Statistics In Football; The Edin Dzeko Case Study

A war between those who think passion is an important, interesting word and those who find football club accountants to be admirable people has broken out this season. Statistics are wonderful. They are being used more and more widely in football as they have been used throughout the world in general for centuries – Namely, to recognise patterns and progressions, worrying falsehoods and encouraging trends. Prozone, WhoScored, Squawka, Opta. Enlightening.

Well, sometimes they are enlightening. Sometimes they suck the life out of a sport which is at its best when putting your emotions through the ringer.

No statistic can explain what City fans felt inside the Etihad that day. No statistic can explain what the players did. Words can hardly do that, let alone numbers. When you’re in the stands and your team score a goal you would have to be inhuman to care about pass completion rates and interceptions. Collective identity and a sense of belonging to a club cannot be defined. The value of a manager and players who understand the club cannot be quantified so easily.

On the other hand, statistics, if used correctly, can explain things about football that our emotions cannot get near, such as measuring players’ fitness or properly analysing an opponent’s defensive weaknesses over a period of time to identify the most vulnerable points.

Where statistics fall down is when people begin blindly accepting numbers without examining them. Not assigning context to them, not asking yourself what secrets they are hiding, is intellectually dishonest. This is using statistics to purposefully mislead yourself, or others, into believing things that are not true because it suits – a great evil of the modern world.

A statistic is an objectively true fact in the purest way which anyone, with a bit of work, can produce. It is a true fact that most tragic spaceship explosions involve human pilots. It is a true fact that vast majority of banknotes in the UK are contaminated with cocaine.

Are the spaceship explosions the fault of the pilots, or is sitting on top of rocket engine inside a giant space firework inherently risky outside of human control? Is everyone in Britain on the gear 10 hours a day, or is the machinery at the bank note printers in Loughton also where they manufacture cocaine? The original facts are true, but they do not inform us of the truth.

When it comes to football especially, it is absolutely vital that statistics are used with context.

Our very own Edin Dzeko provides us with the perfect case study of the above principle.

dzeko vs man united

It is a true fact that Edin Dzeko scores a fair amount of goals.

This season in the Premier League his total compares favourably with Alvaro Negredo, with totals of 11 and 9 respectively. Smilarly, Dzeko scores in the Premier League at a rate of a goal every 134.5 minutes, compared to Negredo’s rate of a goal every 182.9 minutes.

The percentage of what I have unilaterally defined as “important goals” (goals which either draw City level or put them in the lead) is also very similar; 54.50% of the Bosnian’s goals are ‘important’ compared to 55.50% of the Spaniard’s.

Both of them have scored their fair share against higher quality opposition (Arsenal, Tottenham, Chelsea, Southampton, Everton…Manchester United). Dzeko’s goals mostly came away from home, Negredo’s at the Etihad.

So why, until Negredo’s recent run of bad form, was the Spaniard almost universally held to have had a better, more useful season? He was even voted signing of the season by the managers in the league, and yet the differences between the two Premier League goal scoring records are so similar that the difference is negligible?

Because goals are not all that matter. Statistics have to be placed into context.

In his excellent book The Nowhere Men, Michael Calvin visits Brentford and talks to their owner Matthew Benham who offers a different perspective to just looking at goals:

“If I am looking at a striker I absolutely do not care about his goalscoring record. For me [Clive], the only thing that is interesting is how the team do collectively, offensively and defensively, within the context of an individual’s performance […] If you look at a striker, at a random point in his career, and you want to know whether he is going to score in the next game, knowing how many he has in the last ten, twenty, forty games tells you very little” (1)

This offers a window into why Negredo perhaps has been popularly seen to have had a better season. The two strikers’ approach off the ball is, for the most part, wildly different. Negredo is constantly pressing, harrying, bullying the other team; With some notable exception, Dzeko rarely does this.

The ball sticks to Negredo whose outstanding close control allows him to hold the ball up for others very effectively. Dzeko regularly has a first touch which can be described as “expansive” and is generally terrible at holding the ball up, often running into a dead end or the opposition instead.

The Premier League Owl said it best:

“After three years and having watched him play dozens and dozens of times, I’m still really not sure what calibre of forward he really is. For every brilliantly-taken goal there have been three baffling misses, and for every game in which he has been a decisive factor there have been just as many anonymous performances.”

Anyone who says they haven’t felt similar things about Dzeko is lying, but they are much harder to quantify with statistics than simply counting piles of goals. Both are important parts of the whole, but cannot stand on their own. And nor should they.

And that’s the point. It is all very well saying that Dzeko scores goals and therefore should not be sold, or should be played more, or that the club should buy him a new car, but that is not the end of it. Football without gut feeling would be nothing, but without intelligent use of statistics that emotional reaction is an uninformed knee jerk ruled by personal prejudice. By using our head and heart in tandem we can be better at our super expensive hobby of being football fans.

(1) Calvin, Michael, The Nowhere Men (Century, 2013) pages 333-334

All Edin Dzeko statistics provided by the wonderful, resourceful StatCity and are correct as of 10/04/2014

Written by Alex Timperley who is on Twitter

Follow Typical City on Twitter and Like on Facebook.


2 comments on “Statistics In Football; The Edin Dzeko Case Study
  1. As to what Edin Dzeko offers beyond scoring goals, consider some Opta statistics for the 2012-13 Premier League.
    1) Defensive Clearances: Dzeko 17; Aguero 1; Tevez 1.
    2) Blocked Shots: Dzeko 6; Aguero 0; Tevez 3.
    3) Defensive Tackles made / won: Dzeko 10/10; Aguero 26/20; Tevez 28/19.
    4) Duels Won: Dzeko 126; Aguero 113; Tevez 123.
    5) Aerial Duels Won: Dzeko 87; Aguero 5; Tevez 13.
    6) Unsuccessful touches: Dzeko 45; Aguero 44; Tevez 50.
    7) Overran Ball: Dzeko 1; Aguero 4; Tevez 7.
    8) Offsides: Dzeko 7; Aguero 14; Tevez 18.
    9) Goal Assists in Open Play: Dzeko 3; Aguero 2; Tevez 5.
    10) Using the Opta definition of assists, the MINUTES PER GOAL INVOLVEMENT (i.e., goals + assists) were: Dzeko 107; Aguero 139; Tevez 127.
    I do not have comparable statistics for this season. Perhaps StatCity does.
    Manuel Pellegrini has played Edin Dzeko regularly rather than Alvaro Negredo primarily because the former has been playing an integral role in City’s build-ups.
    Recently, when praising Pellegrini’s management skills, Txiki Begiristain singled out the improvement in Edin Dzeko’ form: “I am seeing a different player to the one when I came in one year and five months ago. The talent was there, all the qualities were there in this player but now he has the confidence from the manager.” What do you suppose he meant by “all the qualities”?

  2. Taking all Manchester City’s League and Cup games for this season, the numbers of ASSISTS are:
    Edin DZEKO 15
    David Silva 15
    Sergio Aguero 12
    Samir Nasri 9
    Yaya Toure 8
    Alvaro NEGREDO 8

Comments are closed.

%d bloggers like this: