1937 Back Pass

Last year, thanks to Jurryt van de Vooren and Simon Gleave, I was introduced to what was thought to be the earliest example of football statistics. The game was played on 20 June 1937.

Simon’s forensic skills have come up with an earlier example from a game played at the Olympic Stadium in Berlin, Germany on Sunday 3 January 1937.

The statistics are from the final of the national Tschammerpokal 1936 competition played between VfB Leipzig and FC Schalke 04. VfB Leipzig won 2v1 with all three goals scored in the first half (1:0 Jacob May (20), 2:0 Herbert Gabriel (31), 2:1 Ernst Kalwitzki (42)). The referee was Egon Zacher who officiated games in Germany from 1935 to 1952.

News of the 1937 statistics appeared on a Deutscher Fussball-Bund’s news page about the 2017 Cup competition and the availability of the DFB-Pokal-App to explore performance data.

The 1937 data:

There is a very short film (18 seconds) of the game:

An English Wikipedia entry has a picture of the match program:

There is a detailed German Wikipedia entry about the 1936 competition. In the match report there is this sentence “Der VfB verzichtete auf jede Effekthascherei und ließ nach englischer Art den Ball laufen”. This appears to be a comparison between VfB’s second half pattern of play and an English approach to playing the game.

Kicker has information about the teams. Of the players listed, ten of the eleven Schalke players have Wikipedia entries:

Hermann Mellage, Hans Bornemann, Otto Schweisfurth, Rudolf Gellesch, Hermann Nattkämper, Otto Tibulski, Ernst Kalwitzki, Fritz Szepan, Ernst Poertgen, Ernst Kuzorra. Ernst Kuzorra was the captain of the team. The only Schalke player without a Wikipedia entry is Ernst Sotnow. The trainer, Hans Schmidt, has a page.

The German Wikipedia page about the game has one link to a VfB Leipzig player, Rudolf Grosse. This is incorrect. It is a link to a sociolinguist with the same name. He would have been 13 at the time of the final. Kicker has a date of birth for Rudolf as 25 August 1910.

Kicker provides dates of birth for 8 of the VfB Leipzig team and all of the FC Schalke 04 team. These data give the median age of VfB as 26.5 years ((range 20 to 30) and Schalke at 24 (range 18 to 31). Bruno Waller, the VfB goalkeeper, won the cup on his birthday.

If the start of the birth year was 1 January for German football, then the places in birth year for both teams at the time of the final were:

I am delighted Simon is on the case. I had better start scanning pre-1937.

Photo Credit

FC Shalke 04 (YouTube frame grab)


The Weltfussball page for the game.

Exploring our performance analysis and sport analytics history

A photograph of one of the 47 shots at goal in the West Europe v Central Europe football game played on 20 june 1937.

Yesterday, I was re-reading Ben Alamar and Vijay Mehrotra’s 2011 discussion of the rapidly evolving world of sports analytics (Part 1).

They have a reference to G R Lindsey’s 1959 paper proposing an operational analysis approach to baseball.

By delightful serendipity, Simon Gleave was tweeting at the same time with news of Jurryt van de Vooren’s discovery of some 1937 association football data.

I do not have any earlier record of notating goal scoring in association football but Simon’s connection with Jurryt has added another piece in our jigsaw about the origins of performance analysis and sport analytics.

We have a record of 47 shots (23 first half, 24 second half) from the game played between West Europe (28 shots) and Central Europe (19 shots).

Ferenc Sas scored two of Central Europe’s goals “with assists from György Sárosi and Italy’s Silvio Piola“. (All three of them played a year later in the World Cup Final.)

A list of the 47 shots at goal in the West europe v Central Europe football game, 20 june 1937.


Another corner piece comes from 24 September 1898. This appears to be the oldest football clip in existence. The film is included in a Daily Mail article from 2014. The teams involved were Blackburn Rovers and West Bromwich Albion.

I am hopeful that our crowdsourcing will produce even more genealogy connections. My earliest baseball discovery is from 1910 and Hugh Fullerton’s account of the inside game.

Our timeline in association football now has some important landmarks:

24 September 1898

20 June 1937

… and 18 March 1950 (Charles Reep at Swindon Town v Bristol Rovers).

Photo Credits

Both pictures shared here are from Jurryt’s 17 December 2016 blog post Werden de eerste voetbalstatistieken al in 1937 gemaakt?



Our Charles Reep Project 2


Last month, I wrote about a Charles Reep project.

I have been doing some research following a lead from Simon Gleave.

The project has been given some urgency by a post by Joe Sykes and Neil Paine.

I was surprised that they thought Charles was responsible for the ruin of English football … and that his maths were to blame (“based on a fatally flawed premise”).

My aim is to address the academic issues surrounding Charles’ work and then to consider his impact on the playing of the game.

I had the privilege of meeting Charles at his home in Torpoint in 1996.

He was 92 when I met him. He seemed much , much younger. He had stayed up late the night before I met him and had completed his 2194th hand notation (46 years of real-time and lapsed-time notation).

I am hopeful that the project will add to the conversation about Charles and contribute to a distinction between the value of his work and value judgements about it.

Photo Credit

Shrewsbury v Rochdale Football Match 1950 (Geoff Charles Collection at the National Library of Wales, no known copyright restrictions)