Before the boom

*The Association of Road Racing Statisticians (ARRS) has made researching marathon performances of the past central to their mission. There are two interlinked initiatives underway at the moment in pursuit of this. The Annual Marathon Rankings Project has now documented the last 70 years.* The end result of this research are annual lists – usually going down as far as the 200th fastest performance - for the years from 1940 to 2009. These are posted on the ARRS website (http://www.arrs.net/YR_Mara.htm). The parallel Yearly Marathons Project, co-ordinated by Ken Young, has now produced well developed draft lists for every known marathon held during the years 1940-2009 (see: http://www.arrs.net/MaraList.htm). Together these two projects take marathon research far beyond what has previously been attempted. The ARRS network of statisticians from around the World have pooled their research and knowledge to create the most complete picture of the event which has come to dominate long distance running in the latter half of the twentieth century and into the twenty-first. So how does marathon running way back in the 1940s compare with the present day? Back then, 70 years ago, the only marathon race held in many countries was the annual national championships. Even where marathons were more common, such events were relatively small – a field of 50 runners was a big marathon. It was much tougher to finish a Marathon then than it is now. The conditions under which runners trained and competed were far less favourable. Light-weight canvas shoes with thin rubber soles offered very little protection against impact. Any runners who were not bio-mechanically sound – for example, if they pronated too much or too little – would very easily sustain knee or ankle problems and be unable to finish an event. High-tech running shoes, designed to minimise problems with foot strike, feet or ankles were completely unknown to those runners. Many Marathon organisers did not provide drinks stations. Where they did, service was very restricted so that hot weather conditions caused major problems. Runners themselves were not fully aware of the importance of hydration. Specific dietary preparation was limited to eating a hearty steak alarmingly close to the start of the race. The virtues of “carbo-loading” were yet to be discovered. Marathon runners were a small, select group. Marathons were usually clustered geographically close to each other – in the United States, for example, most of the events took place on the East coast, and with such a small community of runners it was not uncommon for top runners to tackle three or four marathons a year. Marathon courses were not measured to the same accuracy as today. Measurement by today’s approved calibrated bicycle method was unknown until the 1960s and did not come into widespread use until the 1980s. Despite this, it can be argued that within any particular era the variation between one course and another would be similar. In the 1940s accurate measurement would have been by surveyor's wheel. Arthur Newton measured the Comrades Marathon course in the 1930s by that method. Just as today, courses which were obviously short or aided were flagged up and those events were investigated. This happened in the 1880s, the 1900s and, to judge from the information we have researched, ever since. To some degree, marathon runners are a measuring device in their own right. Each runner knows, and perhaps more importantly their opponents know, the times that each runner can produce in a marathon. A notably faster result raises questions, particularly if numerous runners in the same event have significantly faster than expected times. Using the excel program, I have entered all the marathon data for several years in the 1950s and then sorted according to individual cases. It was something of a surprise to find that there were none of the major variations that might have been expected. Only one race was thrown up as notably “faster” - and that was one which had been queried at the time and reportedly re-measured. What is even more surprising is the way that times by the same runner run on different courses were often very similar. It would seem there was a degree consistency in course measurement. However, Marathons were held in relatively few countries and it was in those countries without a long established tradition of marathon events that issues with short courses later emerged. Using excel allows courses that produce faster performances to be identified. An average performance for a specific runner can be calculated and suspicion will fall on any race in which that runner significantly betters his average (in those days there were almost no women running Marathons). In the 1940s and earlier it was not unusual for a runner to run four or five marathons a year - some ran even more. Some were so consistent they could almost be used as standard candles - to use an astronomical term. A sudden improvement in their time raised questions about the race and its course. The data which has been collected and posted on the ARRS website is an immense resource for those interested in analysing marathon, indeed sporting performance data and how it changes over time. It provides insight into how one of the most interesting grass roots sporting events - the big city marathon - evolved in terms of both performance level and participation. It offers sociologists and economists insight into how such a grass roots sport, not dependent on heavy infrastructure costs, was affected by social and economic factors, both nationally and internationally. The intention is to extend the research back to 1900 and earlier, and excellent progress is already being made towards achieving this goal. The research is ongoing and earlier years will be added to the ARRS website as soon as they can be researched and processed. In addition to the Marathons projects, the ARRS website also documents hundreds of long held marathons, road, and track and cross country races, including national and international championships. For further information see: www.arrs.net