Where Our Data Comes From

The Seasoned Score, the leaderboards, and the resort finder quiz are only as good as the numbers underneath them. This is where those numbers come from.

The data pipeline

Every resort stat goes through a four-stage pipeline before it appears on the site:

1. Collection — automated scripts pull data from resort websites, official tourism boards, government sources, and independent aggregators.
2. Normalisation — raw values are standardised (currencies converted to USD, distances to km, areas to km²) so they're comparable across countries.
3. Conflict resolution — when sources disagree on a value, a human reviews them and picks the most credible figure. The chosen source is logged alongside the value.
4. Scoring — the normalised values feed the percentile calculations that produce scores and rankings.

What each stat type comes from

Terrain and mountain stats (skiable area, vertical drop, lift count, piste breakdown) come primarily from resort-published figures, cross-checked against skiresort.info and Ski Club of Great Britain data. Skiable area is particularly prone to inconsistency — resorts measure it differently — so we note where figures differ significantly from independent estimates.

Season length is calculated from resort-published open/close dates for a recent typical season. We use the full season (not the peak weeks), because a seasonaire's earning window is the whole run.

Snowfall figures are historical averages from meteorological data, not resort marketing claims. We use multi-year averages where available. Single-year anomalies (a drought year, a record winter) are smoothed out.

Cost of living — rent and groceries — is the hardest data to get right. Accommodation costs come from a mix of resort staff housing pages, seasonaire forums and community data, and local listing aggregators. Grocery prices use the Numbeo database as a baseline, checked against local supermarket data where accessible. Both are stored in local currency and converted to USD for scoring using ECB-sourced exchange rates from frankfurter.app. Exchange rates are timestamped so you can see how old a conversion is.

Beer price is a single-pint proxy for overall resort pricing culture, collected from resort bar menus and Numbeo. It's a deliberate shortcut — beer prices are easy to collect and reliably track broader cost levels.

Nightlife is rated on a 1–5 scale based on a combination of publicly-available information (venue density, resort reputation in the seasonaire community, published reviews) and AI-assisted reputation scoring. It's the most subjective stat on the site, and the rating notes this where relevant.

Gnarliness is an 80/20 composite: 80% big-mountain and off-piste expert reputation (derived from sources including ski media, freeride competition history, and expert community references), plus 20% steepness density (vertical drop divided by the square root of skiable area). The full methodology is in how the Seasoned Score works.

Airport distance is road distance to the nearest international airport, calculated from the resort centre.

Staff accommodation rating reflects whether the resort actively provides staff housing or has established routes to it — not the quality of that housing, which varies too much to score reliably without review data.

How conflicts get resolved

Sources often disagree. Verbier's skiable area appears as 330 km², 412 km², and 4 Valleys-combined figures depending on the source. When this happens:

• The most conservative credible figure is preferred over inflated marketing claims.
• Where a resort is part of a linked area (e.g. 3 Vallées, Ikon/Epic pass regions), we score the resort-specific terrain rather than the combined area, because a seasonaire's free pass usually covers the resort, not the whole network.
• The chosen source is logged. On resort pages, you can see the source for each displayed stat.

What percentiles mean on resort pages

Raw numbers alone don't tell you much. Is 1,200m of vertical a lot? (Yes — it's top-quartile globally.) Is 80 cm of average snowfall good? (Depends on the field.) The site converts every raw stat to a percentile before displaying it, so the bar next to a figure shows where that resort sits relative to the full database — not just the absolute number.

A resort at the 90th percentile for snowfall gets 90% of the snowfall bar filled. The median resort sits at 50%. This makes it immediately obvious whether a resort is exceptional, average, or poor for that stat — without needing to know that 80 cm of snowfall is mediocre by alpine standards.

How current is the data?

Each stat row on a resort page shows when it was last updated. We update seasonally for time-sensitive figures (costs, season dates) and less frequently for stable ones (vertical drop, skiable area). If you're planning a season, check the collection date for cost stats — exchange rates and local prices shift.

What to do if something looks wrong

If a figure looks off, the fastest route is the feedback link on the resort page (or email us). Include the stat, what you think the correct figure is, and the source you're drawing from. If you've worked a season there recently, your on-the-ground knowledge often beats what we can collect remotely — particularly for rent and accommodation.

Every correction that changes a displayed value gets logged in the data pipeline, so the revision history is always traceable.