In short

Take any two distinct points in a plane. There is exactly one straight line that passes through both of them — no more, no fewer. Drop down to a single point and you get infinitely many lines through it (a whole pencil). Bump up to three randomly chosen points and you usually get no line through all three (you'd need them to be collinear, which is a coincidence). Two is the smallest number of points that pins a line down completely, and it is also the largest number you can place freely before the line is forced. This is one of the foundational claims of plane geometry — Euclid stated it as his first postulate.

You can write a line in algebra as y = mx + c. Two numbers, m (the slope) and c (the y-intercept), are enough to specify the entire infinite line. Two numbers means two unknowns. To pin them both down you need two equations, and each point on the line gives you exactly one equation. So algebraically the count works out perfectly: two unknowns, two equations, one solution. Geometry agrees with arithmetic, and the agreement is not a coincidence.

This article is the visual side of that statement. The widget below puts two draggable dots on a coordinate plane. As you move them, the unique line through both updates live, and its equation appears alongside. Try to draw a second line through the same two points — you cannot, and the widget will not let you fake it. Then collapse the two dots into one and watch the line become indeterminate, because one point is not enough.

The widget

Drag the blue and red dots anywhere on the plane. The black line is the unique straight line through both. Its equation updates on the left; the coordinates of the two points appear on the right. The "collapse" button puts both dots on top of each other — watch the line vanish into "undetermined", because one point alone does not pin a line down.

Drag the blue dot A or the red dot B. The black line through them updates instantly, and the equation in the readout below changes with it. Now try the experiment that proves the claim: keep A and B where they are and try to draw a second straight line through both. You cannot. Any line through A that does not also pass through B misses B; any line through B that does not also pass through A misses A. Only one line threads both — and the widget shows that line and only that line.

Press the collapse button. Both points snap to the same location, and the equation readout reports "undetermined". This is the second half of the result. With one point you have infinitely many lines through it — the line can rotate freely around that single anchor — so a single point gives you no specific line at all.

What the claim says, exactly

The two-point line postulate

For any two distinct points A and B in a plane, there exists one and only one straight line that passes through both.

Two words deserve attention. Distinct — the points must actually be different. If A and B are the same point, the claim collapses, because one point does not determine a line. Exactly one — not "at least one" and not "at most one" but both at once. Existence (you can always draw such a line) and uniqueness (there is no second one) are two separate facts the postulate bundles together.

Why this is a postulate, not a theorem: in his Elements, Euclid took this as one of the five axioms you are simply allowed to assume — "a straight line may be drawn from any one point to any other point." Inside Euclidean geometry there is nothing more basic to derive it from. It encodes what "straight line" means: the unique shortest path between two points. If you change geometries (sphere, hyperbolic plane), the postulate may need to be modified, which is why it had to be stated explicitly rather than proved.

The algebraic side gives an independent argument that points the same way.

Why two points are exactly the right number, algebraically: a non-vertical line in the plane has the form y = mx + c, with two unknowns (m and c). Each point (x_0, y_0) on the line gives one linear equation, y_0 = m x_0 + c. Two distinct points give two distinct equations in the two unknowns, and a system of two independent linear equations in two unknowns has exactly one solution — so m and c are forced. One point gives only one equation, leaving one free parameter (a whole family of lines). Three generic points give three equations in two unknowns — overdetermined, usually with no solution at all.

Why one point is not enough

A single point with many lines through it (ambiguous), two points with one line (unique), three random points usually not collinearThree side-by-side panels. Left panel: a single dot with five lines passing through it at different angles, labelled ambiguous. Middle panel: two dots and the single straight line passing through both, labelled unique. Right panel: three dots scattered such that no straight line passes through all three, labelled usually not collinear. one point: ambiguous infinitely many lines two points: unique exactly one line three points: usually no line collinearity is rare
Left: through a single point pass infinitely many lines — a "pencil" of lines. Middle: two distinct points pin a single line — the goldilocks count. Right: pick three points at random and the chance that they all sit on one straight line is essentially zero; the dashed line shows a candidate that hits two of them and misses the third.

The left panel is what mathematicians call a pencil of lines through a point — every line in the plane that contains A belongs to it, and there are uncountably many such lines (one for every direction). With only A available, geometry has no way to choose between them.

The right panel shows the opposite failure. If three points are placed at random, the chance that all three lie on a single straight line is zero — they are almost certainly non-collinear, and no straight line passes through all three. So three is one too many; the system is over-constrained. That is why surveyors and engineers always quote two reference points, never one, never three.

Worked examples

Through $(1, 2)$ and $(3, 8)$

You want the line through A = (1, 2) and B = (3, 8). The slope is

m = \frac{y_B - y_A}{x_B - x_A} = \frac{8 - 2}{3 - 1} = \frac{6}{2} = 3.

Now use point-slope form anchored at A: y - y_A = m(x - x_A), so

y - 2 = 3(x - 1) \;=\; 3x - 3,

which rearranges to y = 3x - 1. Check the second point: 3(3) - 1 = 8. Both points lie on y = 3x - 1, and the postulate guarantees this line is the only one. Try entering A = (1,2), B = (3,8) in the widget above to see the same equation appear in the readout.

Through $(0, 0)$ and $(5, 10)$

The first point is the origin, which is a freebie — any line through the origin has the form y = mx with no constant term. The slope is

m = \frac{10 - 0}{5 - 0} = 2,

so the line is y = 2x. Check at (5, 10): 2 \cdot 5 = 10. Done. Whenever one of your two points is the origin the algebra gets shorter, because the y-intercept is automatically zero — but the postulate is the same, and there is still exactly one line.

"What if both points are the same?"

Suppose someone hands you "the two points" A = (2, 5) and B = (2, 5). They are the same point. The slope formula gives

m = \frac{5 - 5}{2 - 2} = \frac{0}{0},

which is not a number — it is the algebra screaming that the input is degenerate. Geometrically you have only one point, and the line is not determined: every line through (2, 5) is a valid candidate, so there is no unique answer. This is why the postulate has the word distinct in it. The widget reproduces this exactly: hit the collapse button and the equation readout flips to "undetermined — pick two distinct points".

Where this shows up

In coordinate geometry the two-point fact lets you reconstruct a line from a tiny amount of data — just two ordered pairs. In linear algebra the same fact is the geometric face of "two independent equations in two unknowns have exactly one solution". In physics, drawing a best-fit line through scattered data points uses the same idea in reverse: real data is never perfectly collinear, but you fit the line that comes closest. In carrom, when you aim for a corner pocket you are unconsciously using the two-point principle — your striker's centre and the pocket are two points, and they pin the unique aim line for a straight shot.

The takeaway is short. Two points, one line. One point gives you a pencil; three usually gives you nothing. Exactly two pins the line down, no more, no fewer — and that is why every line problem in school maths starts by handing you exactly that many.

References

  1. Euclid's Elements, Book I, Postulates — the original statement.
  2. Wikipedia: Line (geometry) — modern axiomatisation and the two-point uniqueness statement.
  3. Wikipedia: Collinearity — what it means for three or more points to lie on a single line.
  4. Wikipedia: Pencil (mathematics) — the family of all lines through a single point.