First of all,  we wish all of our readers here at StatsBomb a very happy New Year.

As we are at the midpoint of the Premier League season it seems to be a good time to revisit the shot location charts of the leading teams.
In a series of short articles over the next few days I will provide the plots of shots taken and conceded by each team as a point of reference and  I’ll point anything of note that jumps out at me.

My Zones

A reminder for any new readers of the zones that the summary box below the plots refer to:


Each team plot has two images.  The image on the left reflects the shots (excluding penalties) that the team has taken this season, while the image on the right sets out the shots they have allowed.
The table underneath the image summarizes the shots by zone, and in order to aid comparison the numbers highlighted in yellow are the league averages.

Where better to start off the team by team analysis than with the league leaders – Arsenal?



Despite 5 teams shooting more often than Arsenal, only Man City have taken more shots than The Gunners from Prime locations.  The desire to work the ball into close positions has long been a trait of Wenger’s team and this tactic has continued this season.
Arsenal are the only team not to have taken a single shot from Very Poor locations; yet further affirmation of their shooting discipline.

On the downside, the relatively low volume of shots should be a concern for Arsenal fans, as it is unlikely that they can with a league with only the 6th most shots in the league.


For a team with aspirations of winning the league Arsenal are in the unfortunate position of having to rely on a striker that is clearly not elite at doing what he is paid to do – put the ball in the net.  Giroud’s supporters will undoubtedly point to his support play and the way he can bring Arsenal’s other attacking talent into the game, but the simple truth is that he’s not a very good finisher.

Here are the shots that Giroud has taken this season in the EPL:


Giroud’s shooting positions are superb.  Of the leading Premier League strikers only Aguero has a higher average ExpG figure than the Frenchman, so his 7 goals from open play is a disappointing return.

The following images demonstrate why Giroud is not elite class, and why Arsenal will need an upgrade if they are to continue their attempt to end the season as league champions.

Giroud Curve

This is the first time that I’m using these cumulative probability curves, they were suggested to me by Devin Pleuler and I think they convey in a snapshot the shooting performance of a player – Giroud in this case.

The number of goals are listed on the horizontal axis and the curve represents the probability, based on his shot location and type, that the player will have scored at least those number of goals.
The red dot represents the actual number of goals scored (8 in the current season), and it intersects the vertical axis at about 90%.

Thus, in 90% of the cases our model expected Giroud to have scored at least 8 goals.   Had he converted the goals at the rate our model (created with Constantinos Chappas) expected his red dot would have been at the 50% mark – which is at about 11 goals in this case.
90% is not good – he’s about 3 goals behind where our model expected him to be at given the shots he has attempted.

One poor season could be forgiven, but the image below is from last season and Giroud’s red dot is in almost the same place again.  His 11 goals underperformed our scoring expectation by about 6 goals last year.

Giroud Curve 2012

Arsenal could and should do better in terms of being able to rely on their main striker to convert a fair share of chances that fall his way.  In this instance it may be no bad thing that Wenger’s hand might just be forced due to the injury suffered by the Frenchman at Newcastle on Sunday.

Going Backwards

Defensively Arsenal have been superb; at least in terms of where the shots are struck from.  The paltry 27% of shots they allow from Prime locations is comfortably the lowest in the league and explains why the average chance that Arsenal gives up has the lowest probability of being scored in the league.
Their concession of just 61 shots from Prime locations is almost a shot per game better than any of the other teams in the Premier League – Liverpool fare next best on this measure with 77 shots allowed.

Arsenal’s ratio of Prime Shots for compared to conceded is greater than 2 (123:61) and is an elite number.

Defensively Arsenal are sound, and they certainly have a platform from which they could build a title winning charge.

  • Geraint Morgan

    Is there work that can be done looking at the success of a shot depending on who created the chance? Does a chance created by ozil or walcott have a very different likelihood of being successful?

    • Colin Trainor

      We don’t have sufficient data coverage to be able to begin to answer that question Geraint. It’s tough enough to analyse shooting expectation by locations and shot type, never mind going a level deeper to look at where the pass was played from. To then go deeper again and look on the player that made the pass for the shot would require a serious amount of data – which doesn’t seem to be out there.

      • Geraint Morgan

        Fair enough. Data is both wonderful but annoying when you know it cannot answer the questions you want to ask

  • Phil Markham

    Thanks for a great article, Colin. You say: “The number of goals are listed on the horizontal axis and the curve represents the probability, based on his shot location and type.”

    Can you clarify what you mean by “shot type”?

    • Colin Trainor

      Phil, thanks for the feedback.

      “Shot type” refers to whether it was a header or a shot with the foot, was it a direct free kick etc.

Improve Performance and Productivity in Your Club:
State-of-the-art Football Analytics