Skip to contentSkip to suite
Room Lab
Quick start
Guided start · 1 of 301 · Basis

Compare like with like.

Use the same evidence fields for two rooms, or the same room before and after one intervention.

One room photographed as two halves meeting on a diagonal seam: bare grey plaster, an exposed ceiling and a stripped floor on the left; the same corner finished with a timber slat wall, tiling and a laid wood floor on the right.

08 · Case Compare

What changed, and did the room actually improve?

Put two rooms — or the same room before and after — side by side and read the difference.

Pick two rooms. Differences are shown first; anything identical stays quiet. Example rooms are worked out as you look at them and are not saved unless you load one.

Reading saved cases on this device…

The read

Dining room that fits until people sit is not measurably better or worse than Strangled living room.

Across the 6 conditions both rooms rated, friction averages 58 and 56 out of 100. That gap is inside the noise. Look at what changed below rather than at the totals. The average hides a trade: Comfort and Function rose while Light and Reset fell. Read each condition below, not the total.

Did the original complaint improve?

No. Movement was the problem ranked first in Strangled living room, and its pressure moved from 78 to 79 out of 100 — inside the noise, so it has not measurably changed.

Condition by condition

Pressure on each condition from Strangled living room to Dining room that fits until people sit, out of 100 — higher is worse. Pressure on one condition, not a grade of the room. Nothing on this board is averaged.

Gained
Less pressure on Light (52 → 44) and Reset (70 → 46).
Lost
More pressure on Comfort (48 → 52) and Function (64 → 73).
Unchanged
Movement (78 → 79) and Personality (38 → 40) — a move under 4 points is inside the noise.
Next question
Movement was the original complaint and it did not move, even though Light and Reset did. What would change Movement itself?

What else changed

  • The next move

    Clear the primary walk changes to Move the largest piece once

    The second room has finished the cheaper work and is ready for reversible tests.

  • Budget for this round

    $800 changes to $900

    A larger budget does not move a room up the order of work. It only changes what the later steps can be.

  • Remaining after commitments

    $656 changes to $738

    What is actually left once known commitments and the reserve are taken out.

  • Jobs the room carries

    gather, store changes to dine, host, store

    A room asked to do more jobs needs edges before it needs objects.

  • Zones the plan could place

    Social, Store, Reset / drop changes to Host, Store, Dining + chair pull, Reset / drop, Host / service

    A zone is only drawn where a measurement, an opening, or a placed piece can position it honestly.

  • Still unknown

    exact outlet locations on the north wall; whether the window is operable changes to whether the table leaves can come out permanently; whether the sideboard is doing a job that a wall shelf could do

    The open questions changed. Unknown is never counted as zero.

Field by field

Everything both rooms declared

Each row is marked = unchanged or ≠ changed; unchanged values are also greyed out. Unknown stays unknown — it is never counted as zero.

Answer changed
Strangled living room: Movement looks like the governing problem. Clear the primary walk — no money yet.Dining room that fits until people sit: Movement looks like the governing problem. Move the largest piece once — no money yet.
Why changed
Strangled living room: Movement is the primary friction at 78. Movement scores 78. How people walk: pinch points, blocked routes, and whether a body can pass. Reset at 70 is real, but it is not the thing a body hits first. Overflow pile narrows the primary walk to 18" — 18" under the minimum. Addresses congestion and storage pile with a derived cause behind it. It costs nothing, takes about 30 minutes, and is undone in a minute.Dining room that fits until people sit: Movement is the primary friction at 79. Movement scores 79. How people walk: pinch points, blocked routes, and whether a body can pass. Function at 73 is real, but it is not the thing a body hits first. The primary walk measures 30" at its narrowest, 6" under the minimum. Addresses congestion and scale mismatch with a derived cause behind it. It costs nothing, takes about 45 minutes, and is undone in a minute.
Caveat changed
Strangled living room: Unnamed: exact outlet locations on the north wall; whether the window is operable.Dining room that fits until people sit: Unnamed: whether the table leaves can come out permanently; whether the sideboard is doing a job that a wall shelf could do. Function at 73 is close. If that score is wrong, the sequence may swap.
Safety unchanged
Strangled living room: no named concernDining room that fits until people sit: no named concern
Ranked problem changed
Strangled living room: Movement pressure 78/100 · higher is worseDining room that fits until people sit: Movement pressure 79/100 · higher is worse
Next changed
Strangled living room: Clear the primary walkDining room that fits until people sit: Move the largest piece once
Budget for this round changed
Strangled living room: $800Dining room that fits until people sit: $900
Remaining budget changed
Strangled living room: $656Dining room that fits until people sit: $738
Zones placed changed
Strangled living room: Social, Store, Reset / dropDining room that fits until people sit: Host, Store, Dining + chair pull, Reset / drop, Host / service
Still unknown changed
Strangled living room: 2Dining room that fits until people sit: 3
Jobs changed
Strangled living room: gather, storeDining room that fits until people sit: dine, host, store
Movement pressure changed
Strangled living room: Movement pressure 78/100 · higher is worseDining room that fits until people sit: Movement pressure 79/100 · higher is worse+1 · more pressure
Light pressure changed
Strangled living room: Light pressure 52/100 · higher is worseDining room that fits until people sit: Light pressure 44/100 · higher is worse−8 · less pressure
Comfort pressure changed
Strangled living room: Comfort pressure 48/100 · higher is worseDining room that fits until people sit: Comfort pressure 52/100 · higher is worse+4 · more pressure
Personality pressure changed
Strangled living room: Personality pressure 38/100 · higher is worseDining room that fits until people sit: Personality pressure 40/100 · higher is worse+2 · more pressure
Function pressure changed
Strangled living room: Function pressure 64/100 · higher is worseDining room that fits until people sit: Function pressure 73/100 · higher is worse+9 · more pressure
Reset pressure changed
Strangled living room: Reset pressure 70/100 · higher is worseDining room that fits until people sit: Reset pressure 46/100 · higher is worse−24 · less pressure

Field guide · comparison

Compare like with like before calling it improvement

Before-and-after evidence is strongest when the same room, same scoring definitions, and comparable use conditions are aligned. Two different rooms can still be compared, but that is benchmarking — not proof that one intervention caused the difference.

Before-and-after comparison diagramPaired before-and-after bars aligned by the same definitions and comparable conditions.beforeaftersame definitions · comparable conditionsthe intervention belongs between the two readings
IllustrativeNot to scale. A sketch of the method, not a drawing of your room.

Read it

  1. 01Identify whether this is the same room over time or two different rooms. The inference is different.
  2. 02Check that both sides have the same dimensions, score definitions, and known/unknown fields before interpreting a delta.
  3. 03Read safety, jobs, budget, and the primary friction alongside the headline score so a gain in one dimension does not hide a new problem.

Test it

  1. 01For a before/after case, repeat the reading under a comparable use condition whenever practical.
  2. 02Look first for the condition the intervention predicted it would move, then inspect collateral movement.
  3. 03If the evidence is incomplete, label the comparison provisional instead of filling blanks with assumptions.

What counts as proof

The proof is a traceable delta with comparable inputs and a recorded intervention between them. A visually nicer after image by itself is not causal evidence.

Where this comes from

links reviewed 2026-09-08

Demonstrations are optional learning aids and the standards are the floor under the method. Neither is evidence about your room: Room Lab does not treat a video as proof, and a link never replaces a measured test.

Read the source

NIST TN 1297: Repeatability and reproducibilityNational Institute of Standards and TechnologySame conditions, same definitions, then the difference means something. Change the conditions and the delta is describing the conditions.

Decision matrix · signal → test → proof

Decide whether a delta is evidence, a benchmark, or noise

Do not jump from diagnosis to purchase. Read the signal, run the lightest reversible test, and commit only when the result survives real use. The stop column is part of the decision, not fine print.

Room Lab decision flowA four-stage flow from an observed signal, through one reversible test and proof under real use, to either commitment or stopping the proposed change.signalobservetestone variableproofreal usecommit / stoprecord whyno evidence → no permanent move
Decision matrix for Decide whether a delta is evidence, a benchmark, or noise: signal, reversible test, evidence for commitment, and stop condition.
SignalWhat you actually see TestLightest honest experiment CommitEvidence that earns permanence StopWhat invalidates the move
01The same room has a before and after reading with one recorded intervention between them.Align the readings by comparable use condition and inspect the dimension the intervention predicted it would move.Treat the delta as useful intervention evidence when the predicted dimension improves without a compensating failure elsewhere.Stop short of causal language if other substantial changes occurred between readings.
02Two different rooms have similar headline scores.Compare their jobs, dimensions, known fields, safety state, and primary friction before interpreting the score.Use the comparison as a benchmark only when the relevant inputs are sufficiently comparable.Stop if the similarity comes only from one aggregate number while the rooms solve different jobs or have different constraints.
03One side of the comparison contains missing or unverified evidence.Mark the comparison provisional and collect the missing measurement or repeat the reading before drawing a conclusion.Promote the result from provisional only when the critical fields are known on both sides.Stop if the conclusion requires filling a blank with an assumption or treating unknown as zero.

Evidence map · branch by branch

Why these tests and stop conditions are here

Each matrix branch is tied to a primary standard, public guidance, technical reference, or vetted expert teaching source. The applicability note explains the connection; the source limits prevent a general reference from being mistaken for a room-specific fact.

reviewed 2026-09-08

Decision 01

Before-and-after interpretation is stronger when the same condition is measured consistently and the intervention was defined in advance.

Decision 02

Similar aggregate scores do not make unlike rooms equivalent; comparability depends on the underlying conditions and measurement definitions.

Decision 03

Missing or unverified inputs weaken comparability, so the appropriate state is provisional rather than assumption-filled.

Worked room · method example

illustration, not a recommendation for your room

Bedroom compared before and after rotating the bed away from the door path

01Signal
The after reading shows better circulation and lower visual friction.
02Reversible test
Confirm both readings used the same room dimensions and that the bed rotation was the only meaningful layout change.
03Observed result
The route widens and the circulation score improves while sleep, storage, and safety remain stable.
04Decision
Keep the rotation and attribute the circulation improvement to the tested layout change; do not generalize it to other rooms.