Run the sun-hours test once, the shape passes, and the number starts to travel. It goes into the deck, then into the pitch, then quietly into the pro forma as a fact about the building — rather than a fact about a block of extruded volume that nobody has drawn a single unit into yet.
That happens under schedule pressure, not dishonesty. A massing is fast to build and fast to test, so it's what gets tested first. The trouble starts when the result gets treated as final before the thing it describes has become an actual floor plate.
A massing win is a hypothesis, not a result
A blocked-out volume is the cheapest way to test a shape, which is exactly why the first daylight pass runs against one. Stagger the towers, open gaps between them, hold the same built volume and search for a better arrangement — then check whether the shape claws back direct sun hours a denser or more literal option would lose. When it does, that's a genuine result, and it's worth acting on.
What it isn't yet is a result about units. A massing has no core position, so the test can't say which apartment actually loses light to its own lift shaft. It has no podium, so it can't say what the lower floors see once a plinth and its setbacks sit inside the sun path. The number that comes out of a massing-stage study describes a volume. Nobody buys a volume — they buy a floor plate with a window in a specific place.
The plan changes what the massing promised
Draw the typical plate — place the core, split the floor into real units, decide which face gets the larger unit and which gets the smaller one — and the geometry under test changes completely. The core that made the massing efficient now sits exactly where a deep unit needed its window. The podium that made the ground floor work now throws a setback shadow the blocked-out volume never had to account for. Neither of these is a flaw in the earlier study. It's the difference between testing an idea and testing a building.
Re-running the same direct-sun-hours check — same peak-day date, same minimum-hours threshold — against the resolved typical plan and podium, not just the massing, is the only way to find out whether the gain the massing promised is still there once the core, the units and the plinth actually exist in the model.
Sometimes it holds exactly as promised. Sometimes a unit that looked fine at the massing stage turns out to sit behind the core's own shadow once the plate is actually drawn, and the layout has to move before anyone builds it. Either way, that's information the massing study was never going to give you, because the massing study wasn't testing that geometry.
Skip the second pass and the risk stops being abstract. A daylight advantage that gets written into a sales narrative — or into an investor's read of what a unit is worth per square foot — before the plan confirms it is a number someone eventually has to walk back. Walking back a claim after it's already in a buyer's head costs more than the extra study would have.
Where the judgment stays
Neither test decides anything by itself. The software scores whatever geometry it's handed against a threshold someone set — annual cumulative hours, a peak-day minimum, whatever the brief calls for. It can't say whether a unit that misses the threshold by a narrow margin is worth moving a core for, or whether the real fix is a different plate altogether. That call belongs to the people who know what the unit is worth to sell and what moving a core actually costs to build, and it gets made after the number, not instead of it.
What the second test buys is timing. It puts the real question — does the plan still deliver what the massing promised — in front of the architect and the developer while the plate is still a drawing, not after it has already shipped as a floor someone has to discount.
A daylight claim is only as good as the resolution it was tested at. If a scheme you're evaluating has a promising massing-stage number and no plan-stage confirmation behind it yet, closing that gap before you commit to it is exactly what the 7-day review is for. Start a review.