How to split a ski house when the bedrooms are not equal

Six people, one house, and a basement bunk room nobody wants to talk about. The usual advice is to give each room a fixed percentage of the rent. On a ski house that advice does not just get the amounts wrong — it can get the order wrong, and charge less for the bedroom people actually wanted.

The house sleeps six. There is a master with an ensuite bath that faces the peak, a queen on the top floor that shares the hall bath, a third bedroom with two twins, and a bunk room in the basement with no window, next to the boot room and the dryer. It is $3,600 for three nights and somebody has to say a number in the group chat.

Whoever says it first is going to propose one of three things, because those are the three things the internet will tell them to do.

The three usual answers, and where each one breaks

Give each room a percentage

This is the most common recommendation, and the most specific: the master suite pays 25–30% more, the mid-tier rooms add 10–15%, the small ones get discounted proportionally. It sounds rigorous because it produces numbers.

The problem is what the percentage is attached to. It prices the room, as though how good a bed is were a fact about the house. It is not. It is a fact about the house and the person in it together, and on a ski trip those come apart badly, because a ski group is almost never made of people who want the same thing. The person who is out the door at eight and back at four wants the bed nearest the boot room and could not care less about the view. The light sleeper wants a door that closes and quiet behind it. Somebody is there mostly for the hot tub and the town and would pay real money not to sleep in a basement.

A fixed percentage cannot express any of that. It decides, on everyone’s behalf, what the master is worth.

Draw for the master, split evenly

Everyone pays the same, and a random draw decides who gets the good room. It is popular because it is obviously impartial.

It is also the version where the most value goes missing. If the couple who would genuinely have paid an extra few hundred dollars for the master lose the draw to somebody who is indifferent, both sides are worse off than they would have been under almost any trade, and there is no trade on offer — the losers get nothing. This is the same failure as a room auction, wearing a fairer-looking hat, and it is covered at length in why splitting rent by room is harder than it looks.

Divide by person-nights

Total cost, divided by the number of nights each person actually slept there. This one is genuinely good, and it is the right answer to a question nobody in this group asked. It settles who owes what when people arrive on different days. It says nothing whatsoever about who sleeps in the basement, which is the argument you are actually having.

The house, with real numbers

A ballot says two things, and they are separate on purpose. Where you rank a room says whether you want it. What you put on its price says what you think it should cost the group. Those are routinely opposite, and the form is built so they can be.

That separation is the part a percentage table cannot reach. Sam, travelling alone, thinks the master is plainly the best room in the house and says so — $1,500, the highest number anyone puts on it. Sam also ranks it last, because alone he would be paying for all of it. High value, no desire. Both halves of that are true at once, and both are information the group can use.

Dana and Priya are one household with one bed, so their two ballots are added together rather than averaged: each writes half of what they think the room should cost. Here is what the five households said.

Read the bunk column against the ranking beside it. Marcus ranks the bunk first, because it is ten feet from the boot room and he is carrying wet gear twice a day. Sam ranks it first too. Nobody else wants it at any price — Tomás puts $350 on it and ranks it last. The room a percentage table calls the worst in the house is the first choice of two of the five households staying in it.

What the numbers actually produce

Beds go to the assignment that makes the total of what people said as large as it can be, and then prices are set so that nobody would rather have another household’s room at that room’s price. For this group, that comes out as:

Those are room totals, and all three columns add to $3,600. The fixed percentages are the ones the guides actually recommend, applied to each room’s even share — master up 30%, the queen up 12%, the twins down 15%, the basement down 25% — then scaled so the column lands on the house price.

Now compare the last two columns on the bottom three rows. The percentage rule prices the bunk room at $437.60 and each twin at $495.95: the basement is the cheapest bed in the house, discounted about a fifth, exactly as the rule intends. The mechanism prices the bunk at $462.29 and the twins at $378.70 and $375.49. The bunk is now the dearer room, by about $84.

That is not a rounding difference. It is the order changing. The windowless basement next to the dryer costs more than a bedroom upstairs — not because anyone decided a basement is nice, but because two of the five households reached for it and only one could have it, while the twins went to people who ranked them fourth and third. A fixed percentage cannot produce that result for any choice of percentages, because it prices the room before it has heard from anybody.

Why this version is checkable

Publish the whole table — every room, every price. Anyone can then read across it, compare their own room and price against everyone else’s, and confirm they would not rather swap. Tomás can see that the bunk room costs $84 more than his twin and decide, correctly, that no discount was on offer for the basement anyway. Sam can see the master at $1,820.49 — which he would have been paying alone — and be glad he ranked it last, whatever he thinks of the room.

That property has a name and a fair amount of maths behind it — what “envy-free” actually means covers why a solution always exists, whatever people happen to want. The practical point is narrower: an outcome the group can inspect does not need anybody to be trusted. A percentage table handed down by the person who booked the house does.

Two ski-trip costs a bed cannot carry

This is where ski houses differ from beach houses, and where groups that get the bedrooms right still end the weekend annoyed.

Lift tickets and rentals are per person, not per bed

Keep them on their own line and settle them separately. A five-day pass and a two-day pass differ by more than any bedroom in the house, and rolling them into one “what you owe” number buries the bed question underneath a much bigger one. Marcus, who took the bunk room to be near the boot room, is on the first lift every morning and may well end up owing the most of anyone once a full pass and rental gear are on the sheet. That is fine. It is just not a lodging fact, and mixing it in makes the lodging split look unfair when it is not.

Staggered arrival is a separate pot

Ski trips stagger more than any other kind. Two people drive up Thursday after work; the rest come Friday morning. The clean handling is to treat the extra night as its own small settlement between the people who were there for it, then run the bed allocation over the nights the whole group shares. One number cannot do both jobs, and trying to make it is how a spreadsheet acquires a fourth tab and nobody checks it.

The order to do it in

Agree the method before anybody sees the listing photos. Once people know which room is which, every proposal about how to divide is also a proposal about who gets the peak view, and it will be read that way — sometimes correctly. List every bed with its bad parts included. Collect the numbers privately, so nobody is bidding against a friend in public. Run it once, publish the result, and rerun only if an input was actually wrong.