Do Olympic Hosts Really Win More? Home Advantage Explained

Dakar opens the Youth Olympic Games on 31 October, the first Olympic event hosted in Africa, and Los Angeles 2028 is already on the calendar. Each time, the same claim returns: hosts win more medals. It sounds obvious, but proving it takes more than noticing a host's good year.

Counting medals as proportions

Suppose a country normally wins 40 of 1,000 available medals, or 4%, but wins 65 of 1,000 as host. That is 6.5%. The difference of two proportions calculator gives a standardized gap of about 2.5 and a p-value near 0.012, and the chi-square test for independence on the same two-by-two table agrees. The confidence interval for two proportions runs from roughly 0.5 to 4.5 percentage points, a reminder that the true boost is uncertain.

Compare each host with itself

A single country proves little, so researchers compare many hosts with their own earlier results. Take ten hypothetical hosts whose medal share rose in nine cases, by 1.2 points on average. The Wilcoxon signed-rank calculator for paired data gives p ≈ 0.004, because pairing removes the large baseline differences between big and small sporting nations.

Beware regression to the mean, though. Countries are not chosen at random, and a strong team can have an unusually good or bad year regardless of venue.

Crowds, judges or investment?

Why would hosting help? A study of Summer Games from 1896 to 1996 found significant home advantage in subjectively judged and officiated events, but little in objectively measured ones such as athletics (Balmer, Nevill and Williams, 2003). That points to crowds influencing officials, a theme we explore in our gymnastics judging article. Winter Games analysis found a similar pattern, with greater home advantage in judged events (Balmer et al., 2001).

Judges are only part of the story. Post-war hosts also performed better in the Games before and after hosting, which suggests heavy sporting investment rather than a one-off crowd effect (Nevill, Balmer and Winter, 2009).

A host's medal bump is real in many datasets, but its cause is mixed: officials, familiarity, money and selection all play a part.

Shares, not counts

Raw medal counts mislead because the number of events changes from Games to Games. A country can win more medals simply because more were on offer. Using a share of the medals available, as in our example, removes that drift, which is why researchers favor proportions or odds.

Ten hosts is a small sample

Our paired example uses ten hosts, which is enough to illustrate the method but not to settle the question. With so few pairs, a single odd case can change the result, and testing many sports one by one makes some "significant" findings appear by chance alone. A convincing analysis uses every host, adjusts for national strength, and reports intervals instead of a bare p-value.

When LA 2028 arrives, expect a wave of headlines about a home boost. These tools let you check what they actually show. Check that the analysis uses shares, covers many hosts and reports intervals.

So treat any "home advantage" headline with three questions: advantage compared with what, measured over how many events, and attributable to which mechanism? Counting medals is the easy part. Explaining them is the statistics.