22 completed World Cups of Elo, what past rankings tell us about future favorites.
I miss FiveThirtyEight. As an homage to one of sports and politics greatest websites, I started a small analytics project for this year's world cup. For the last Women's Euros I built an Elo-based model for a family wager. It was pretty basic but performed surprisingly well. With the 2026 World Cup around the corner, I wanted to revisit that model and dig into 22 editions of men's tournament history.
First question: does the best team actually win?
Usually, yes. Sixteen of 22 World Cups were won by teams ranked in the top 5 of that tournament's Elo field. Elo adjusts for opponent strength and margin of victory, a running tally of footballing quality that goes back nearly a century. The first chart shows how individual teams fared in each tournament.
If we zoom out, a bigger question emerges: does coming in as the favorite actually matter? Conventional wisdom says the best team wins. But football has a way of humbling even the mightiest. Let's look at where World Cup winners actually stood within the qualified Elo field when the tournament began.
Loading winner rank summary...
Winning a World Cup takes more than just showing up as the favorite. It requires performing when it matters most. But some nations have mastered the art of consistency: showing up tournament after tournament with roughly the same level of quality.
Which nations are the steady hands of international football? And which ones ride an emotional rollercoaster from one World Cup to the next?
How to read: Lower standard deviation (σ) = a steadier pre-tournament Elo profile across appearances. The bar shows the Elo range (min to max), with the dot marking the average. Tighter bars indicate nations whose ratings varied less from one completed World Cup to the next.
Consistency is admirable, but football's greatest tragedies belong to the teams that came so close: the nations with the talent, the rankings, and the moments of brilliance, yet somehow never lifted the trophy.
These weren't plucky upstarts. They were elite squads who mistimed their peaks.
Where does this leave us? Let's review what the data told us:
As a starting point for further analysis, we converted the 2026 field into a simple Elo-only prior.
This is not a bracket simulation. It converts each 2026 field Elo score into a strength value, then normalizes across all qualified teams before draw path, injuries, travel, and matchups.
| Rank | Team | Elo | Elo-only prior |
|---|---|---|---|
| Loading Elo prior... | |||
Now don't take these priors as win-probabilities just yet. We'll get to that in the next post.
These numbers are a starting point which we'll use to build a more sophisticated model. As part of that we already started gathering some additional data about our contenders and their performances throughout the years. Looking forward to sharing more soon! (If you are part of my family, stop reading here. There is nothing interesting to see.)