Measuring performances in the Eredivisie – Goal scoring chances

Regular followers of the blog know by now that I try to take a particular interest in measuring the performances of football teams, and should more and better data be available in the future, of individual players too. This post continues that trend, being the first of a planned three-part mini series looking at team performances in the Eredivisie, with 13 of all 34 games played.



As always, a word of thanks goes out to Infostrada Sports for courteously providing the data used in this analysis. As of recently, the Eredivisie Live website provides these match statistics for every Eredivisie match played, should you be interested. Please note that all matches played from the first until the thirteenth Eredivisie match round of the present 2011/12 season have been used for this analysis, apart from today’s Excelsior – AZ as that match was broken off at half time due to foggy weather conditions.

This mini series will break down team’s performances in offense and defense, focusing on the process of either scoring goals or preventing goals being scored. Both processes will be broken down in three elements, presented as three different parts of this series: creating/preventing goal scoring chances, finding the target and conversion.

Note that Infostrada provides goal scoring chances as goals, saves or off target shots. Blocked attempts are not included in this analysis. This may help to explain potential differences with analyses made using different data sets.


Creating and preventing shots

The table below specifies the average amount of shots taken and conceded per match for each team. The average amount of shots per match was 13.8, with a reasonable difference between teams.

 One team absolutely dominated the Eredivisie in terms of creating chances while preventing their opponents from doing so: PSV tops both the offensive and defensive version of the creation table. On the other side of the table, VVV leaves just De Graafschap behind them with regard to the amount of shots created, while conceding most shots too. Interesting observations concern teams whose offense slightly outperforms their defense, at least in terms of the sheer amount of shots created and allowed (Twente, NAC, Heerenveen and Utrecht) or vice versa (AZ, Heracles, Vitesse, ADO and De Graafschap).

For clarity we could calculate the net amount of shots a team had per match. This would,  of course, underline PSV’s dominance with regard to shot creation in both the offensive and defensive department, having created on average over 10 shots per match more than they have allowed. The reverse is true for VVV, who have allowed on average almost 12 shots per match more than they have created.

A surprisingly low net amount of shots per match of 5.1 expresses that AZ’s leading position in the Eredivisie table does not simply stem from the amount of chances created or conceded, and raises interest in their efficiency in terms of finding the target and converting those shots. Other teams that find themselves lower in the ‘net shots’ table in comparison with the regular league table are Vitesse and Heerenveen, who both occupy comfortable top half positions in the league table.

The prime example of a team with a higher spot in the ‘net shots’ table compared to the regular league table is Ajax, whose 8 net shots per match should be enough for more than their disappointing fourth league spot, where they trail league leaders AZ, who also have a game in hand, by 10 points. This points towards shortcomings in shot accuracy and/or conversion, either offensively or defensively, or both. Another team with a lower league position than their amount of net shots would suggest is N.E.C., who are ranked 15th in the present league table.



Studying the amount of shots created and conceded has severe limitations. Most importantly, the amount of shots does not tell anything about the quality of the chances created or allowed. This type of information should come from planned part 2 and 3 of this mini series, specifying on shot accuracy and conversion. Furthermore, the average amount of shots may not always be the best indicator for performances. An interesting idea for the future would be to look at the median number of shots, rather than the mean, as a few outlying positive of negative match performances will easily disturb observations based on mean numbers, while median numbers are more robust.


In the end

These limitations, however, do not mean that the information is useless. When taking the limitations into account, interesting observations regarding the teams’ performances can be made, as have been detailed in the text. This will be even more true when, in part 2 and 3, the quality of the goal scoring chances will be presented, answering questions like why Ajax’ performance lets them down despite creating an excellent amount of net shots and why AZ comfortable tops the table despite coming in only sixth in terms of the net amount of shots.


This post could never have been created without the support of Infostrada Sports, who’ve provided the data for the analysis.

5 thoughts on “Measuring performances in the Eredivisie – Goal scoring chances

  1. Jop

    Thanks for the great post. Do you now of any data about specific strikers? Topscorers are most likely to play at the top teams, but I guess they get more goalscoring chances as well (even correct for different kind of chances/oppertunities and chances created individually). Who would be the most efficient strikers?

  2. bart

    cheers for the series … couple of mistakes though above … in Heerenveen and Utrecht the net shots are negative and at AZ the net shots are positive (offense out performing defense!)
    cheers and keep ’em coming,

    1. 11tegen11 Post author

      I’m not sure that I catch what you mean here…
      Indeed, Heerenveen’s and Utrecht’s opponents had more shots on target per match than Heerenveen and Utrecht themselves produced.
      And the reverse is true for AZ.

      Whether net shots are positive or negative is a reflection of the combination of (I) having a defense that prevents the opponent from (A) creating shots at all and (B) finding the target with their shots and (II) having an offense that (C) creates shots in the first place and (D) finds the target regularly.

      The fact that AZ’s net shot are positive does not mean that their offense outperformed their defense, just that AZ outperformed their opponents with regard to creating chances and finding the target with them.

      1. bart

        no worries … I was wrong because I found that to be a strange reasoning because Ajax, Excelsior and VVV should also be in that first category and N.E.C. and RKC in the second with the teams Feyenoord, Groningen and Roda JC being equal.

        I just don’t really see the point of these two tables being compared. I understand what you are trying to do but I’m not sure the second column is of any use as the amount of shots does not really reflect the quality of the defense, just the willingness of the teams faced to shoot at your goal … if a team is willing to take lots of shots from a distance a defense will usually say, “no problem go right ahead.”
        There have been various blogs around the net debating just that with respect to ManU, where they have had a crazy amount of shots on goal but their defense has been fairly stout … apparently it came down to the fact that De Gea was seen as a week shot stopper when he moved to the PL but hey, he’s not that bad.
        There are so many ifs ands and buts to this that I would just present the data but not draw any analysis from it … the net shots is interesting as it gives an overview of how the defense and offense of a team are combined … as you mention this is a basis for the next two … part 2 is already a lot more interesting …

        I hope you see this as constructive criticism … it is great to collect a ton of data but it is not always interesting or relevant to analyze each single number … such as in baseball, hits are interesting and fun to keep track of but only get analyzed in correlation with other data and not as raw numbers … I could have 500 hits and player B 200 hits, okay, I have more hits but it says nothing on how player B is better or worse than me, it gets interesting when they get used in BA and OBP and more complicated stuff.

        Anyway, cheers and keep up the good work, you may not think it by this post but I really like your site.


Leave a Reply