When England kicked-off against Scotland at Wembley in April 1961 there were over 97,000 fans in attendance, including dozens of men taking notes with pads and pencils. Among these observers were Wally Barnes (ex-Wales manager), Ted Fenton (manager of West Ham), Jack Mansell (the ex-Portsmouth player who later coached overseas), Phil Woosnam (then a West Ham player) and Alan Dicks (ex-Chelsea player who would go on to have a long managerial career). This gathering was part of the FA’s decade-long quest to gather as much data from the game as possible, and this was probably the pinnacle.
Soon after the Second World War, the FA began to take a limited interest in football data. Initially it was fairly bland: investigations into the percentage of games won by the home team and average league attendances plotted against the weather (surprise: more people attended when it wasn’t raining). FA publications at the time often displayed a ‘Pictorial Summary’ to illustrate these findings. By the early 1950s more actual on-field information was being collected including the first examples of position-specific data, though few shocks were forthcoming: England’s centre-forwards were found to score more goals than the inside-forwards who in turn scored more than the wingers, but things were progressing.
Solid arguments were put forward regarding the information that could be gleaned and, more importantly, used by coaches. However, some breakthroughs while cutting edge at the time seem frustratingly vague by today’s standards. One report mentioned “twenty international and football league matches” were used in a study. But which ones were included and why the mix? A survey of shooting broke the shots down into ‘goals’ and ‘unsuccessful’, but did the latter go wide, get blocked or were saved? The report concluded that, on average, nine shots were required to produce a goal - a ratio that translates to a modern day ‘expected goals’ figure of 0.11, a rate that is still broadly held as the standard today.
Early defensive measures were also being noted with percentage of tackles won on the ground and in the air, while crude measurements were made to link pass lengths and their success (or otherwise). However, just as the use of data in football has had its share of naysayers in recent years, similar arguments were being fuelled over 60 years ago.
Walter Winterbottom, as the head of FA Coaching, knew of these studies but he faced reluctance from professional players to listen to anything new. Undaunted, he dedicated a chapter of his book, Soccer Coaching, to “Analysing Play”. Using language that would gladden any current analyst he explained that he wanted to ‘study each separate movement’ while ‘analysing play and measuring the value of importance’. Perhaps most importantly he tried to collect data over a series of matches and use the information to predict future success. He refused to write-off any particular style of play, making note of the pros and cons of the long ball game as well as playing-out short from the back. In that analogue-age he also tried to make sense of the information as a whole and how the individual passages of play linked together to form a complete story of how a match unfolded.
תגובות