Wednesday, 8 August 2012

What is home advantage worth at the Olympics?

Regular readers will know that I am not the biggest fan of the Olympics, but I am keen on statistics and so I wondered how much of an advantage hosting the event provides in terms of medals.

I took the average number of medals won by each host in the years before their Olympics and compared this to the number won when they were the host. Thanks to the cold war boycotting of the 1980 and 1984 games, the USA and South Korea were excluded because they missed a preceding Olympics and the USSR is excluded because its success at the 1980 Moscow games was at least partly due to the absence of the other superpower. West Germany was excluded because it did not compete under that name until 1968.

Here are the results. Click to enlarge. Each nation's home performance is shown in the middle. Figures to the left show preceding Olympic performances going back up to seven Olympics.

As you can see, every host nation wins significantly more medals when they are playing at home. Plausible explanations include increased investment in athletics prior to the the games, favourable climatic conditions, psychological advantages from crowd support and judges being influenced by partisan crowds.

The win rate increases from 41% (Australia, 2000) to 756% (Spain, 1992). This, of course, is a large margin, but the average increase in medal tallies is 222%. Therefore, if 'Team GB' benefits from an average home advantage, we would expect them to win 64 medals this year.


Jonathan Bagley said...

I saw a similar graph recently. I remember thinking that all benefitted except the USA with the Atlanta Olympics and tried, without success, to figure out why. On further investigation, that turns out to be the case. see here, where 59 is the prediction.

Snowdon said...

Good link. Thanks.

Kairosis said...

Interesting Data.

It looks like there may be a significant increase in medal counts preceding the 'home game'? One that gradually diminishes in the years post home game.

This seems to suggest that the increase in investment provides the most plausible explanation for variations in medal counts?

The post home game decrease in medal counts might mean that if the meteoric increase in elite sports investment is not sustained following the home game the cost of an Olympic medal will show a similar meteoric trajectory!

Certainly the Australian Olympic medals are getting rather expensive:

Some Australians such as myself are now thinking that it maybe time to investigate more cost effective Olympic formats:

Love your work, especially your tobacco industry blogfest.

BriarTuck said...

Nice piece of work -- and as it turns out, pretty bloody accurate. Once upon a time, one used to get stuff like this in newspapers.

Christopher Snowdon said...

They won 65 medals in the end, so this sketchy system was only one out.