Charles Reep: How data analysis kicked off in the 1950s

Post-war Britain was a time of austerity. Rationing was still in effect, so everything had to be done efficiently and with minimum waste. The public craved entertainment and football attendances rose as the sport delivered. Attacking football was never more in vogue and goal scoring levels were the highest since the offside law had been changed in the 1920s.

Football took its first steps towards the game we know today. The use of floodlights to allow evening games, the spread of television coverage, European competitions starting up and the expansion of domestic leagues pushed the game in new directions.

Amid the growth in interest, England came to the realisation they weren’t the global leaders of the game, losing at Wembley for the first time to continental opposition and being humbled in the World Cup. Things had to change and, ever so slowly, they began to.

Embed from Getty Images

The FA began tracking data in the 1950s, but like many new ideas, they didn’t know how to use the results to the fullest advantage. Likewise, in the Netherlands Lou Hofland researched how the game ebbed and flowed. He used a converted weather barometer to track the movement of the ball up and down the pitch during a match. Twenty years later, the Premier League website used the modern-day version of this as a graphic to illustrate which team was exerting the most pressure on their opponents during the course of a game.

Charles Reep was the key British character in the post-war explosion of football analysis. Within the confines of technology available at the time – no GPS tracking, no tablets or phones to collect data, not even video, or cinefilm playback for the majority of games – the 1950s saw the field of football analytics expand to cover more ideas than at any time until the 2010s.

Reep’s data work with RAF teams came to the attention of Brentford, a team who use analytics to inform their decisions in the 21st century and he was invited to give a presentation to manager Jackie Gibbons. By then, Reep had already advanced his work through several levels. He had distilled the game down in his mind to ask the simplest questions – how many goals do you need to be successful and how can you give yourself the best chance to score that many goals?

He reckoned that, at the time, a team needed an average of three goals per match to win enough games to finish in a promotion place or win a league. He looked at how many times a team had possession in a game, how many shots were needed to bring a goal and where those shots should be taken from. He further expanded this to look at the length of passing sequences that brought about goals and where on the pitch these scoring moves started.

Reep suggested that winning the ball near an opponent’s goal led to more goals than a complex passing move starting in your own half. The advice was to get the ball forwards quickly and if you lose it, press hard to win it back swiftly. Not that different to the high-intensity tactics employed by many top teams today.

Embed from Getty Images

Brentford shot up the league table and their goals per game rocketed. Due to a change in his RAF posting, Reep ended up in the Midlands and worked with Wolves manager Stan Cullis as his side became the team of the decade.

Reep was all about efficiency – create as many chances as possible without making things too complex. A match that he documented and analysed in great detail was the ‘Match of the Century’ between England and Hungary at Wembley in 1953. The visitors were Olympic champions and in the midst of a long unbeaten streak, full of technically superior players they won 6-3.

As Reep pointed out, four of the Magyars goals came from winning the ball back near England’s goal. While disappointed with the result, he felt vindicated with his analysis, but unknown to him, Hungarian officials had been analysing football, especially optimal shooting positions for more than 20 years. Hungarian coach Gustav Sebes kept detailed notebooks of formations and tactics while the state government held files of match and training data designed to help the team reach peak performance.

Some fundamentals of football don’t seem to change with time. In the 21st century, an analysis of goals at the World Cup showed that more than 70% of all goals come within 20 seconds of winning the ball.

The FA were also trying to find new insights into football during the 1950s. By the end of the decade they sometimes had tens of analysts at games, furiously noting down data with pencil and notepad.

At the same England-Hungary game in 1953, the FA took note of the amount of running done by the teams. Using landmarks around the pitch and the markings on the playing surface as guides, they calculated how far players had run and the number of sprints that players made during the match.

The FA looked at a direct comparison between the movement of Ferenc Puskas and Jackie Sewell. They found that while Puskas covered only 90% of the distance completed by Sewell, he made more than twice as many sprints which covered almost twice the distance that Sewell sprinted.

Puskas’s 200 sprints during the match still look impressive when compared to the average modern performance in the Premier League, where most strikers make between 50 and 80 sprints in a match.

Embed from Getty Images

The modern-day player moves much more overall though, with 9,000-10,000 metres per match covered for an average striker. In 1953, Puskas covered 5,966m and Sewell 6,562m.

In his book, Soccer Coaching, Walter Winterbottom used data to cover a whole chapter on analysing play. One of the details he focused on was the amount, and type, of movement players made during a match.

It highlighted, as expected, a large difference between players in differing positions and compared to today, they didn’t move about as much. An average centre half would only run around 475 metres in a game, a winger would cover 700 metres running and a further 3,000 metres walking. An average midfielder would cover around half the distance of today’s players.

Winterbottom was also interested in where the players ran and commissioned a number of charts that mapped the zigzag movements of players across the pitch in timed intervals. Today these can be compared to heat maps that are produced by computers using GPS tracking.

Other areas were also analysed. The English national team has been haunted by failure in penalty shoot-outs, but the FA were analysing these decades ago. Goalkeepers’ dives were measured to see which parts of the goal could not be reached and shooters’ kicks were timed to see how fast a shot would reach different target areas. It’s not known how this data was used – or not as the case may be.

Other FA analysis included more general observations of matches: passing sequences, interceptions, dribbles and crosses. But it is the work done by Reep that still resonates. He preached that most goals were scored with close-range shots predating xG by 60 years and in fact his data was used by Richard Pollard in the 1990s. Pollard published work that analysed angle of shot, type of shot, distance from goal and so on. This led directly to xG as we know it now.

Many other things that Reep started to address in the 1950s and early 60s was only held back by the lack of technology available to him at the time. He worked on ideas that today are called Field Tilt (which team has the most of the ball in their opponents’ final third), he looked at goalkeeper distribution, where attacks started and where they most often broke down – and with some help from fellow researchers, he began looking at individual player metrics, all using the human eye and a notepad.

Reep, among others, never got the praise he deserved and many of the ‘new’ metrics we see today actually had their beginnings 70 years ago.

Rob Haywood is the author of the forthcoming book Many Impossible Things: The Ingenious Evolution of Football Data

Charles Reep: How data analysis kicked off in the 1950s
5 (100%) 2 votes