2026 Winter Olympics Data Analysis: What AI Found in the Milano-Cortina Medal Table
Brazil won alpine skiing gold. One man swept every men’s cross-country event. And the Netherlands — a famously flat country — finished third in the medal table.
The Milano-Cortina 2026 Winter Olympics delivered surprises worth digging into. So we did what any reasonable data nerd would do: we pointed PlyDB and Claude Code at the raw Olympic datasets and asked questions in plain English.
The results were more fun than we expected.
Who is this for? Sports fans who want more than a medal table, data enthusiasts curious what AI-powered sports analytics looks like in practice, and anyone who wants to ask their own questions about Olympic data without writing code.
What is PlyDB, in one sentence?
PlyDB is an open-source tool that lets AI agents query your data directly — no pipelines, no data wrangling, no ETL. You point it at your files (CSV, Parquet, Excel, databases, whatever), and your AI can ask SQL questions against them in real time.
For this project, we connected PlyDB to five Kaggle datasets: athletes, coaches, medals, schedules, venues, plus World Bank GDP data and country latitude coordinates. Then we just… started chatting.
Finding 1: Norway Is in a Different League
Norway sent 80 athletes to Milano-Cortina and came home with 41 medals — 18 of them gold. That’s a 51% medal rate. One in two Norwegian athletes stood on a podium.
Johannes Hoesflot Klaebo alone won 6 gold medals — every single men’s cross-country skiing event. All six. The man did not lose once.
Norway also swept Nordic Combined entirely (3 events, 3 golds, all Norway) and won 7 of 12 cross-country skiing golds. The other 5 cross-country golds? Sweden. Together they shared all 12 events — no other country won a single cross-country gold.
Finding 2: The Netherlands Is Quietly Absurd
The Netherlands sent 39 athletes and won 10 gold medals.
Ten. Golds. From thirty-nine people.
That’s 25.6 golds per 100 athletes — the best conversion rate of any nation. Their secret was total domination of the ice oval: 5 golds in Short Track Speed Skating, 5 more in Speed Skating. Jens van ’t Wout alone won 3 golds.
The Netherlands is also famously flat. They just decided that going very fast in circles on ice was their thing, and apparently committed.
Finding 3: France Owned Biathlon
France won 6 of 11 biathlon gold events. In a single discipline. At a single Games.
Biathlon combines cross-country skiing with precision rifle shooting, which sounds like it should be wide open. France disagreed. France’s biathlon haul alone — 6 golds — would rank them 8th on the overall gold table as a standalone country.
Quentin Fillon Maillet and Julia Simon each won 4 medals (3 gold). The French biathlon team walked away from Milano-Cortina looking like they run a very specific finishing school.
Finding 4: Brazil Won Alpine Skiing Gold
Lucas Pinheiro Braathen — a Brazilian alpine skier — won the Men’s Giant Slalom.
Let that sink in. Brazil, which sits at 14°S latitude and has essentially no alpine skiing infrastructure, produced the Olympic champion in one of winter sports’ most traditional events. It was the only Latin American gold medal of the entire Winter Games.
From a pure data perspective: Brazil’s $2.19 trillion economy produced 1 medal. Norway’s $483 billion economy produced 41. By the arithmetic of GDP-per-medal, Norway’s entire economy is smaller than the cost of a single Chinese medal ($18.3T / 15 = $1.22T each). But Brazil’s single alpine skiing gold carries the weight of an entire continental economy behind it — which makes it arguably the most remarkable result of the Games.
The Question: If You’re Starting From Scratch, Which Sport Should You Choose?
We asked Claude to figure out which Olympic discipline gives an athlete the best statistical odds of winning a medal — assuming they’ve already qualified for the Games.
The answer required three separate analyses: medal rate (what fraction of athletes in each sport actually medal), country concentration (does one nation sweep everything), and field size (fewer competitors = better odds).
The winner: Short Track Speed Skating
- 33.9% medal rate — the highest of any discipline
- Gold spread across 4 countries (genuinely competitive)
- Relay formats let one athlete stack medals across multiple events
- Only 112 athletes in the field
Short Track is the rare case where the headline number holds up: the medal rate is high, the competition isn’t monopolised by two countries, and the relay structure multiplies your opportunities.
The dark horse: Ski Mountaineering
It debuted at Milano-Cortina 2026. Only 36 athletes competed. The competitive field is still forming. Get in early.
What to avoid
Nordic Combined: Norway won all 3 gold events. Every single one. Unless you’re Norwegian, your gold odds are mathematically near zero.
Alpine Skiing: The most popular winter sport has the worst medal rate of any discipline — just 9.2%. That’s 306 athletes competing for 28 medals. Expensive sport to train for, brutal to medal in.
Ice Hockey: 530 athletes, but gold goes to 23–25 players from a single country per gender tournament. The pool looks big; the actual odds aren’t.
Medal rate = unique medalists / registered athletes. Assumes qualification for the Games. Source: Milano-Cortina 2026 · Analysis via PlyDB + Claude Code
The GDP vs. Medals Question
One more analysis: does money buy winter Olympic medals?
The short answer is no. Or at least, not very efficiently.
Norway — whose entire GDP is roughly the size of the state of Washington — produced 41 medals. China — the world’s second-largest economy — produced 15. Norway is 71x more efficient per dollar of GDP than the United States, and 103x more efficient than China.
After controlling for latitude (northern countries have a natural advantage — it’s hard to build a cross-country skiing culture without snow), the real story emerged. Georgia and Bulgaria, both sitting at roughly the same latitude as Rome, 15x and 9x outperformed their geographic peers. Latvia — population 1.8 million, GDP $43 billion — won 2 medals in bobsled and luge.
Great Britain, at the same latitude as Scandinavia, scored a latitude-adjusted index of 0.11 — the lowest of any country in the dataset. Britain’s sporting culture has long pointed toward summer: athletics, rowing, cycling, tennis. Every winter medal British athletes win is a genuine act of going against that grain.
The conclusion across all three analyses: latitude sets the ceiling; culture determines how close you get to it. You can’t buy a ski jump in your DNA.
Try It Yourself
The full repository includes:
- All five Olympic datasets (athletes, coaches, medals, schedules, venues)
- GDP data from the World Bank
- Country latitude data
- A pre-configured
plydb-config.jsonso you can start querying immediately - Three full analysis writeups from Claude
To run it yourself:
- Clone the repo
- Follow the Quickstart Guide to install PlyDB and connect your AI agent
- Tell your AI to use the
plydb-config.jsonin the repo and start asking questions
The whole setup takes a few minutes. The rabbit holes take considerably longer.
Frequently Asked Questions
What data does the Olympics analysis use? Five Kaggle datasets covering athletes, coaches, medals, schedules, and venues from the 2026 Winter Olympics, joined to World Bank GDP data and country latitude coordinates. All datasets are pre-normalized and ready to query.
Do I need to know SQL to explore this data? No. PlyDB lets your AI agent handle all the SQL. You ask questions in plain English — the agent translates them into queries and returns the answer.
What is PlyDB? PlyDB is an open-source tool that gives AI agents unified SQL access to local data files — CSV, Parquet, Excel, databases, and more. No cloud required. See the PlyDB quickstart to get up and running.
Is there a similar project for other sports or events? Yes — check out F1 Analyst for Formula 1 telemetry data, Baseball Analyst for MLB Statcast and FanGraphs data, and Oscars Analyst for 97 years of Academy Awards history.
Did you know? PlyDB can connect your AI to boring data too!
Whether it’s business data in a dusty Excel sheet or a complex DevOps log in S3, AI can be surprisingly good at making sense of a mess. PlyDB acts as the bridge, letting your AI query across Postgres, MySQL, CSV, Excel, Parquet, Google Sheets, and more - locally or in the cloud.
Open source and free. Give PlyDB a spin!