Investigating the 12th Man Effect in the World Cup
▶ Watch Demo VideoEvery four years, one country hosts the FIFA World Cup. Thirty-two nations play across five weeks. About 1.5 billion people watched the 2022 final. It is the biggest single sporting event on earth. For the host, every match is a home match.
Quatar spent an estimated $220 billion preparing for 2022. South Africa spend $3.5 billion in 2010. Brazil, over $15 billion in 2014. Hosting the world's biggest sporting event is a national-scale financial gamble.
The money is only part of the bet. There's a belief in football that playing at home makes you better at the game.
England hosted in 1966 and won. The only time in their history. Argentina hosted in 1978 and won. France in 1998, same result. South Korea were ranked 22nd in the world when they hosted in 2002. They reached the semifinals, their best finish ever.
The pattern is obvious. The question is whether it holds up once you account for how good the host already was, or whether it's a string of coincidences we've chose to believe in.
In football, home advantage is real, but small. In club football, like the Premier League, home teams win about half the time — maybe a few percentage points above what you'd expect on neutral ground.
International tournaments mostly eliminate this. Games are at neutral sites, neither team has a home crowd, and the baseline win rate is about 44%.
The World Cup is the exception. The host plays every group stage match in its own country while every opponent is a visitor. How large is that edge? And does it hold after you control for how good the host actually was?
Six host nations.
All six went further than anyone predicted. Each one is a data point in the same argument — that hosting the World Cup does something to a team that their talent alone cannot explain.
Click the card to flip through them.
In international matches, teams playing at a neutral venue win about 44% of the time. Draws account for 22%.
This will be our baseline: what you'd expect if location didn't matter at all.
A home crowd changes the odds. With a genuine home side, teams win about 51% of the time — 6 points above the neutral baseline.
Then there's the World Cup.
When a nation hosts, it wins 61% of its matches — a 17-point gap over neutral, nearly triple ordinary home advantage.
But strong teams tend to host.
West Germany in 1974, France in 1998, Brazil in 2014. They were all among the best teams in the world when they hosted. If strong nations tend to get in the tournament, the high win rate could say more about talent than about home advantage.
The Elo rating lets us test this. It tracks a team's strength over time based on all their results, and it gives us a prediction for where each host should have finished. Did they beat that prediction?
Here's every World Cup host nation since 1930, plotted against where their Elo rating predicted they'd finish. The hollow circle is the prediction. The filled circle is where they actually ended. The line between them is the hosting premium.
Four hosts gained nine or more finishing positions above their Elo prediction. Russia in 2018 (ranked 27th, reached the quarterfinal, +19). South Korea in 2002 (ranked 16th, finished 4th, +12). Chile in 1962 (ranked 13th, finished 3rd, +10). Sweden in 1958 (ranked 11th, reached the final, +9).
But the effect isn't universal. Brazil in 2014 was predicted to win but only finished 4th. Japan in 2002 ranked 10th as co-host but exited in the Round of 16. Spain in 1982 and Italy in 1990 both met their Elo prediction exactly - no gain, no loss.
21 of 23 host nations met or exceeded where their Elo predicted. The average gap is 4.35 finishing positions; roughly the difference between a quarter-final exit and winning the whole thing.
Twenty-one of twenty-three host nations met or exceeded what their Elo rating predicted. The average gain is 4.35 finishing positions. Roughly the difference between a quarterfinal exit and winning the whole thing.
The raw win-rate gap (61% vs. 44%) could reflect stronger teams hosting more often. But the Elo-adjusted numbers still hold even after you control for talent.
Why? Crowd pressure on referees is documented across sports. No travel across a six-match, four-week tournament is a genuine physical advantage. And there's something harder to measure about playing for your country in your country. Be it pride, pressure, or whatever, it shows up in the results.
Japan in 2002 and Brazil in 2014 are the two clear exceptions. But if you're a nation deciding whether the football advantage is worth the billions it costs to host, the data says yes.