Telling a red flag from a difference
Somewhere in the second week of a room search, most people become mildly paranoid. Every listing has a tell, every housemate has a look, and the checklist that was supposed to help has turned into a reason to reject everything.
That happens because nearly all advice in this genre puts two unrelated categories in one list. One is a red flag: evidence that a problem is likely, independent of who you are. The other is a difference: a fact about the household that only matters measured against you. Confusing them is the single most expensive mistake in a search, in both directions.
Two different kinds of statement
A red flag is a claim about the world. The lister won’t say who else lives in the house predicts trouble for anybody, in any market, whatever you’re like. You don’t need to know yourself to evaluate it.
A difference is a claim about a fit. Everyone in the house works nights predicts nothing at all until you know your own hours. For one person that’s an empty flat and a quiet evening; for another it’s someone making dinner at four in the morning through a thin wall.
The practical test: can this be wrong for you and right for someone else? If yes, it’s a difference. If it’s wrong for everyone, it’s a flag.
Why the two get merged
Because lists are easier to write than judgements, and because a flag and a difference can look identical from the doorway.
The genre also has an incentive problem. “Watch out for X” is more compelling than “X may or may not matter depending on your schedule”, so the advice drifts towards alarm. The result is readers who reject good rooms over harmless differences while walking straight past the actual flags, which are usually procedural and boring.
There’s a third category that muddies it further: the difference that becomes a flag when it’s hidden. A house where two of five residents are the same person’s partner is a difference. A house that gives you a headcount which changes between the message and the viewing is a flag — not because of the number, but because of the change.
What is actually a flag
The reliable ones share a shape: they’re about process and disclosure, not preference.
Money moving before you’ve stood in the property. Every version of this is a flag, and it doesn’t degrade with a good explanation. It is the load-bearing test in how the scams work.
Refusal or inability to answer a factual question. How many people live here, who holds the tenancy, what happened to the last person, when is it actually available. Every legitimate household can answer all four. Not knowing is different from not saying; both are worth a second question, and only one is fine.
An answer that changes. Headcount, availability date, what’s included, who the lister is. A shifting account is the strongest in-person signal there is.
Pressure to decide before you’ve had the information you asked for. Urgency in a tight market is genuine. Urgency plus “I’ll send the details after” is not.
Unwillingness to let you meet the people you’d live with, in a share where they exist and could reasonably be present.
That’s most of the list. It’s short, and none of it is about vibes.
What is only a difference
Schedules. Nights, shifts, early starts, working from home. All neutral in themselves; all consequential against yours and against the thickness of the walls.
Sociability. A house that eats together and a house where everyone keeps to their room are both fine houses. Getting the wrong one is the most common source of quiet misery in a share, and it is a fit problem, not a fault.
Cleanliness standards, up to the point where something is a health matter. Two tidy people and two relaxed people both work. One of each does not.
Age gaps, life stage, occupation, how long people have lived there. Differences. Sometimes strong ones. Not flags.
Being nervous, awkward, blunt, or bad at small talk. This is a strange conversation for both sides. Social smoothness predicts approximately nothing about whether someone is a good housemate.
And a boundary that isn’t negotiable here: protected characteristics are never in either category. Race, religion, national origin, disability, family status, sex and sexual orientation are not screening criteria on this site, and if you’re the one choosing, selection and advertising law may apply to you — it varies by jurisdiction, and your local tenant or housing service is the place to check what applies where you are.
Flag check
FLAG CHECK — sorting what you noticed
· Factual question gets no answer
→ flag. Applies to everyone.
· Answer changes between message and viewing
→ WALK AWAY. The change is the
signal, not the content.
· Household's schedule is opposite to yours
→ NOT A FLAG. A difference.
Weigh it against your walls.
· Housemates are much older or younger
than you
→ NOT A FLAG. A difference, and
often a good one.
· Very quiet, private household
→ NOT A FLAG unless you wanted
the other kind.
· Money requested before you've been inside
→ WALK AWAY. No version of this
is a difference.
· Household won't discuss what happened to
the previous occupant
→ flag. Everyone knows and
everyone can say.
Using it in practice
Take whatever you noticed and sort it into two columns before you decide anything. Most searches have one or two items in the flag column and eleven in the difference column, and the difference column is where the actual decision lives.
Then rank the differences against yourself rather than against an abstract ideal, which is what makes comparing two rooms tractable. You are not looking for the objectively best house. You are looking for the one whose differences you don’t mind.
One flag ends it. Any number of differences is just information.
What the distinction can’t do for you
It won’t classify the ambiguous cases, and there are plenty. A housemate who was a bit off with you at the viewing might have been having a bad day or might be the reason three people left this year, and nothing available to you at the time distinguishes those.
It also can’t stop the failure mode where the flag column is empty and the house is still wrong. Most bad shares aren’t caused by anything that was detectable in advance; they’re caused by two reasonable people whose differences turned out to matter more than either expected. The sorting exercise catches the knowable problems, which is a genuine subset and not the whole of it.