Section formula in 3D: Question bank — Maths deep dive

Before you begin, keep this anchor in front of you. For $A (x_{1}, y_{1}, z_{1})$ and $B (x_{2}, y_{2}, z_{2})$ divided in ratio $m : n$ with $A P : P B = m : n$ : $P = (\frac{m x _{2} + n x _{1}}{m + n}, \frac{m y _{2} + n y _{1}}{m + n}, \frac{m z _{2} + n z _{1}}{m + n}) .$ Here $m$ is the first number of the ratio (the one touching $A$ 's side of the walk), yet it multiplies $x_{2}$ , the far endpoint $B$ . That single cross-over is where most traps live.

True or false — justify

The heavier weight

m

multiplies the coordinate of the near endpoint

A

False. The weight

m

(the

A P

part of

A P : P B

) multiplies the far endpoint

B

, because loading weight on

B

pulls

P

toward

B

For internal division,

P

always lands strictly between

A

and

B

True. Internal division requires both

m > 0

and

n > 0

, so

P = \frac{m b + n a}{m + n}

is a genuine mix of both endpoints and never leaves the segment.

The midpoint is the special case

m = n

, not just

m = n = 1

True. Any equal ratio like

5 : 5

gives the same point as

1 : 1

; the ratio

\frac{m}{n}

is what matters, so

m = n

(any common value) yields the midpoint.

Swapping the order of the endpoints (

A \leftrightarrow B

) but keeping ratio

m : n

gives the same point

False. Now

m

multiplies the other endpoint, so

P

moves. Order and ratio are linked: dividing

B A

m : n

equals dividing

A B

n : m

In external division the denominator is

m + n

False. External division replaces

n \to - n

, so the denominator is ==

m - n

== and the numerators subtract.

m > n

in external division,

P

lies beyond

B

True. A larger

m

leans the point toward

B

; pushed outside, it overshoots past

B

rather than falling short before

A

Each of

x, y, z

needs the same ratio

m : n

True. 3D section is three independent 1D averages sharing one ratio; you never change

m, n

between coordinates.

The vector form

p = \frac{m b + n a}{m + n}

secretly depends on the origin you chose

False. It is coordinate-free: it uses position vectors, but the relation

A P = \frac{m}{n} P B

holds for any origin, so

P

is a fixed physical point.

Ratio

1 : 2

and ratio

2 : 4

divide

A B

at different points

False. Only the value

\frac{m}{n}

matters;

1 : 2

and

2 : 4

are the same ratio and give the identical point.

For external division, negative coordinates in the answer signal an arithmetic error

False. External points routinely have coordinates outside the segment's range, including negatives, since

P

sits off the segment — that is expected, not an error.

Spot the error

"Dividing

A, B

2 : 3

, I wrote

x = \frac{2 x _{1} + 3 x _{2}}{5}

The weight

2

m

) must multiply

x_{2}

(the far point

B

), not

x_{1}

. Correct:

x = \frac{2 x _{2} + 3 x _{1}}{5}

"External division in

3 : 1

: I used

\frac{3 x _{2} + 1 \cdot x _{1}}{3 + 1}

Two errors: external needs

n \to - n

, so it should be

\frac{3 x _{2} - 1 \cdot x _{1}}{3 - 1}

, with a minus and denominator

2

, not

4

"Midpoint of

A, B

M = (\frac{x _{1} - x _{2}}{2}, \dots)

Midpoint is the average, so it adds:

\frac{x _{1} + x _{2}}{2}

. Subtraction gives half the displacement vector, a direction, not a point.

"I found the ratio using

x

, got

k = \frac{1}{2}

, and stopped there."

You must verify with a second coordinate (

y

z

). If they disagree,

P

is not on line

A B

at all, and the single-coordinate ratio was meaningless.

"For internal division I mixed coordinates:

x = \frac{m x _{2} + n z _{1}}{m + n}

You pulled a

z

into the

x

-line. Each coordinate is its own independent average —

x

uses only

x_{1}, x_{2}

; keep the three computations on separate lines.

"To check collinearity I computed the distance

A P

and

P B

and added them."

Distances add to

A B

only if

P

is between them; equal sum-of-parts is necessary but you should confirm direction too. Better: verify

P

satisfies the section relation, or use direction ratios.

Why questions

Why does the same ratio

m : n

work for all three coordinates at once?

Because

p = \frac{m b + n a}{m + n}

is one vector equation; reading off components splits it into three copies of the same weighted average, so

x, y, z

inherit identical weights.

Why does replacing

n \to - n

turn internal into external division?

Externally

A P

and

P B

point in opposite directions, so

A P : P B

carries an effective negative sign; writing

n \to - n

is exactly how that reversed sense enters the algebra.

Why must

m + n \neq = 0

for the formula to make sense?

m + n

is the denominator; if it were zero we'd divide by zero. Physically

m + n = 0

(external with

m = n

) means the "point" runs off to infinity along the line's direction.

Why is finding an unknown ratio best done by setting it to

k : 1

One free number

k = \frac{m}{n}

captures the whole ratio, so a single coordinate equation solves it linearly; the leftover coordinate then independently checks the result.

Why does more weight on an endpoint pull

P

toward it, not away?

A weighted average leans toward the heavier term, just like a seesaw balances closer to the heavier child; raising

m

(attached to

B

) drags the mixed point

P

nearer

B

Why can the centroid be found by a

2 : 1

section on a median?

A median's centroid splits it

2 : 1

from vertex to opposite midpoint; applying the section formula with that ratio to those two points lands exactly on the centroid.

Edge cases

What point does external division with

m = n

give?

None — denominator

m - n = 0

, so

P

is a "point at infinity": the direction of line

A B

but no finite location.

What if

P

coincides with

A

— what is the ratio

A P : P B

A P = 0

, so

m = 0

, giving ratio

0 : n

. The formula returns

p = \frac{n a}{n} = a = A

, consistent.

What if the two endpoints are identical,

A = B

The "segment" is a single point; every ratio returns that same point, and there is no meaningful division since there's nothing to divide.

If a computed "ratio" comes out negative from a single coordinate, what does it mean?

It signals external division (or that

P

lies outside segment

A B

); a negative

k

means

A P

and

P B

oppose, placing

P

beyond an endpoint.

Does the section formula still hold if

A, B, P

are given but

P

is not actually on line

A B

No. The formula assumes

P

lies on the line; if a second coordinate check fails,

P

is off-line and no single ratio exists — see Distance formula in 3D to confirm collinearity.

For internal division as

m \to \infty

(with

n

fixed), where does

P

go?

Toward

B

. The weight on

B

dominates, so

p \to b

; symmetrically,

n \to \infty

drives

P \to A

Connections

Section formula in 3D — the parent formulas these traps stress-test.
Position vectors — why $p = \frac{m b + n a}{m + n}$ is origin-independent.
Distance formula in 3D — the collinearity check behind "is $P$ really on $A B$ ?"
Centroid of a triangle in 3D — the $2 : 1$ median edge case.
Direction ratios and direction cosines — verifying $A, P, B$ collinear.
Section formula in 2D — the same traps, one dimension fewer.
Equation of a line in 3D — internal/external division as line parameter.