Arsenal News
Arsenal News & Transfers
As featured on NewsNow: Arsenal newsArsenal News 24/7

Arsenal News, Only Arsenal, Blogs, Transfer News

Archives

December 2016
M T W T F S S
« Nov    
 1234
567891011
12131415161718
19202122232425
262728293031  

Why CIES data ranking clubs by the average age of their signings is not all it seems

By Tony Attwood

There are lies, damn lies and statistics. Or so it is said (although usually without any data to back up such a claim).

Ask a statistician however and he/she will tell you that there are statistics used properly and statistics used improperly or with incomplete data.   Football, sadly, is fully of the latter two groups, and we seem to have just been offered another batch of such “analyses”.

The CIES Football Observatory has an impressive name, and appears to present serious data.  Its academic credentials look impeccable working as it does within the International Centre for Sports Studies which is promoted as “an independent study centre located in Neuchâtel, Switzerland.”  It claims a staff of  four full-time permanent researchers who specialise in the statistical analysis of football.

All well and good and presumably highly believable.  But why then does it keep coming up with figures and statistics that I find hard to believe?

Undoubtedly many will think it is because I am a rabid nutter who not only can’t proof read but also can’t add up.  But stay with me for a moment because although such comments might be valid, I think there is something else going on here.

The answer, in terms of the Observatory’s analyses is, I think, that there is nothing wrong with their original data, but there is something amiss with their presentation.  For so keen are they on the headline grabbing stats that sometimes a little of the vital background information gets, well, sort of… lost.

For example they are currently trumpeting an analysis of the age of player recruitment by clubs, and they say that “Real Madrid squad members were on average recruited at the age of 22.5 years,” which is apparently the lowest figure at big 5 league level.   They then go on to tell us that “the best clubs focus their recruitment on younger players than less competitive teams.   In the top five positions of clubs whose transfer policy is most centred  on young footballers are four Champions League participating teams: Real Madrid, Leverkusen, Dortmund and Tottenham.”

So no Arsenal.  I find that a bit surprising.  In fact Arsenal are not even in the top 10, although Bournemouth and Liverpool are.   In fact Arsenal are shown as having an average age of recruitment of 24.07 years, compared with 23.35 years for the Tiny Totts.

That just doesn’t seem right to me from what I know of the two teams.  So time for a bit of digging.

In another analysis presented at the same time by the “Observatory” there is an analysis of the type of recruitment used for players in the squads of teams, shown by leagues.  In the Premier League the numbers are broken down into four groups

  • Internal promotion 8.5%
  • Loan from another club 3.7%
  • Free Transfer 16.6%
  • Paid Transfer 71.2%

What makes that interesting is that for Arsenal, in a squad of 30 first teamers such as Arsenal have (that is the club’s official figure – the 25 players registered as their first team squad who are over 21, and five players who are under 21 who are considered part of the first team), a fair number joined as youth players including…

Player Position DOB
1 Kieran Gibbs* LB 26 Sep 1989
2 Aaron Ramsey* Mid 26 Dec 1990
3 Theo Walcott* Forward 16 Mar 1989
4 Alex Oxlade-Chamberlain* Mid 15 Aug 1993
5 Carl Jenkinson* RB 8 Feb 1992
6 Emiliano Martinez* GK 2 Sep 1992
7 Francis Coquelin* DM 13 May 1991
8 Matt Macey* GK 9 Sep 1994
9 Yaya Sanogo Forward 27 Jan 1993

Plus all the under 21s included in our First Team Squad

Player Position DOB
10 Rob Holding* CB 12 Sep 1995
11 Alex Iwobi* Forward/MF 3 May 1996
12 Hector Bellerin* RB 19 Mar 1995
13 Chuba Akpom* Forward 9 Oct 1995
14 Jeff Reine-Adelaide* Forward 17 Jan 1998

So that makes 14 out of 30 players who were recruited as youth players, or 46% of the current Arsenal squad.  The players with asterisks are also considered “home grown”.

Thus Arsenal has an internal promotion level of 46% compared with the Observatory’s average figure for the Premier League of 8.5% and yet we have an average age of recruitment at over 24 years – putting us in the lower reaches of the league for youth recruitment.  It just doesn’t add up.

In fact it seems quite ludicrous, and the problem is there is no sure fire way to resolve the anomaly because CIS simply don’t provide (at least anywhere I can find it) an analysis of how they are compiling this data.

What I think is happening (but I stress this is a guess because the data is not revealed) is that they are only counting players who are brought into the club as paid transfers.   So a player like Kieran Gibbs who joined Arsenal at the age of 15 is not counted anywhere at all in the chart.  Rob Holding might be counted because he has gone straight into the first team squad, but Hector Bellerin who joined aged 16 is not because he went into the youth team first.   There are of course many such examples of players who, through this sort of approach, are simply not counted.

I stress, that is what I think, and I can’t be sure because CIS is not releasing the information we need to find out what the figures make no sense.  And that is the biggest problem – data without a full explanation of how it was gathered.

The point has some importance when the aaa and their allies in the media start their ranting along the lines of “spend some ******* money” (sorry I know that you know what the word is, and putting a load of asterisks in is a bit pathetic but apparently if I write it, then the news now corporation might well not print a link to this story, and then I get emails asking why we are not on their feed and …. well it’s all a bit of a pain) the whole analysis of how many players Arsenal buy is based on the premise that only players who cost money and go straight into the first team squad, actually count.  Which is a nonsense.

Arsenal are using a totally different approach, of bringing in young players and promoting them through the ranks.  It is a perfectly valid approach – but just doesn’t fit in with the CIS definition of how things should be.

It is a shame that CIS are not giving us clear information about how they do their analysis (or if they are that they are not giving links from the analysis to the research details, so I can find it all).  This type of information is useful – but in its present form it is quite possibly misleading.

Recent tales from Untold 

Wenger ponders whether Yaya Sanogo will ever really be good enough for Arsenal. 

Arsenal to change the line up for Sunderland game: the tactical preview.

Sunderland v Arsenal Saturday October 29th – The Match Officials. Expect serious fouls to go unpunished.

Reasons to be cheerful parts 1, 2, 3, 4, 5, 6, 7, 8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19

Reading in the library: watching the EFL cup from the posh seats

Empty seats at the Ems? You should try the Tax Payers’ Stadium!

Alisher Usmanov makes his move – and it is not the one predicted.

Why 26 October is such a special day in the history of football, and of Arsenal (and the Royal Family)

 

6 comments to Why CIES data ranking clubs by the average age of their signings is not all it seems

  • Yiannis

    Can’t you challenge them by sending them the article? Perhaps they will respond and we will get the answers to your questions

  • Notoverthehill

    Tony, you should read that report in conjunction with Recruitment strategy throughout Europe.

    Quite:

    3. Arrival timeline

    The arrival timeline in a club refers to the semester during which current squad members were recruited. Players promoted from the academy to the first team squad are not included in the analysis. On average, footballers present in the big-5 leagues in October 2016 are in their employer club since 1.9 years.

    The analysis by league and club level again bring to light important gaps. Players from English clubs were recruited since a longer period than footballers from other championships: 2.3 years on average on the 1st October 2016.

    IMHO!

  • OlegYch

    “The analysis of the age of players when they were recruited by the club represented in October 2016 also allows us to highlight important differences in transfer policies. Footballers from the youth academy are not included in the analysis.”
    this seems pretty clear
    thanks for the link though

  • OlegYch

    btw Arsenal has the largest extent of internal recruitment in PL (fwiw) according to their report

  • OlegYch

    the issue of data sources not being publicly available is of course a shame, but it’s mostly because those sources are either commercial (like Opta) or not really credible
    it is pertinent to pretty much any contemporary media (including Untold)

  • Robido

    Thank you for highlighting this source of analysis. It is a gem for datanerds. They identify opta as a source for some data but it may be that their researchers have also had to do a considerable amount of digging. They may share the raw data if you ask nicely.

    As you say data can be presented misleadingly. In any table there has to be a top and bottom and without appropriate description data is just that, it is not information.

    To be fair to CIES they explicitly state that academy graduates are excluded but only when you get behind the headline. It would be interesting to understand why this is the case, possibly as some European teams have such high numbers the average squad recruitment age would be ridiculously low? I would agree however that some reflection of this would be helpful.

    Arsenal are shown as the second highest internal promoters in the premiership and this will affect their position in the table but it is no bad thing. The odd very experienced recruit also pushes up the average which given the position as 12th of 98 (missing the top 10 by .07) means we are still very much in the upper reaches of this table.

    We also show very favourably in continuity and do not have the largess of the bottomless pocketed.

    The full monthly report is good read if for those who are sad enough and like most data driven reports raise many questions around what lies behind the stats. Whilst it mostly reinforces “accepted knowlege” there is the odd surprise.

    http://www.football-observatory.com/IMG/sites/mr/mr18/en/phone/index.html

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>