Every quadratic equation reduces to one question: what values of x make this expression equal to zero? The quadratic formula answers it for any quadratic, always, whether or not the factors are tidy. It is also the one piece of algebra that has a real claim to being useful forever — the same formula times a launch, a break-even point and an intersection of two curves.
This article derives it rather than presenting it, works three examples you can check by hand, and then covers the thing the textbook usually skips: what the discriminant means before you take a single square root.
Starting from a quadratic
A quadratic equation is any equation of the form
ax² + bx + c = 0
where a is not zero (if it were, the equation would be linear and there would be nothing quadratic about it). That single constraint is worth remembering, because the whole formula divides by 2a and a calculator that accepts a = 0 is handing you a division by zero dressed up as an answer.
Any quadratic can be written in this form, including the ones that arrive without an x² term visible. The equation 3 = x + 2x² is a quadratic in disguise: rearrange to 2x² + x − 3 = 0 and it is ready for the formula.
Deriving it by completing the square
Start from ax² + bx + c = 0 and divide everything by a:
x² + (b/a)x = −c/a
To turn the left side into a perfect square, we need to add (b/2a)² to it — because a square of the form (x + k)² expands to x² + 2kx + k², and we already have a 2kx term where k = b/2a. Adding the same amount to both sides keeps the equation balanced:
x² + (b/a)x + (b/2a)² = −c/a + (b/2a)²
Now the left side factors perfectly:
(x + b/2a)² = (b² − 4ac) / 4a²
Take square roots of both sides. The ± matters here: a square root has two values, and dropping the ± is the single most common error in the whole topic.
x + b/2a = ±√(b² − 4ac) / 2a
Subtract b/2a from both sides and you have the formula, arrived at by operations that are all reversible — which is exactly why it works for every quadratic, including the ones that do not factor nicely.
Example one: two real roots
Solve x² − 5x + 6 = 0. Here a = 1, b = −5, c = 6.
Discriminant: b² − 4ac = 25 − 24 = 1.
x = (5 ± √1) / 2 = (5 ± 1) / 2, giving x₁ = 3 and x₂ = 2.
Check by factoring: (x − 3)(x − 2) = x² − 5x + 6. ✓ Both roots verified, and because the factors are integers, factoring was the faster route — a useful reminder that the formula is the reliable method, not always the quickest one.
Example two: one repeated root
Solve x² − 6x + 9 = 0. Here a = 1, b = −6, c = 9, so the discriminant is 36 − 36 = 0.
x = (6 ± 0) / 2 = 3. One root, counted twice.
Factoring confirms it: (x − 3)². Graphically this is a parabola whose vertex sits exactly on the x-axis — it touches the axis at (3, 0) and turns around there. This is also the boundary case that matters in optimisation: when a business problem's discriminant is exactly zero, the break-even point is precisely the maximum or minimum.
Example three: no real roots
Solve x² + 2x + 5 = 0. Here a = 1, b = 2, c = 5, so b² − 4ac = 4 − 20 = −16.
√(−16) is not a real number, so the two solutions are complex: x = −1 ± 4i. The parabola y = x² + 2x + 5 has its vertex at (−1, 4) — always above the x-axis — so it genuinely never crosses.
This is where the discriminant earns its keep. Before computing anything, the negative discriminant told you the answer shape. In a word problem, that is the difference between "no sensible answer exists" and going down a rabbit hole of arithmetic.
The vertex is hiding in the same formula
Rewriting a quadratic in vertex form, y = a(x − h)² + k, gives the turning point directly: h = −b/(2a), and k is whatever the polynomial evaluates to at h. Because a square is never negative, the vertex is the minimum when a > 0 and the maximum when a < 0.
For the first example: h = 5/2 = 2.5, and 2.5² − 5(2.5) + 6 = −0.25, the lowest point of that curve. Every real use of a quadratic — pricing, reaction rates, projectile paths, maximising revenue — is really asking for this vertex. The formula and the derivative agree, and at the point where they do, algebra and calculus stop feeling like different subjects.
Three mistakes worth naming
1. Dropping the ±. A quadratic has two roots; taking only the plus sign throws one away. If you only need one, say which one and why.
2. Forgetting the sign of b. The formula uses −b, not b. With b = −5, −b is +5; typing b directly is a silent error that still returns plausible-looking numbers.
3. Setting a = 0. The equation stops being quadratic. Catch this before the formula divides by zero.
Numerical stability: the form nobody teaches
There is a subtlety that bites in real work. For large b and small c, the two roots are −b and −c/b — and if the discriminant is nearly zero, the standard formula suffers catastrophic cancellation: subtracting two almost-equal numbers, losing most of the significant digits. The fix is to use the algebraically equivalent form
x = 2c / (−b ∓ √(b² − 4ac))
where you take the sign that keeps the numerator and denominator small. This is why some calculators give you one root reliably and the other not at all. If you are fitting curves to data rather than solving homework, this matters more than anything else in this article.
Where quadratics show up
The break-even point of a business is a quadratic root: set revenue equal to cost, and the solutions are the sales volumes where profit crosses zero. The maximum of any parabola is its vertex, which is how pricing is optimised — a classic derivation that needs no calculus. In physics, the flight time of a projectile follows from solving a quadratic, and so does the range. In finance, the time to repay a balance under a payment schedule reduces to solving one of these. Seeing a quadratic in the wild is more common than it looks.
Worth knowing: the quadratic formula extends to the complex numbers, and over the complex field every quadratic has exactly two roots. The discriminant is still the useful test — a non-zero discriminant guarantees two distinct roots, a zero discriminant guarantees a repeated one, and a zero discriminant is the defining feature of a defective matrix.
For arithmetic, the quadratic equation solver returns both roots with the discriminant and vertex alongside, so you can check your working without retyping the numbers.
Frequently asked questions
Why does the quadratic formula work?
Because it is completing the square in disguise. Starting from ax² + bx + c = 0, dividing by a and moving the constant gives x² + (b/a)x = −c/a. Adding (b/2a)² to both sides makes the left a perfect square, and the square roots that follow produce the formula. Nothing is assumed — the algebra is reversible at every step.
What does the discriminant tell me before I solve?
It predicts how many real roots exist. b² − 4ac > 0 means two distinct real roots, = 0 means one repeated root, and < 0 means no real roots — the two solutions are complex conjugates. In optimisation problems it is also the quickest test for whether a real, sensible answer exists.
What if a is zero?
Then it is not a quadratic equation. The x² term vanishes and you are left with a linear equation bx + c = 0, solved by x = −c ÷ b. This is why the formula divides by 2a.
Do I need the formula if I can factor?
Factoring is faster when the factors are integers, and you should use it. The formula is the reliable route when factoring is awkward, the roots are irrational, or you only need one root — then use the stable form x = 2c / (−b ∓ √(b² − 4ac)) to avoid cancellation error.