Assigning Chores in a Shared House

Why a fresh weekly draw fails

Generating a new random allocation each week means every week is independent, which sounds fair and produces exactly the outcome that ends rotas. Over ten weeks with four people and four chores, the totals will spread, and somebody will have drawn the worst job three or four times.

The arithmetic is the same square-root behaviour that shows up everywhere on this site. The expected number of bathroom weeks per person over ten weeks is 2.5, with a standard deviation of about 1.4 — so somebody getting five is under two standard deviations out and entirely ordinary.

Ordinary is not the same as acceptable. The person with five weeks experiences a run of bad luck as a broken system, and they are not being unreasonable — they are correctly observing that the outcome was unfair, even though the process was not.

Generate in blocks and print the totals

The chore rota assigner builds a multi-week rota in one operation, deals the jobs round-robin within each week, and enforces one rule: nobody takes the same job in consecutive weeks. Round-robin dealing does most of the balancing work that independent weekly draws leave to chance.

It then prints the totals for the whole period, which is the column the house actually cares about. Twenty-four jobs across four people is six each; a rota showing 7, 6, 6, 5 is fine and a rota showing 9, 6, 5, 4 is not, and you can see which you got before it goes on the fridge.

Generate once for the whole term or the whole quarter, print it, and stop regenerating. A rota regenerated weekly has no memory of what it said last week, which defeats both the consecutive-week rule and the totals.

Weighting the jobs nobody wants

Not all chores are equal, and pretending otherwise is the second most common way a rota loses legitimacy. The bathroom and the bins are not the same job, and a rota that counts them identically is fair on a measure nobody uses.

Two workable approaches. Split the bad job into more slots — list the bathroom twice in a four-week rota so it rotates faster — or handle it separately with the straw picker each week, which makes the unpleasantness explicit and shared rather than buried in a table.

The third approach, weighting the draw so people who did it recently are less likely to get it again, is what the weighted picker is for. It works and it requires somebody to maintain the weights, which in practice is the same person who was already keeping score.

• List an unpopular job twice so it rotates through the house faster.
• Or draw it separately each week with the straw picker, so it is visible rather than buried.
• Or weight the draw against whoever did it recently — effective, and needs maintaining.
• Whatever you choose, agree it before generating rather than after the first complaint.

What actually keeps a rota alive

Almost every rota dies for a non-mathematical reason: it stops being visible. A printed rota on the fridge survives; a rota in a group chat scrolls away within four days and stops existing.

The second killer is a job whose completion is ambiguous. "Kitchen" means different things to different people, and a rota with ambiguous jobs generates arguments about whether a turn was done rather than about who has it.

Neither of those is solved by a generator. What a generator solves is the allocation being suspected of favouritism, which is a real problem and a smaller one than most houses think.

Frequently asked questions

Tools for this job

Background reading