Presentation on theme: "Can statistics demonstrate a home advantage in the Olympic games? Stephen R Clarke."— Presentation transcript:
Can statistics demonstrate a home advantage in the Olympic games? Stephen R Clarke
Home advantage in the Olympic Games This presentation is based on the paper: Clarke, S. R. (2000). Home advantange in the Olympic games. Proceedings of the Fifth Australian conference on Mathematics and Computers in Sport. G. Cohen and T. Langtry. Sydney, University of Technology Sydney : 76-85.
Home Advantage Simple study done prior to Sydney Olympics Uses only Secondary School descriptive statistics The results of the study gained widespread publicity with several newspaper articles, 2 television segments and almost 20 radio interviews
Home advantage Investigate the extent of home advantage in the Olympics All countries – The number and type of medals won – The percentage of available medals won Host countries – Comparison of home and away performance – comparison of home with expected performance – mix of medals
History of the Ancient Olympics The ancient games began about 824 BC initially consisted of a single foot race from one end of the stadium to the other they accumulated sports such as discus, javelin, boxing and wrestling continued for 12 centuries. from 776BC they were held every four years. abolished in 394 AD.
The Modern Olympics The Modern Olympics began in 1896 It now vies with the World Cup of soccer as the World’s premier sporting event.. The first modern games did not allow women. They now make up 30% of the athletes. The event list has steadily grown An increasing number of medals awarded. – In 1896, 122 medals including 44 gold – In 2000 928 medals including 301 gold.
Table 1: Venues for the Modern Olympic Games. 80928
Past studies Stefani - improvement in the various sports. Condon et al - neural networks and regression to model the number of medals won at the Atlanta, based on various economic variables such as area, population and length of rail road track. Sommers measures the success of nations at Atlanta per unit of population Leonard The "Home Advantage": The Case of the Modern Olympiads
Home Advantage The existence of HA in sport is well documented Home teams win majority of the matches (~60%). Causes of HA are – positive effects for the home side due to ground familiarity and a partisan crowd. – negative effects on the visitors mainly due to travel. Such effects are present during the Olympics- possibly to a greater extent for Australia
Home Advantage Visitors have change in season and time zones Home crowd Stefani shows that in soccer HA effects increase with international travel The home country has some choice in the sports that will be offered larger teams and competes in a larger range of events than usual. boycotts reduce the strength of competition.
Home Advantage Usual causes of HA clearly apply to all athletes from the host country Research Question Does the home team have an advantage in the Olympic games If so, can we quantify it
Home advantage In any analysis there are many statistical traps that await the careless – The increasing number of events – Increasing number of countries – Changes in the general level of sports performance of particular countries – Boycotts Need to ensure results cannot be attributed to other causes
Method Compare the home performance of host countries with their away performances. Simple study, uses school mathematics, yet results achieved widespread publicity Analysis done on SAS/JMP, but Excel could have been used.
Data The final gold silver and bronze medal tallies of all countries that won a medal. – Wallechinsky has details for all events up to 1984 - including 1906 – Later Olympics from the WWWWWW – Collected by undergraduate students - 841 observations. There were no data on the countries that competed but won no medals. (Latest edition includes Sydney)
Measure of Performance Number of Gold? Number of Medals?
Total number of medals won against number of gold medals won for Australia
Measure of performance Less % variation in a total than individual items – Use total medals rather than gold Total number of medals awarded has steadily increased – Use % of total medals won as the measure of a countries performance
Percentage of total medals won by the USA Outlier??
Original Conclusion Large random element in the performance of countries in the Olympic games. Percentage of available medals won is a better measure of a performance than the number of gold The home team wins more medals than expected on adjacent games performance (1.5 times?) The home team usually wins a richer mix of medals (more gold) than when away.
But?? Is HA decreasing with time? Does it depend on other factors – the hemisphere of the host nation. – success rate of country – team or individual sport – the actual event Press only wanted prediction of Australian medal tally
How many medals for Australia in Sydney? Science becomes guesswork – At its last home Olympics, Australia gained 7.6% of available medals, twice the percentage they achieved in the games immediately before and after Melbourne. – However the isolation of Melbourne in 1956 resulted in a low number of athletes attending, and the games were also weakened by two boycotts – In Sydney Australia was coming off a strong performance in Atlanta of nearly 5% so to expect a repeat performance of their last home games effort might be optimistic.
How many medals for Australia in Sydney? There is also strong evidence that the mix of medals is richer for the home teams, so Australia can expect to win a proportion of the gold medals greater than both their long term average of 30% and the actual percentage available at Sydney. In making predictions there is always the effect of randomness. Canada, a country similar to Australia, won no gold medals at the Montreal games.
Prediction 5% of 910 medals = 46 7% of 910 medals = 64 9% of 910 medals = 82 (record 41 in Atlanta) 33% of 46 = 15 gold 33% of 64 = 21 gold (record 13 in Melbourne)
Conclusion Simple statistical analyses can provide reasonable estimates. Many other variables could be taken into account in predicting medal tallies. Randomness means accuracy as much a result of good luck as good planning
Conclusion Athens? Beijing? Over to you, or your students
Statistics can demonstrate a Home advantage in the Olympic games Stephen R Clarke