You rewrite a quadratic function in vertex form by taking the coefficient of x, halving it, squaring the half, and adding and subtracting that square inside the expression so the first three terms collapse into a perfect square. The vertex sits at (h, k) = (−b/2a, c − b²/4a) for y = ax² + bx + c, and every step below is built to land on those two numbers. The method works whether the leading coefficient a is 1 or not, and whether the numbers come out clean or fractional.

Read the Standard Form and Set Up the Rewrite

Every quadratic you will handle arrives as y = ax² + bx + c, and the three coefficients tell you what kind of answer to expect before you lift a pencil. The sign of a decides which way the parabola opens, so a positive a means the vertex is a minimum and a negative a means it is a maximum. The ratio −b/2a is the x-coordinate of the vertex, and it exists for every quadratic where a is not zero. If b is zero, the vertex lands on the y-axis and the rewrite is shorter than you might think.

Write the expression with the x² and x terms grouped and leave a visible gap where the completing term will go. Keep the constant c outside the grouping, because it will rejoin the expression at the end. For y = 2x² − 12x + 7, you write 2(x² − 6x ) + 7 and treat the empty space as a slot you are about to fill. That slot is the single most useful habit in this method, since it stops you from losing track of what you added.

Check the coefficient on x² before anything else. When a equals 1 you can skip the factoring step entirely, which saves time and a common error. When a is not 1, every term inside the parentheses must be divided by a, including the b term, and that division is where most mistakes happen. Rewriting 3x² + 18x − 4 as 3(x² + 6x ) − 4 is correct; writing 3(x² + 18x ) − 4 is not.

Confirm the sign of each coefficient by reading it directly from the function rather than from memory. A minus sign in front of the b term means b itself is negative, so −b/2a becomes positive when you divide. Sign slips account for a large share of wrong vertices, and a two-second check each time prevents them cleanly.

Halve the x Coefficient, Square It, and Balance Both Sides

Take the number sitting in front of x inside the parentheses, divide it by two, and square the result. For y = 2x² − 12x + 7, factoring gives 2(x² − 6x ) + 7, so the number in front of x is −6, half of it is −3, and its square is 9. That 9 is the completing term, and it goes into the gap. This single calculation determines the vertex's x-coordinate, since h always equals the halved value with the sign flipped back.

Balance is where the leading coefficient matters, and this is the part most textbook treatments underplay. Adding 9 inside the parentheses adds 9 to the grouped expression, but because the group is multiplied by 2, the whole function actually gains 18. To keep the value unchanged you subtract 18 outside the parentheses, not 9. The rule: subtract a times the completing term, where a is whatever sits in front of the parentheses.

For y = x² + 10x + 3, the coefficient of x is 10, half is 5, and the square is 25. Adding 25 inside and subtracting 25 outside gives y = (x² + 10x + 25) + 3 − 25, which simplifies to y = (x + 5)² − 22. The vertex is (−5, −22), and both numbers fall straight out of the arithmetic you just did.

Failure case worth knowing: if you add the completing term when a is fractional, the balancing term changes again. For y = ½x² + 4x + 1, factoring gives ½(x² + 8x ) + 1, the completing term is 16, and you subtract ½ × 16 = 8 outside. Add too much or too little and the parabola's vertex drifts vertically while the shape stays the same, which is the sneakiest kind of error because the graph still looks plausible.

Factor the Perfect Square Trinomial Into Squared Binomial Form

The three terms inside the parentheses are now a perfect square trinomial, and they always factor to the squared binomial whose constant is half the original x coefficient. For x² − 6x + 9, that is (x − 3)², because the halved value was −3. For x² + 8x + 16, that is (x + 4)², and for x² + 10x + 25, (x + 5)². The sign inside the binomial matches the sign of the original b term inside the parentheses.

Write the full vertex form with the factored group and the constant that remains from the balancing step. Finishing the y = 2x² − 12x + 7 example: 2(x² − 6x + 9) + 7 − 18 becomes 2(x − 3)² − 11. The vertex is (3, −11), which you can read directly without calculating anything further. The coefficient 2 stays in front of the square, and it controls how wide or narrow the parabola is.

Verify by expanding back to standard form, which takes about fifteen seconds and catches every sign and balancing error. Expanding 2(x − 3)² − 11 gives 2(x² − 6x + 9) − 11, which is 2x² − 12x + 18 − 11, and that returns 2x² − 12x + 7 exactly. If your expansion does not match the original, the mistake is almost always in the balancing term, not in the factoring.

One misconception to retire: the vertex is not always at (h, 0). The k value comes from the balancing arithmetic, not from the inside of the square, and it is almost never zero unless the original constant happens to land there. Students who assume the vertex sits on the x-axis produce answers that fail the expansion check every time.

Derive the Vertex Coordinates Without Graphing

You do not need a graph, a table, or a calculator to get (h, k), because the standard-form coefficients hand them over through two short formulas. The x-coordinate is h = −b/2a, and the y-coordinate is k = c − b²/4a, which you can also get by substituting x = h back into the original function. For y = 3x² + 18x − 4, h = −18/6 = −3. Then k = 3(−3)² + 18(−3) − 4 = 27 − 54 − 4 = −31.

Confirm that the formula version and the completing-the-square version agree, since they must. The completed form for that example is 3(x + 3)² − 31, and the vertex reads as (−3, −31), matching the formula output exactly. Agreement between two routes is your proof the arithmetic is right, and it takes less than a minute.

The discriminant b² − 4ac, computed inside the k formula, tells you where the vertex sits relative to the x-axis. When b² − 4ac is positive, the parabola crosses the axis twice and the vertex is above or below that crossing depending on the sign of a. When it is zero, the vertex touches the axis exactly once, and that single point is also the only real root.

Watch the sign of a when interpreting the vertex. A positive a makes the vertex a minimum, so k is the lowest y-value the function reaches. A negative a makes it a maximum, so k is the highest y-value. That distinction changes how you read every answer, and it matters more than the coordinates themselves for graphing and optimization questions.

Worked Example One: Convert a Function With a Leading Coefficient of One

Start with y = x² − 4x + 9, where the leading coefficient is already 1 and no factoring is needed. The x coefficient is −4, half is −2, and the square is 4. Add 4 inside and subtract 4 outside: y = (x² − 4x + 4) + 9 − 4, which collapses to y = (x − 2)² + 5. The vertex is (2, 5), and because a is positive, that point is the minimum of the function.

Check the result with the formula route. Here b = −4, a = 1, c = 9, so h = −(−4)/2 = 2 and k = 9 − 16/4 = 5. Both routes give (2, 5), and the expansion check from vertex form returns x² − 4x + 4 + 5 = x² − 4x + 9, matching the original. Three independent confirmations for a problem that took under a minute.

Try a version where the constant is negative, since subtraction trips people up more than addition. For y = x² + 6x − 1, the halved value is 3, the square is 9, and the rewrite is y = (x + 3)² − 1 − 9 = (x + 3)² − 10. The vertex is (−3, −10). A quick warning: keep the −9 separate from the −1, because confusing the two is the most frequent slip when the constant is already negative.

Read the vertex form for shape information while you have it. The squared term dominates at large |x|, so the parabola opens upward here with axis of symmetry x = −3. The y-intercept comes from setting x = 0 in the original function, giving −1, and plugging x = 0 into the vertex form gives 9 − 10 = −1, which is the same number from a different direction.

Worked Example Two: Handle a Leading Coefficient Arrow Than One

Take y = 2x² − 16x + 5, where the leading coefficient is 2 and the factoring step is mandatory. Factor: 2(x² − 8x ) + 5. The x coefficient inside is −8, half is −4, and the square is 16, so the gap gets 16. The grouped expression gains 2 × 16 = 32, so you subtract 32 outside: 2(x² − 8x + 16) + 5 − 32. That factors to 2(x − 4)² − 27, and the vertex is (4, −27).

Run the formula check. With a = 2, b = −16, c = 5: h = 16/4 = 4 and k = 5 − 256/8 = 5 − 32 = −27. Both routes land on (4, −27). Expanding 2(x − 4)² − 27 gives 2(x² − 8x + 16) − 27 = 2x² − 16x + 32 − 27 = 2x² − 16x + 5, which is the original function exactly.

If the leading coefficient is negative, factor the negative out and keep track of the sign. For y = −3x² + 12x + 1, factor as −3(x² − 4x ) + 1. The completing term is 4, and the balancing subtraction becomes minus (−3 × 4) = plus 12 outside: −3(x² − 4x + 4) + 1 + 12. That is −3(x − 2)² + 13, with vertex (2, 13), which is a maximum because a is negative.

If the leading coefficient is not a clean integer, do not round. For y = ½x² + 3x − 2, factoring gives ½(x² + 6x ) − 2, the completing term is 9, and the balancing subtraction is ½ × 9 = 4.5. The rewrite is ½(x + 3)² − 2 − 4.5 = ½(x + 3)² − 6.5, with vertex (−3, −6.5). Rounding 4.5 to 5 shifts the vertex by half a unit, which is visible on any decent graph.

Vertex Form vs Standard Form: What Each One Tells You Fast

Each form has a distinct job, and switching between them is faster than trying to read everything off one. Standard form y = ax² + bx + c exposes the y-intercept instantly at c, which vertex form never does directly. Vertex form y = a(x − h)² + k exposes the vertex and the axis of symmetry instantly, which standard form hides behind arithmetic. The table below lines up the trade-offs.

Feature Standard form y = ax² + bx + c Vertex form y = a(x − h)² + k
Vertex Requires −b/2a and substitution Read directly as (h, k)
Axis of symmetry Requires calculation at x = −b/2a Read directly as x = h
y-intercept Read directly as c Requires expanding or substituting x = 0
Direction of opening Sign of a Sign of a
Maximum or minimum value Requires vertex computation The value k, with max or min set by the sign of a
Number of real roots Discriminant b² − 4ac Solvable by isolating the square and taking square roots
Ease of graphing Needs multiple sample points Two anchor points from the vertex

Graphing is the biggest practical difference. From standard form, you need five or six sample points or a symmetry calculation before the shape is clear. From vertex form, the vertex goes down first, the axis of symmetry is a mirror line, and two points on either side finish the curve. That is roughly three plotted points instead of six, and the symmetry is exact rather than estimated.

Root-finding favors vertex form when the numbers co-operate. Setting (x − h)² = −k/a and taking the square root gives exact solutions without the quadratic formula. When −k/a is negative, the function has no real roots, and the sign flip tells you immediately instead of burying the domain question in a discriminant calculation.

Choose based on the question: intercepts from standard form, vertex and graph from vertex form. Converting costs one short arithmetic sequence, so there is no reason to stay in a form that hides the feature you need. This is why algebra courses routinely ask students to convert rather than just evaluate at points.

Spot the Balancing Error Before It Cancels the Whole Answer

The most common error with a leading coefficient above 1 is subtracting the completing term itself instead of a times the completing term. For y = 3x² + 6x + 1, the completing term is 1 inside, but the balancing subtraction must be 3, because the group is multiplied by three. Subtract only 1 and the vertex sits at −1 + 1 = 0 instead of the true −2, which is a full two units off.

Sign errors rank second. When the coefficient of x is negative, half of it is negative and the squared binomial has a minus inside, as in (x − 3)². When the coefficient is positive, the binomial has a plus, as in (x + 4)². Flipping that sign moves the vertex from (3, k) to (−3, k), a mirror image on the wrong side of the y-axis.

Forgetting to divide the b term by a when factoring is the third frequent mistake. Writing 2(x² − 12x ) instead of 2(x² − 6x ) for the original 2x² − 12x produces a completing term four times too large and a vertex in the wrong place. Divide every term by the leading coefficient, and the b term is no exception.

The fixing procedure is the same every time: re-expand the proposed vertex form and compare it term by term to the original. If the x term matches but the constant differs, your balancing subtraction was wrong. If the x term itself is off, you factored incorrectly at the start. One expansion catches both classes of error, and it is faster than re-reading your own algebra.

Use the Finished Form to Graph, Solve, and Optimize

Graph y = (x − 2)² + 5 by plotting the vertex at (2, 5), then moving one unit right to (3, 6) and one unit left to (1, 6), since the square gives 1 when x is one step from h. Two more steps each way gives (4, 9) and (0, 9). Five points and a smooth curve finish the graph, and the axis of symmetry is the vertical line x = 2. No table, no substitution marathon.

Solve (x − 2)² + 5 = 0 for the roots, if any exist. Isolating the square gives (x − 2)² = −5, and since a real square cannot be negative, this parabola never crosses the x-axis. That conclusion takes two lines from vertex form. From standard form, you would compute the discriminant 16 − 36 = −20 to reach the same place, which is more work for the same answer.

Optimize by reading the vertex as the extreme value. If a company's profit model is P = −2x² + 40x − 150, completing the square gives −2(x − 10)² + 50, so maximum profit is 50 at x = 10. Completing the square is not a classroom ritual; it is the fastest route from a cost or profit model to a decision about the optimal quantity.

Handle fractional and decimal outputs carefully in this context. For P = −0.5x² + 3x + 1, the completed form is −0.5(x − 3)² + 5.5, so the optimum is 5.5 at x = 3. Keep the decimal exact to the model's units, and only round at the final reporting step, because early rounding moves the reported optimum.

Practice Set With Answers Checked Two Ways

Work these six problems with the method above and verify each against the formula h = −b/2a and k = c − b²/4a. The set moves from a leading coefficient of 1 into fractions and negatives, which is the order that builds the skill fastest. Every answer in the table has been checked by expansion, so any disagreement means the arithmetic slipped somewhere.

Function Vertex form Vertex
y = x² + 8x + 1 y = (x + 4)² − 15 (−4, −15)
y = x² − 10x + 20 y = (x − 5)² − 5 (5, −5)
y = 2x² + 8x + 3 y = 2(x + 2)² − 5 (−2, −5)
y = −x² + 6x − 2 y = −(x − 3)² + 7 (3, 7)
y = 3x² − 12x + 8 y = 3(x − 2)² − 4 (2, −4)
y = ½x² − 2x + 1 y = ½(x − 2)² − 1 (2, −1)

Check one row by hand before trusting the table. For y = 2x² + 8x + 3: factor to 2(x² + 4x ) + 3, half of 4 is 2, square is 4, balance subtracts 2 × 4 = 8, giving 2(x + 2)² + 3 − 8 = 2(x + 2)² − 5. Expanding returns 2(x² + 4x + 4) − 5 = 2x² + 8x + 8 − 5 = 2x² + 8x + 3, which matches. The formula route gives h = −8/4 = −2 and k = 3 − 64/8 = 3 − 8 = −5, agreeing.

The fractional row is the one to slow down on. For y = ½x² − 2x + 1: factor to ½(x² − 4x ) + 1, half of −4 is −2, square is 4, balance subtracts ½ × 4 = 2, giving ½(x − 2)² + 1 − 2 = ½(x − 2)² − 1. The vertex is (2, −1), and the formula route gives h = 2/1 = 2 and k = 1 − 4/2 = −1. Both agree, and the decimal-free nature of the answer is a sign the work is clean.

FAQ

What is the difference between vertex form and standard form of a quadratic?

Vertex form is y = a(x − h)² + k, where the vertex reads directly as (h, k) and the axis of symmetry is x = h. Standard form is y = ax² + bx + c, where the y-intercept is c and the vertex requires −b/2a. Converting between them is one arithmetic sequence, and most exam questions allow either form as long as the vertex is correct.

Does completing the square work when the leading coefficient is negative?

Yes, and the balancing rule adapts cleanly. For y = −3x² + 12x + 1, factoring gives −3(x² − 4x ) + 1, the completing term is 4, and the balancing subtraction becomes plus 12 because −3 × 4 is −12 and you subtract it. The completed form is −3(x − 2)² + 13, with a maximum at (2, 13) rather than a minimum.

How do I find the vertex without completing the square?

Use h = −b/2a and k = c − b²/4a directly on the standard form. For y = 4x² − 8x + 3, h = 8/8 = 1 and k = 3 − 64/16 = −1, so the vertex is (1, −1). This formula route matches the completed form 4(x − 1)² − 1 and is faster when you only need the coordinates rather than the full rewrite.

What happens if the completing term is not a perfect square?

The method still works and the answer turns into a radical or decimal. For y = x² + 3x + 1, half of 3 is 1.5, the square is 2.25, and the rewrite is (x + 1.5)² − 1.25, with vertex (−1.5, −1.25). The vertex form does not require integer outputs, and the expansion check still proves the result correct.

Can a quadratic have no real roots, and how does vertex form show it?

Yes. If the completed form is a(x − h)² + k and −k/a is negative, setting the expression to zero produces a negative on the right after isolating the square. That function never crosses the x-axis, so the discriminant b² − 4ac is negative and the quadratic has complex roots only. For y = (x − 2)² + 5, isolating gives (x − 2)² = −5, which has no real solution.

Why do I subtract a times the square rather than just the square?

Because the completing term is added inside a group that is multiplied by the leading coefficient a, so the function's value rises by a times that term. For y = 2(x² − 6x + 9) + 7, the extra 18 must be removed outside as −18, and 2 × 9 = 18. Subtracting only 9 leaves the vertex 9 units too low for this example.

Is vertex form useful outside algebra class?

It appears in business profit models, physics projectile equations, and any optimization problem where a quadratic represents cost, height, or profit. For example, a revenue model R = −5x² + 40x gives completed form −5(x − 4)² + 80, so the maximum revenue is 80 at x = 4. The form converts a messy quadratic into a direct statement of the best case.