From geometry to a materials list

The last translation on any job is from geometry to a pile of lumber — and the lumber yard doesn't sell square feet. It sells boards, in lengths, in a count you have to be right about before the truck leaves.

First, name the two directions

Everything below depends on one decision you make before any arithmetic: which way the boards run. Take a 16 × 12 deck.

• Joists span the 16 ft direction, spaced across the 12 ft: that's a stud count turned flat — ⌈144 ÷ 16⌉ + 1 = 10 joists, rims included.
• Decking crosses them, so every row spans the 12 ft width and the rows stack along the 16 ft length.

Two counts, perpendicular, each dividing its own direction. Feed one of them the wrong side and the order still looks plausible — which is exactly why it's worth saying out loud before you multiply anything.

A row is a board plus its gap

Decking doesn't cover its own width. Every row occupies its face plus the gap after it, and that's what divides into the direction the rows stack — the length:

$\text{rows} = \left\lceil \frac{\text{length} \times 12}{\text{face} + \text{gap}} \right\rceil$

A 16 ft length, 5½" boards, ¼" gap: 192 ÷ 5.75 = 33.39 → 34 rows.

Look at that 33.39 for a second. Thirty-three rows cover 189¾ inches — you are 2¼ inches short. Those two and a quarter inches cost a whole row of decking, which you will then rip. This is what the ceiling function means on a job site: the fraction doesn't average out, it becomes a board.

And the gap earns its keep in coverage, not just drainage:

Butting them tight buys you an extra row of material and a deck that has nowhere to shed water. Wet lumber will shrink its own gap later — which is an argument for tighter spacing when you're laying green stock, not an argument for zero.

Boards per row is a stock-length decision

Each row has to reach across the width, so the stock length gets measured against the width, not the length:

$\text{boards per row} = \left\lceil \frac{\text{width}}{\text{board length}} \right\rceil$

A 12 ft row out of 16 ft stock is one board: 34 boards, no joints, clean sightlines — and 4 ft of drop per row you'll cut off.

Buy 8 ft stock instead and every row needs two — 68 boards — with 4 ft left over on each one anyway. Same drop, twice the joints, and every butt joint has to land on a joist (or get blocking under it). The cheaper board is often the more expensive deck.

Fasteners are a crossing count

Two screws at every place a board crosses a joist: 34 rows × 10 joists × 2 = 680 screws. Nobody estimates that by eye, and everybody underestimates it. The takeoff multiplies it out so the screws show up as a bucket count, not as a second trip.

Waste is a decision you declare

Notice what the arithmetic above does not include: breakage, culls, the board with the wandering crown, the miscut. Every real order carries a waste allowance — commonly 10% on decking, more for diagonal layouts, less for a simple rectangle in dry stock.

The tool applies it as a toggle you can see and switch off, on purpose. A hidden waste factor is how an estimate quietly becomes wrong in both directions: you can't tell whether you're over because of waste or over because the count was bad. Declared, it stays yours to argue with.

Run it

Open the deck board takeoff with the 16 × 12 deck loaded:

1. Confirm the 34 rows, the 10 joists and the 680 screws.
2. Switch the board length to 8 ft. Watch the board count double and ask where those joints are going to land.
3. Turn the waste allowance off and back on. That difference is the number you're deciding on — not one the tool decided for you.