Why taking up the PSV or Groningen vacancy is a good idea, and the Heerenveen job is not…

With the regular matches of the 2011-12 Eredivisie season over, and only the promotion/relegation play-offs and the Europa League qualifier play-offs still in contention, several teams are either appointing or releasing managers right now. Groningen fired Pieter Huistra, whose contract they renewed as recently as during the past winter break. Veteran manager Dick Advocaat took up the PSV job, after Philip Cocu indeed proved to be just an interim solution for the job vacancy after Rutten quit. Ron Jans announced his departure from Heerenveen earlier this season, with Marco van Basten returning to a managing job here.

This post will outline why the first two jobs, at Groningen and PSV, are excellent opportunities, while the Heerenveen job is a pitfall. The same parameter that was introduced recently to differentiate between sustainable and unsustainable performance at club level, PDO, is used again here. For a full description of PDO, read the introductory post, written a few weeks ago.

 

PDO from season to season

The key concept for this post is the fact that PDO has an enormous influence on a team’s performance throughout a single match and also over the course of a single season. This becomes clear when we look at two teams with very different PDO’s in the 2010-11 and the 2011-12 season. However, PDO fluctuates a lot from match to match and from season to season, as has been demonstrated before by James Grayson on his excellent blog in a much larger data set than the two Eredivisie season that I have available here.

By now it’s very much clear that last year’s over-performers, Groningen had a terrible run this year, leading to the sacking of their manager, Pieter Huistra. Groningen’s PDO dropped dramatically, coming from 1045 and 2nd best in the league and finishing the 2011-12 season at a dramatic PDO of 936, the worst in the league at some distance. The best example of the reverse trend is Feyenoord. Their PDO last year was 988, not dramatic, but still indicating that the team had more quality than their 10th place in the 2010-11 league table represented. Feyenoord’s 2011-12 PDO is 49 points higher at 1037 and they finished the season in a much improved 2nd spot in the table.

PDO data from all clubs over the past two seasons are presented in the next table. Note that this table only contains 17 clubs, as there was one promotion/relegation in between these seasons.

PDO

2010-11 2011-12
Ajax

1031

1026

AZ Alkmaar

995

1026

Den Haag

1050

976

Excelsior

977

968

Feyenoord

988

1037

Graafschap

974

963

Groningen

1045

936

Heerenveen

1019

1057

Heracles

997

977

NAC Breda

1018

996

Nijmegen

1027

980

PSV Eindhoven

1019

978

Roda

1042

1019

Twente

1014

1038

Utrecht

1013

1010

Vitesse

970

1005

VVV Venlo

935

982

 

A quick mind will have noted that there is no correlation between both seasons. In other words, a high (or low) PDO in one year indicated nothing about the level of PDO in the next year. This is well illustrated in the next graph, depicting both seasons in a scatter plot. Note the flat trend line with a near-zero correlation coefficient.

 

 Now, what does this mean? Coming off a low-PDO season, things can only get better at clubs like PSV and Groningen, while teams like Feyenoord, Twente and particularly Heerenveen, who come off extremely positive PDO’s are in for a disappointing year.

Good luck, Marco!

 

Data: Infostrada Sports

3 thoughts on “Why taking up the PSV or Groningen vacancy is a good idea, and the Heerenveen job is not…

  1. Bas

    Isn’t the N too small to say legitimate answers?
    PDO fluctuates a lot every match, which would lead to different outcomes each season. Couldn’t you use the data of the last 10 seasons? Nice article nontheless!

    Reply
    1. admin Post author

      Completely agree with you… For the Eredivisie I presently only have these small numbers, but the same principles have been tested in a big data set of 702 data pionts by James Grayson on his excellent blog. This already showed team PDO to have a marginally low correlation from season to season… I’ve added a small comment to explain this further to the full body of my article now, for clarity.

      Reply
  2. sean

    A bit late to the party, and not at all related to this post, but I’ve been meaning to post for a while that your link in the Landmark Articles section to Swiss Ramble is broken.

    Reply

Leave a Reply