’t think I didn’t notice Diego putting 1/6 shots on target). But one of the areas I’m surprised to hear cacophonous criticism: the midfield. Specifically Cesc Fabregas.

There are a variety of things to talk about here, including: the supposed lack of Chelsea midfield dominance (note that by ‘supposed’ I’m not saying Chelsea did dominate the midfield, but rather that it’s more complicated than a Boolean), the lack of culpability attributed to Matic for this lack of dominance, the apparent willingness to buy into player narrative, and the idea that Luiz would be the solution to these problems. This piece ended up being far larger than a single article, so those issues will be addressed in a separate piece in a couple days.

For now, let’s start further back.

Chelsea have had a midfield problem ever since Ballack left. This seems to be the general consensus from pundits and fans regarding Chelsea. Turns out when your midfield gets rotated between guys like Makelele, Essien, Lampard, and Ballack for half a decade and then all of those guys fade away through age or injury, there will be a drop-off in performance. Chelsea’s lack of true central midfield dominance has seen them play Ramires as part of a midfield two with an aging Lampard (probably better than you think but not good), Mikel with Lampard (bad), Mikel with Ramires (bad), a sizable amount of Meireles, and oh yeah s superbly talented center back in the midfield as a last resort.

Let me repeat that for clarification: as a last resort. David Luiz is not and has never really been a midfielder. That he was able to play in midfield with varying degrees of success does not mean that he is a midfielder. This is an important point to remember going forward.

So Chelsea had this midfield problem. They finally managed to alleviate it somewhat by restoring Matic to the fold from his Portuguese-themed level-up party; throwing him in against Manchester City helped us figure out that he was good. Luiz played midfield in that game, and was decent (here’s a link to the We Ain’t Got No History community ratings from that game, where he was rated about the middle of the pack). Keep this in the back of your mind.

The other issue to come out of this game (and the focus of this piece) is the true knelling of Chelsea as big game specialists. Last season, Chelsea’s record in big games was nothing short of spectacular: wins against the other two members of the top three in the home and away legs, a massive win over Arsenal (again with a decent-but-middle-of-the-pack player rating from Luiz; keep this in mind) at home, and… uh…

Well. I guess that PSG game. That’s interesting. There is the idea that Chelsea have lost the status of big game giants. I don’t necessarily disagree with this theory, but I think that this theory, as you might say, has some ‘splainin’ to do. Don’t get me wrong: I don’t think it’s incorrect to say that Chelsea have been worse in big games this season. They somewhat objectively have been, with two draws with City having virtually guaranteed that, at least in the league, and the draw/loss to PSG being the final nail in that particular coffin. What I do think must be discussed in some level of detail is the nature of big games, the nature of fortune, and the nature of narrative.

Big games take on a special meaning to the fans. They are inherently arbitrarily defined: Manchester United probably still count as a big game despite their being pretty irrelevant to the title race. Southampton, despite their strong start to the season, probably don’t. But, are all cup games big games? Or is it just a change in threshold, where say a cup game against Stoke is big but against Derby is not? Do all games against traditional rivals count? Fellow title rivals? What about cross-town lovable loofs (looves? leef?) with weird grins?

I think the concept of a big game is pretty nebulous, and you can get away with some hedging at the fringes depending on what you want to get across. This is not, in itself, a bad thing. If there was no real disagreement in soccer it would be spectacularly boring. Having said that, we can probably agree that games against Manchester City are big games, and that Champions League knockout ties are big games, unless maybe they’re against Galatasaray (also keep this in mind).

But PSG was a big game. Having taken too many words to establish that, let’s dive a little deeper.

Consistently performing well is hard. It is hard against small teams where you can get away with coasting. It is particularly hard to do against good teams, where strong motivation might not be enough. It is especially hard to do year after year team after team.

Chelsea are running into that this season. Bill Barnwell writes about the “Plexiglass theory” often with regards to American football (the term originated with Bill James, I believe), and it’s possible that Chelsea last season just ran full force into an elastic window. They managed to stretch the window pretty far — for instance shutting down Man City at home and throwing two goals past PSG thanks to Demba Ba. But eventually the basic principles of energy conservation would demand that Chelsea please stop bending the window.

You look at those City performances, and that Liverpool performance (quick notice that Luiz isn’t in the midfield for that one), and that PSG performance, and then you look at this year and the results differences are stark. Two wins against City, none this season. Two wins against Liverpool, and… wait. This year we have one in the league (in the only game), and a draw away from home in a cup competition, and another win.

Well certainly you look at that Arsenal home game (6-0! Luiz was in midfield, you can remember that too) and compare it to this season: a paltry 2-0 with no real drama.

OK but fine. That PSG game versus this one. We beat them by two last season, and drew this season, so clearly we are worse at big games. Except, well, we lost to PSG last season. We, uh, didn’t do that this season.

I guess my point here is: we fell victim to the Plexiglass principle from last season to this season. Sustaining 100% points records against the best teams in the league is nigh-on impossible. It’s also worth remembering that we weren’t even that good in big games last year.

Because you know we had some big games that we lost. You know we lost to Man City in the FA Cup, right? We didn’t lose to Man City this season. And you know after the PSG game we did play more games in the Champions League, right? And we didn’t win those games; in fact we lost one of them by two goals. And did I mention that we lost to PSG last year?

We were hardly invincible in big games last year. It just seems that way when you don’t win anything and have nothing else to really hold on to.

Big games have a hold on our minds, but perhaps they really shouldn’t. Because big games are often won in the way that many other games are: decided by luck and some measure of furtune. We beat City at home, but only because City screwed up in the last minute (whaddup Joe Hart + Demichelis?). And we beat Liverpool away from home pretty fair and square (because shooting from 20 yards is hardly shooting at all), but the game broke open because of a slip. Maybe if there were a no-slip boundary condition, the flow of that game is different (terrible engineering joke… now come back). Big games require some level of fortune. This season, Chelsea lost a lead to United because of a weird sending off and a subsequent mismatch of set piece marking. We had a penalty denied during the PSG game along with the aforementioned sending off shenanigans. There’s a lot of luck going on in this games.

Chelsea as a big game team primarily relies on the idea that we are good at limiting our opponents to sub-par situations even when we are, from pure amount of chance creation, being outplayed. For instance, though Liverpool took a metric ton of shoots against us at Anfield a combination of defensive shape and eventual desperation drove the Reds to attempt a lot of long-distance shots that had low probabilities for success. In fact, we’re still pretty good at this; despite being outshot in raw terms by City in both league legs, the expected goals* were pretty roughly similar. And the loss against Tottenham actually shows that we outcreated Spurs (though as Caley mentions the game state** has an effect on these expected goal numbers).

*Expected goals are a metric used to determine the quality of shots taken by utilizing information about the shot type and location based on historical averages. It’s sort of like how many goals a team should score in similar situations.

**Game state is essentially the net score of the match; for instance, a game where the home team is winning by two goals would be a +2 game state. This matters in that teams that are losing/chasing a game typically shoot more than teams that are even or winning.

And of course, against PSG…

Note that we 1) were even in terms of expected goals (with the addition of a penalty, and really what should have been another penalty) and 2) were not really outplayed in raw terms. We were outplayed in the sense that we perhaps should have been much better given that we had a man advantage for nearly ninety minutes. On the other hand, PSG were forced to chase the game all the way through, meaning there are some mitigating game state factors on that end (recall that teams trailing tend to shoot more).

But without being able to see visualizations from past games, we’re left to wonder: were we truly much worse against PSG than we have been in big games in the past? Maybe, but then again maybe not. We were certainly bad, but at least we didn’t lose by two (which happened three times in big games last season).

On a larger scale, we won 34 (63% of available) points in arbitrarily defined big games last season (I included all games against City, Arsenal, and Liverpool, CL knockouts, cup finals, and a couple other key games such as Stoke in the FA Cup and any cup loss, just to balance things). We’ve won 18 (54% of available) this season. The key point is that we’ve played less games (still have games against Liverpool, Arsenal, and United in the league). A couple wins would give us about 60% of available points in “big games” this season as in the last. I don’t have to tell you that that’s basically the same number.

There are some mitigating factors, of course: drawing twice to PSG doesn’t count as a loss to the total win-loss record but is really a loss, and we played tougher teams in the CL last season as we progressed. Of course, we didn’t get to play Galatasaray this season, either, so these things balance somewhat.

I guess if you really wish to believe so you could make the case that we’re demonstrably worse in big games this season than we were last season. Perhaps conceding a late equalizer to City makes you question our steel (but conceding an equalizer in the 49th doesn’t, as happened at home last season before Demichelis gifted us a win). Maybe those concerns are valid, but to me they seem fairly inconsequential.

See, last year we didn’t win anything. And I tell you what, I kinda like the change.

One Comment

  1. Sage

    March 16, 2015 at 1:15 PM

    Squad is better this year so we are winning titles now despite playing worse. You may be right that the PSG tie is getting overanalyzed and criticized due to everything being decided by set pieces. Getting two draws rather than a minor loss and a compensating win are not equal situations as you assert though. The away goals rule means it is important to attack well away, and to defend well at home. Chelsea did worse at the first task in particular in both fixtures this time around.

    But it is also true that this team is set up tactically poorer than last season. Conceding a winning state against 2 teams with 1 less man is a new low for Mou teams, and this year’s Chelsea has had some other bad losses which have been Mou/Chelsea records in terms of poor defensive performance.

Leave a Reply

Your email address will not be published. Required fields are marked *