clock menu more-arrow no yes

Filed under:

Breaking down Branislav Ivanovic's terrible, horrible, no good, very bad day against FC Porto

New, comments
Julian Finney/Getty Images

First things first: Branislav Ivanović is not having the greatest season.  He's not even having a good season.  He's having a bad season.  Pretty, pretty, pretty... very bad.  Just like everyone else, especially the likes of Eden Hazard and Nemanja Matić and Cesc Fàbregas and Diego Costa and John Terry and César Azpilicueta and on and on and on.  Should Ivanović be dropped?  Maybe?  Probably?  Baba Rahman has looked enthusiastic going forward, less so staying back and defending.  Could he be worse than Branislav?  I think there's a sizable gap left to exploit below, but maybe the sinkhole isn't so deep after all.

Regardless, here we have two award-winning full backs, one voted best in England, the other best in Germany.  Such awards can be fairly ridiculous, as we've seen many times with UEFA's selections.  Recipients owe as much to their reputations as they do to their individual performances.  Even more crucially, they owe the system in which they operate on a day-to-day basis.  For example, Piqué is an excellent fit for the Barcelona system, which emphasizes possession and pressure over what we would think of a "traditional" defending.  So is Mascherano, who would not be a good fit for most defenses, thanks in part to his small stature.  Similarly, what has made Ivanović so vital the last couple years is the system that we've used.  He's been one of our main avenues for outlet and attack, while the gaps behind him were generally covered by midfielders shifting over, wingers tracking back, and center backs mopping up.  How many times have we seen Cahill slide over and clear a ball from in-behind the fullback?  (And, in contrast, how many times have we seen Zouma do that lately?)

It's all about the system and the system is breaking down in multiple places.  Ivanović has been bad, but our midfield has been even more non-existent. The support from other pieces has been inconsistent at best.  Porto's first goal started with Ivanović isolated 2-v-1 (or even 3-v-1?), for example.  How many times has Eden Hazard had the luxury of going 2-v-1 at the other end?

ivanovic 3-v-1

Ivanović, quite literally, is the only Chelsea player on the right-hand side (lower half, as we look at it) of the pitch.  The two players in the middle are Mikel and Pedro, the latter covering for both Ramires (who's jogging back after a mad dash forward) and Fàbregas, too, who's not even on Chelsea's half of the pitch.  Cesc had actually been pressuring the goalkeeper in the previous passage of play and he does eventually recover to keep an eye on Danilo(?).  But Ramires is operating in a nonchalant vacuum as the action flows around him.  Thus Pedro's decision to vacate the wing and come inside to help Mikel.  To his credit, Pedro does later recover his wide position (Porto slow it down and Bane manages to delay Brahimi as well), but then fails to actually help out Ivanović.  Same goes for Zouma, who basically just stands there.  No, literally, just stands there through the whole play.

Ivanovic Pedro Zouma Brahimi

This is as close as Pedro gets.  Compare with how close Willian was tracking back on the opposite wing most of the day.  That's not to say that Ivanović is faultless.  But he's not alone.  Brahimi already attempted to go down the line a couple times by this point, so the cutback was a surprise.  Plus, that's where the right back is supposed to have help!  Brahimi was the only dude on their team whom we really had to worry about.  Where's the help?!  (In fairness, we did manage to play similar situations correctly a few times as well.)

While he does get blamed unfairly in multiple instances (see: Newcastle's first goal), this wasn't the first game in which, after a while, Ivanović just had no idea how to play his man.  We see Hazard do this all the time to defenders (well, at least we used to); a fullback having to repeatedly go 1-v-1, 2-v-1 against attackers is a surefire recipe for disaster, regardless of the names involved.  And when it's a fullback who's out of form, it's an even more surefire recipe.

So, to reiterate, again and again, Ivanović did not have a good day.  He was never going to, left 1-v-1 with Brahimi all the time.  That's slightly hyperbolic (never say never), but it still not as hyperbolic as the "stat" that made its Twitter rounds after the match, thanks to the wonderful trolls at Squawka, who apparently not only specialize in the most misleading headlines in all the business, but also in some of the most misleading "stats."

Cherrypicking aside, it's not all their fault.  We've talked many times before about the inadequacy of Opta football statistics, which capture strictly on-the-ball events in a game that's at least 75% about space, positioning, and movement.  The situation is even worse for defenders, where the name of the game is probably more like 90% positioning and anticipation.  So, for example, all those times that Brahimi was turned away by good closing-down or good shape being kept in defense and midfield are not recorded anywhere.

Even on-the-ball events leave us scratching our heads regularly — things like possession being equal to pass volume or "chances created" meaning "pass before shot" or where the line might be between a cross and a pass or how do you even judge the quality and success of a cross, just to name a few of the most obvious ones.  In Branislav's case, the one that everybody concentrated on, let's see how Opta (which is the common source for sites like Squawka, StatsZone, or WhoScored) defines a "tackle".

A tackle is defined as where a player connects with the ball in ground challenge where he successfully takes the ball away from the man in possession. All tackles are really a successful event.

A Tackle Won is deemed to be where the tackler or one of his team-mates regains possession as a result of the challenge, or that the ball goes out of play and is "safe".
A Tackle Lost is where a tackle is made but the ball goes to an opposition player.

In case you missed it:  "All tackles are really a successful event."  When Opta says that Ivanović "won" just 1 tackle, they actually mean that he made 8 successful tackles, but Chelsea only ended up with possession once.  Or at least that's how it's supposed to work.  Squawka then also include fouls as part of their missed tackles count (two committed), so we end up with 1/10.

squawka ivanovic

But still, great for Ivanović, right?  Not too bad with 8 tackles and 2 fouls.  Well, not so fast.   Turns out, Opta's definition isn't really being used as intended.  Let's run down Bane's on-the-ball challenges and see what we find out.

This seems like splitting hairs, but I suppose you have to do that in discrete statistics sometimes.

Ivanović clearly wins the ball, but it ends up with somebody in a Porto shirt.  So, tackle "lost" which is absolute nonsense in the way I'd guess most people would commonly think of a "tackle."  He won the ball!

Ivanović gets no credit for this play.  He does get a completed pass eventually, and Brahimi probably gets a turnover or some such.  I didn't check.  The stats we have are wholly inadequate to describe this bit of action from the Chelsea defender.

Ball ends up with Porto man, so tackle lost.  That's either a good tackle or a foul.

As somebody on Twitter said:  Lost?  Yes, yes he is.  Tackle?  I don't see one.

Turns out, when an opponent "takes-on" a defender and wins, the defender is automatically assigned a lost tackle.  Which runs contrary not only to Opta's definition, but also to common sense and what we see here, where Ivanović didn't actually attempt to tackle.  Because that would've been a very foolish thing to do inside the area.  I've seen some places go with a "dribbled by" stat, which perhaps is a better descriptor.  Either way, we're dealing with inadequate statistics.

In Squawka's world, this foul is a lost tackle.  Double whammy.

Yeah, this one's just sad in all sorts of ways.

Foul.

Ivanović gets no credit for this play, nor is he punished by a foul call.  It has to be one or the other, no?

Inside help, Bane beaten down the line not for the first time.  Brahimi's good, we know this.  Still, that's not a tackle in any sense of the word tackle.

There's a lot going on here.  Ivanović steps in initially, then Brahimi slips.  Then there's a tussle with Imbula and the ball's up for grabs for a minute as Bane gets a toe in (this is a non-event for the statisticians), then Imbula turns on beast-mode.  Bonus LOLs: Fàbregas attempts to win the ball back by breathing heavily on Imbula.

Ends the day on a high note.  Though of course it's only a "won" tackle because Azpilicueta recovers the ball.

So, if you're still with me, here are the counts:

  • Opta:  1/10 tackles
  • StatsZone:  4/8 tackles — they don't include fouls as lost tackles.  They also seem to use more common sense when it comes to won/lost tackles.
  • My count:  5 tackles ("all tackles are successful events"), 2 fouls, 2/5 1-v-1s ... or really, just scrap this whole system and come up with something better altogether.

Again, none of this is meant to exonerate Ivanović.  He was not good.  Though, I maintain that his two biggest mistakes were the ridiculously bad clearing header in the first half and laying out for an attempted clearance on the ball over the top in the second, which eventually led to the corner from which Porto scored.  Ivanović's tendency to go to ground is a nasty habit as well.  Stay on your feet, son, as Ray Wilkins used to say.

What all this is meant to do is indict — or at least get you to think about it a bit — this ridiculous "tackling" statistic and people bringing it up as some sort of meaningful indicator of a player's quality of play (especially while ignoring other meaningless numbers assigned to things like interceptions, blocks, recoveries, clearances, etc).  Yes, Ivanović was bad.  Probably not as bad as you think.  And certainly not as bad as the Squawka stats make him out to be.