World Cup Qualifying Math

TL;DR: Here's some math on World Cup Qualifying, who deserves to be there, and maybe some hints for the predictors out there.

Preface

This started with a simple grievance: Italy. They had a tough group stage qualifying campaign, lost a single-leg playoff final on penalties, and are staying home while teams with worse underlying records are going to the World Cup. It felt like the format was doing something unfair — but to actually test that, you have to be able to compare records across confederations, and that means accounting for schedule strength. Which meant building the whole thing from scratch.

So with the help of AI, that's what I did.

**What I ended up with**

- 908 qualifying matches across all 6 confederations + intercontinental playoffs, with scores

- Full W/D/L records for all nations who participated in qualifying

- Strength of Schedule (SOS) for every team, using ELO-based win probability

- Performance vs Expected analysis for all 45 qualified teams

**The methodology in brief**

For Strength of Schedule, I used the median ELO of the 48 qualified teams (1775) as a reference point — so the question is "how hard would this schedule be for a typical World Cup team" rather than against some arbitrary average. Lower SOS score = harder path.

For Performance vs Expected, I built two metrics:

1. **Δ ELO** — actual points % minus what the team's own ELO predicts they should earn. Did they beat their own benchmark?
2. **Δ Schedule** — actual points % minus what a median World Cup team (ELO 1775) would earn on the same opponents. Did they outperform a WC-calibre baseline?

I also filtered to only count games against top-100 ELO opponents (ELO ≥ 1400) — so early-round blowouts against minnows don't inflate anyone's numbers.

**A few findings that stood out**

- Norway is the most obvious overperformer — perfect record, +33% above their own ELO prediction, against genuinely decent Group I opposition

- Panama qualified but massively underperformed their ELO (-29%). One of the more worrying numbers for a team heading to the tournament

- Argentina, Brazil, and Colombia all show negative Δ ELO — they underperformed their own ratings. But their Δ Schedule is strongly positive because CONMEBOL is genuinely brutal. They were stress-tested in a way most qualifiers weren't

- Nigeria's numbers are striking — -28% Δ ELO with a reasonable sample. They didn't just miss on bad luck

- Algeria has a ⚠ flag — only 2 qualifying games vs top-100 opponents after filtering, so their numbers are tough to really dig into. A clear indicator there is likely a problem in African Qualifying.


** Sorry Canada, USA, & Mexico - you didn't play any qualifying games.

**On the AI disclosure**

I used Claude (Anthropic) extensively to help compile and cross-verify the match data, build the ELO calculations, and construct the spreadsheet. The methodology decisions — what to measure, how to define the reference point, what to filter — were mine, but the heavy data lifting was AI-assisted. I think that's worth being transparent about, especially for something where the data integrity matters. The match counts were validated against Wikipedia confederation H2H matrices and known format specs, but I'd encourage anyone who spots errors to please flag them. I did not do an audit and am trusting the data.