Line items (same as the original site): round-trip flight, bags $150est., airport transfer, 7 nights’ lodging, tourist tax, lift pass, food & drink, local transport, mountain insurance where required or recommended (Italy’s mandatory RC cover⁶³, France’s Carré Neige⁶⁴), travel insurance, tips, and a 10% contingency. Each line is computed in its original currency, converted at the ECB reference rates of 9 October 2026⁵³ and rounded to whole dollars; the contingency is 10% of the rounded sum.

Tiers. 1 Shoestring (hostel bed; where no hostel is open in winter, the cheapest bookable budget room for one adult; self-catered, public bus/train) · 2 Economy (2–3★ hotel or basic apartment, often in a valley town, B&B) · 3 Comfort (well-reviewed 3★ or in-resort apartment, mix of cooking and restaurants) · 4 Premium (= the original “Best value”: 4★ half-board near the lifts or a catered chalet, private transfer) · 5 Luxe (= the original “Splurge”: 5★ or luxury catered chalet, ski-in/out, private transfer).

Lodgingest.. For every resort and tier we sample real 2026/27 rates from named properties of the right class on Booking.com⁶⁸ for two weeks (Jan 9–16 and Feb 13–20, 2027; Christmas was not sampled), and store the property, link, rate and date checked (listed on each resort page). Each rate is converted to € per person per night (÷ occupancy, ÷ 7 for weekly rates). Because Premium and Luxe assume half-board and Economy breakfast, a sample without those meals gets an allowance added (breakfast €12, dinner €35 per person per night at EU-average prices × the country price level⁵⁷)est.. A seasonal curve per country is the median, over properties sampled in both weeks, of the same property’s January ÷ February rate (Christmas ÷ February where sampled; without Christmas samples Christmas weeks stay at the February level), and interpolated across the 21 weeks: early December = January level; Christmas weeks = Christmas level; 2 Jan = midway; 9 Jan → 13 Feb rising linearly to the February level; 13–20 Feb = February; 27 Feb → 27 Mar falling back to the January level; April = January level. The resort’s price for a week is the median of its samples (each first converted to its February equivalent) × the curve, and the range is the cheapest to dearest sample. With fewer than 3 named properties the country median for that tier is used, and if a country has no samples the Europe-wide median × the Eurostat restaurant-and-hotel price level⁵⁷; both show a low confidence chip. The day-ticket price index and the school-holiday price bands of v1 are gone. Tier order: modelled lodging is never priced below the tier beneath it +10%, nor above a quoted tier above it −10%; and on the trip total, a tier with modelled lodging is lowered to at least 3% below the tier above and raised to at least 3% above the tier below (quoted lodging is never changed). In some resorts real 3★ hotels cost more per person than the quoted Premium apartment or chalet — there the Comfort total is held just under Premium. Whenever a guard bites, the lodging line says so. A test checks that totals rise from tier 1 to 5 for every resort × week × group size (1–24) × Ikon/flight setting.

Coverage now: 8968 lodging samples (8952 new named-property samples plus the original site’s quotes), 255 resort × tier prices from ≥ 3 sampled properties, 0 from the country median, 0 from the Europe-wide median, 0 with no data.

Quoted prices. For La Thuile, Chamonix, Zermatt, Ischgl and St. Moritz the quoted 2027 Premium and Luxe lines are kept exactly (hotel and chalet rates, transfers, tourist taxes, day tickets), so at the default settings these ten totals reproduce the original site to the dollar. A quote is used as-is for the weeks it covers; other weeks are scaled by the same seasonal curveest..

Group size changes the math. Per-vehicle costs (private transfers, day-trip buses) use ⌈N/16⌉ 16-seat vehicles; groups of 8 or fewer use one 8-seat van at 60% of the 16-seat priceest.. Whole-chalet quotes are divided by N and only used when N is between 60% of the beds and the capacity; otherwise the tier’s modelled lodging is used and the swap is shown. Rooms are double occupancy; with an odd N one guest pays a +50% single supplementest., averaged across the group.

Flights. Choose a departure city (Boston, Denver, New York, Chicago or Los Angeles) and optionally a second city with its own headcount; the flight line is the head-count-weighted average. Fares come from a table of dated Google Flights spot checks by origin × gateway airport × week type (Christmas, January, February, March, April)⁶¹. Without a fare for that combination we use the median of the same origin’s other gateways for that week typeest., and without any, the original $900 US East Coast allowanceest.. Turning flights off removes the flight and bag lines (and their share of the contingency).

Transfersest.: road distance = great-circle distance to the gateway airport × 1.3⁵⁸. Private 16-seat minibus €4.15/km¹⁴; public bus/train €0.174/km¹⁵; Swiss resorts go by train at CHF 0.58/km³⁵, Luxe × 150/130. Foodest.: tiers 1–3 use weekly allowances of €200 / €420 / €520 × the Eurostat price level⁵⁷; tiers 4–5 the original site’s “beyond half-board” allowances. Tourist taxest.: France €2.50, Italy €3, Austria €4, Switzerland CHF 4 per night; other countries “no data”. Allowancesest.: travel insurance $120–$320 and tips $20–$160 by tier; local transport $20–$60.

Lift pass and Ikon. With “We hold Ikon passes” on, Ikon resorts cost $0 within the allotment: Base = 5 days at European destinations, so the 6 ski days (Sunday–Friday) need one day ticket; Full = 7 days, no ticket²². Ikon’s own information says Base blackout dates do not apply at European destinations^117,118,119; Base holders only buy the 6th-day ticket. With the switch off, or at a non-Ikon resort, a 6-day adult pass is added: the published 2026/27 price for the week where we found one, otherwise the day ticket × 5.26 (median 6-day ÷ 1-day ratio of four official tariffs)est.⁶². Hotel Pollux (Zermatt Premium) is sold only with a 6-day pass, so that pass appears as its own line. Ikon Pass prices themselves are never added.

Powder odds — the original site’s method, run identically for every resort and week. Daily snowfall from ECMWF ERA5 reanalysis for 1996–2026 at each resort’s mid-mountain elevation⁵¹ is rescaled so each month’s mean matches Snow-Forecast’s published weekly average⁵². For a week starting Saturday D we count every 7-day window starting between D−5 and D+9 across 30–31 winters (about 450–465 windows). Big-storm week (8+ in) = at least 20 cm (7.9 in) in the window, shown as a percentage first and as “about N in 10” second. Powder day (4+ in) = a day with at least 10 cm; we show powder days per week and the share of weeks with at least one. Rating: Best 60%+, Great 50–59%, Good 40–49%, Fair under 40%; missing data shows “no data”, never a rating. Historical odds, not a forecast.

Snow reliabilityest. (0–100, per resort and week) is a separate score: 35% historical snow depth for the week (0 at ≤ 12 in, full marks at ≥ 59 in) where a sourced depth series exists; 25% cold base = 1 − share of days in that week whose daily mean temperature at the base is above 0 °C (ERA5 2016–2026)⁵¹; 25% share of terrain above 2,000 m (published where available, otherwise the share of the lift-served vertical above 2,000 m from the resort’s base and top elevations¹¹, est.); 15% snowmaking cover where the resort publishes it. Weights are renormalised over the inputs available; at least two inputs including depth or cold base are needed, else “no data”. Very reliable ≥ 75, Reliable 55–74, Variable 35–54, Marginal < 35. Temperatures are 2016–2026 ERA5 averages for the same Sunday–Friday at the base and top elevations^51,11.

Holiday pressureest. (the v1 “crowds” measure): for each country we take the official 2026/27 school-holiday calendars and compute, for every week, the share of pupils on break during the stay’s Monday–Friday, weighted by regional population⁵⁶. A resort’s holiday pressure is the weighted average over its visitor markets — a resort-specific mix where a tourist office or operator publishes one, otherwise the national figures (Austria, Switzerland, France⁶⁶, Italy (Trentino as proxy)¹⁰¹, Andorra¹⁰⁵, Åre and Hemsedal¹¹⁶). If less than half the market weight has holiday data the score is “no data”. Quiet < 15%, Moderate 15–39%, Busy ≥ 40%. “Avoid school holidays” hides Busy resorts and gives crowds more weight in the ranking.

Queue riskest. (0–100, per resort): 40% holiday pressure from a resort-specific visitor mix, 35% bed base (or skier visits) per 1,000 people/hour of lift capacity, as a percentile among the resorts with data, and 25% day-trip exposure (≤ 1 h by road from a big city = 100, ≥ 3 h = 0). Each input is footnoted on the resort page; with fewer than two sourced inputs the score is “no data”. Low < 35, Medium 35–59, High ≥ 60.

Ski-Bag scores every resort in the selected region (All, North America or Europe; North America is coming soon) on eight factors, each 0–1 relative to the other visible resorts: low cost (cheapest all-in per person = 1), powder odds (big-storm week share), snow-sure (the reliability score), fewer crowds (mean of holiday pressure and queue risk where available), fits our ability mix, easy travel (1 for a 1-hour transfer from the gateway airport, falling to 0 at 4 hours), nightlife (an editorial tag) and steeps & off-piste (difficult-piste share and top elevation)¹¹.

Ability mix: “Mostly beginners” rewards easy-piste share (40%+ = 1); “Mixed levels” rewards easy + intermediate share (60% → 0, 100% → 1); “Mostly advanced” blends intermediate + difficult share with top elevation; “Experts” blends difficult share (35%+ = 1) with top elevation. Missing data scores a neutral 0.5.

You set each factor’s importance from Off (0) to High (3). Defaults: cost Med, powder Med, snow-sure Low, crowds Low, ability Low, travel Low, nightlife and steeps Off. “Avoid school holidays” adds one step to crowds. The score is the weighted average Σ(weight × factor) ÷ Σ weights, shown out of 100; “Why this ranks #1” lists every factor’s points so the total can be checked by hand. Passes: Ikon lift-pass savings are in the cost; Epic is a placeholder until North American data lands. The ranking is a convenience, not a recommendation, and has no hidden inputs.

US units throughout: feet (rounded to 10 ft), inches, miles and °F. Prices in US dollars with the original currency in brackets, at ECB reference rates of 9 Oct 2026⁵³: €1 = $1.1206, CHF 1 = $1.2033, £1 = $1.3221, NOK 1 = $0.1046, SEK 1 = $0.1003.

• verify Hotel Pollux (Zermatt Premium) is only sold as a package with a 6-day pass; the pass is shown as its own line, valued at the lowest published 6-day price^37,77
• verify Hotel Schweizerhof (St. Moritz Luxe) is a 3-night package scaled to 7 nights⁴⁹
• verify Dolomiti Superski’s €453 is the high-season 6-day price, applied to every week¹⁰⁴; Zermatt’s CHF 384 is the lowest date-based price⁷⁷; Åre’s SEK 715 day ticket is the Christmas-week price¹¹⁵
• verify Day tickets for Engelberg, Laax and Obergurgl are not published in our sources; the country median is usedest.
• verify Flights to OSD: for 19 origin × week combinations (BOS january, BOS april, BOS christmas, BOS february, CHI january, CHI april, CHI christmas, CHI february, DEN january, DEN april, DEN christmas, DEN february, LAX january, LAX christmas, LAX february, NYC april, NYC march, NYC christmas, NYC february) Google Flights listed only fares of $2,960–$27,675 (mostly 2-stop), over twice the median of the same origin’s other gateways; those fares are not used and the other-gateway median is used insteadest.⁶¹
• Tourist tax: no data for Andorra, Spain, Bulgaria, Slovakia, Sweden and Norway (not included).
• Deep-dive text (snow, temperatures, sample week, hotel quotes) describes the example week Feb 13–21, 2027 and is labelled so.