How we score · Discover

How Discover ranks properties

Every property in Discover carries a score out of 100. It is not a rating of the house — it is how closely that house matches the brief you set. Change your priorities and the same house scores differently. This page explains how to read the number, and what is behind it.

TL;DR

  • The score measures fit to your brief, not house quality. The same property scores differently for a different search — neither number is wrong.
  • 80+ is a strong match, under 40 is weak — see the coloured bands below. Compare scores only within one search, never across two.
  • Your priorities set the weights — there is no fixed formula. Pick the three that genuinely matter; turning everything up cancels out.
  • Read the dimension breakdown, not just the number. The score is the sort order; the notes underneath are the actual finding.

1. Reading a score

The score answers one question: how well does this match what you asked for? A four-bedroom house on a big block might score 91 for a family and 54 for a couple who wanted a short commute and no garden. Neither score is wrong. They are answers to different briefs.

80+

Strong match

Meets nearly everything you asked for. Expect few compromises, and check the breakdown to see which dimension gave up the remaining points.

60–79

Good match with trade-offs

Most of your brief is met and one or two things are not. This is the band worth reading closely — the trade-off is usually a deliberate one you can live with, or a dealbreaker you should filter out.

40–59

Partial match

Something substantial is off — often price, commute, or a missing requirement. Worth a look only if you are flexible on whatever the breakdown flags.

Under 40

Weak match

It passed the filters but little else lines up. If a lot of your results sit here, your brief is probably too narrow for this suburb.

Compare scores within one search, never between searches. Because the weighting is rebuilt from your priorities each time you search, a 78 from yesterday’s brief and a 78 from today’s are not the same measurement. Rank within a result set; do not track a number over time.

The number on its own is the least useful part. Every property carries a dimension breakdown showing where the points came from and where they went, and short notes explaining each one — “22 minutes to work by car”, “zoned for both primary and secondary”, “12% above the suburb median”. Read those. They are the actual finding; the score is just the sort order.

2. Filtering, then ranking

Two different things happen, and it helps to keep them apart.

Step one · filtering

Some properties are removed entirely

These never appear and never get a score. A property is excluded if it is the wrong property type, sits outside your price range beyond the tolerance described below, misses your bedroom or land-size floor, or — when you have asked to avoid them — carries a high flood or bushfire risk.

Step two · ranking

Everything left is scored and sorted

Nothing is excluded at this stage. Every remaining property is scored across the eight dimensions below, those scores are combined using weights built from your priorities, and the results are ordered. A property that is weak on one dimension is ranked lower, not hidden.

This is why loosening a filter and lowering a priority do different things. Loosening a filter brings new properties into the running. Lowering a priority reshuffles the ones already there.

3. The eight dimensions

Each is scored from 0 to 100 independently, before any weighting is applied.

DimensionWhat it measuresWhat moves it
Property fitHow closely the home matches the beds, baths, parking, land size and type you asked for.Every requirement met lifts it; each compromise costs a little.
ValueThe asking price against what comparable homes in that suburb have actually sold for.Priced near or below the local median lifts it; more than about 25% above it drops it.
CommuteDoor-to-door travel time to each place you told us you go, by the mode you chose.Time, nothing else. See the bands below.
AmenitiesWhat is genuinely close by — shops, parks, cafés, medical, childcare, gyms.Only the categories you said you cared about. Others are ignored.
SchoolsWhether the address falls inside a government school zone, and what else is within about 2km.Being zoned counts for considerably more than being merely nearby.
LocationSuburb-level Census signals — household make-up, ownership, income, education, employment.Compared against Victorian baselines, not against other suburbs in your results.
RiskFlood and fire exposure, traffic and crime, and market stability, in equal halves.A recorded flood or bushfire overlay is the heaviest single factor.
LifestyleA read on how the area matches the way you described wanting to live.The softest dimension, and weighted accordingly.

Commute, in full

Commute is the one dimension with no judgement in it at all, so it is worth showing completely. Each destination you add is timed separately, by the mode you picked for it, at both peak and off-peak. Where you give several destinations, the worst one sets the score — a home that is quick to work but an hour from school is scored as an hour.

Travel timeScore
15 minutes or less100Excellent
16–25 minutes80Good
26–40 minutes60Acceptable
41–60 minutes40Long
Over 60 minutes20Very long

4. Your priorities set the weights

The eight dimension scores are not simply averaged. Each carries a weight, and the weights are rebuilt for every search from the priorities you set. There is no fixed formula that applies to everyone.

It works in three steps:

  1. Start from a base. Property fit begins at 0.15, commute at 0.10, and risk, location and lifestyle at 0.05 each. Value is set by how hard you push the value-for-money slider.
  2. Add your priorities. Each priority you turn up adds weight to the dimension it belongs to, in proportion to how far you turned it up. Several priorities pointing at the same dimension stack.
  3. Normalise. Everything is scaled so the weights total 100%. This is why raising one priority lowers everything else — the share has to come from somewhere.

Worked example

A search that pushes healthcare and gyms high, family-friendliness moderately, quiet a little, and value-for-money fairly high produces roughly this split:

  • Amenities22%
  • Property fit19%
  • Value18%
  • Commute13%
  • Risk10%
  • Location6%
  • Lifestyle6%
  • Schools5%

Amenities ends up heaviest because two separate priorities pointed at it. Schools ends up lightest despite family-friendliness being on, because only one priority fed it. Illustrative — your search will differ.

Turning everything up is the same as turning nothing up. If every priority sits at maximum, they all contribute equally, normalising cancels the effect, and you get close to the default ranking. The signal is in the difference between your priorities, not their absolute level. Pick the three that genuinely matter and leave the rest low.

5. When a priority becomes a dealbreaker

Two priorities change character when you push them past roughly 60%: schools and parks. Below that, a property lacking them is ranked lower. Above it, a property lacking them is removed from the results.

The reasoning is that past a certain point “important” stops meaning “rank it higher” and starts meaning “do not show me these”. No other priority does this — healthcare, shops, cafés, transport and the rest only ever affect ranking, however high you set them.

If your result count drops sharply when you nudge one of those two, this is why. Pull it back below the threshold to get the properties back as ranked results rather than exclusions.

6. Why you see homes outside your budget

Deliberately, and within limits. We show properties from about 20% below your minimum to about 30% above your maximum, then apply a scoring penalty that grows with the distance from your range.

Two reasons. Advertised prices are frequently ranges or guides, so a strict cut-off hides homes that would sell inside your budget. And a home slightly over budget that fits everything else is information worth having — you can decide whether to stretch. A property well outside the range still appears but scores poorly, so it sorts to the bottom rather than crowding the top.

If you want a hard edge instead, the exact-match options turn bedrooms, bathrooms and parking into strict filters.

7. What a score is not

It is not a valuation. The value dimension compares an asking price to recent local sales. It is not a certified valuation and must not be used as one.

It is not advice. The score ranks matches against the brief you set; the judgement is yours. See our terms.

It is not a quality rating. Nothing in the score sees condition, light, layout, noise on the day, or how a street feels at 6pm. A 90 can still be wrong for you the moment you walk in — which is the part of the search we cannot do.

It is only as current as its inputs. Travel times are modelled, not measured on the day. School zones move. Census data is from 2021. Every figure on a property card carries its source and date — check them where they matter.

Something look wrong? Tell us — a wrong number is a bug we can find.

Data sources

Demographic figures from the Census of Population and Housing © Australian Bureau of Statistics, used under Creative Commons Attribution.

Cadastral and boundary data contains information © State of Victoria (Vicmap), licensed under CC BY 4.0.

Sale figures are aggregated from recorded transactions and shown only as medians and counts.

Narrative summaries are AI-generated from the underlying data and may contain errors.

Informational only — not financial, legal or planning advice. Verify zoning, overlays and catchments with the relevant council or school before relying on them. Data vintage: 2021 Census.