Advanced Math Unit M3

Quadratic Equations

Zero-product cracks them open, one formula solves the rest, and the discriminant counts your answers before you find a single one.

An equation that equals zero confesses — and for the ones that won't, there's a formula that did completing the square once so you never have to again.

Zero is the only number that confesses

Suppose I tell you two hidden numbers multiply to 1212. What are they? Could be 33 and 44. Could be 22 and 66, or 1-1 and 12-12, or 12\tfrac12 and 2424 — you know almost nothing. Now suppose I tell you they multiply to 00. Suddenly you know something for certain: one of them is zero. There is no other way to build a zero product. Twelve keeps secrets; zero confesses.

That one confession is why M2 taught you to factor. An equation like x2+5x+6=0x^2 + 5x + 6 = 0 asks a genuinely hard question — which xx makes this whole expression collapse to nothing? — and factored form answers it almost by accident.

Zero-product: factoring finally gets paid

factor
M2’s hunt: multiply to 66, add to 55 — it’s 22 and 33. So the equation says (x+2)(x+3)=0(x+2)(x+3) = 0.
confess
A product of two things is zero only when one of them is: x+2=0x + 2 = 0 or x+3=0x + 3 = 0.
solve each
Two one-step equations from A1: x=2x = -2 or x=3x = -3.
check
Plug 2-2 back in: 410+6=04 - 10 + 6 = 0 ✓. Two answers — quadratics usually keep a pair.

The SAT often skips the factoring and hands you the confession directly: (x3)(x+5)=0(x - 3)(x + 5) = 0, solve. Read the roots straight off — each parenthesis dies at the value that zeroes it: x=3x = 3 and x=5x = -5. Note both signs flipped. The factor (x3)(x - 3) has root +3+3; answering 3-3 is the wrong answer the test is fishing for.

When the hunt comes up dry

M2 left you with a promise about x2+6x+2x^2 + 6x + 2: no integer pair multiplies to 22 and adds to 66, but completing the square still rewrites it as (x+3)27(x+3)^2 - 7. Watch what that buys you the moment there’s an equals sign:

x2+6x+2=0        (x+3)27=0        (x+3)2=7x^2 + 6x + 2 = 0 \;\;\Rightarrow\;\; (x+3)^2 - 7 = 0 \;\;\Rightarrow\;\; (x+3)^2 = 7

Now the question is one you finished training for in F7: what squares to 77? Two numbers do — 7\sqrt{7} and 7-\sqrt{7}. That’s where the famous ±\pm is born: not from the 7\sqrt{\phantom{7}} symbol (which always means the positive root), but from the equation, because both signs square to the same thing:

x+3=±7        x=3±7x + 3 = \pm\sqrt{7} \;\;\Rightarrow\;\; x = -3 \pm \sqrt{7}

Two exact answers, no factoring required. And nothing about this move needed the numbers to be friendly — completing the square solves any quadratic. Which raises a beautifully lazy question: if the recipe always works, why do we keep running it from scratch?

The formula: completing the square, done once, forever

Run the exact same recipe on ax2+bx+c=0ax^2 + bx + c = 0 — letters left in, nothing else new — and out comes the recipe’s own answer, prepackaged:

That’s all the quadratic formula is: completing the square, performed once in general so nobody ever has to do it again. Watch the whole build happen at the board — every move named as it is made, from ax2+bx+c=0ax^2 + bx + c = 0 to the boxed formula:

Press play — the board writes itself.

0:00 / 6:34
Transcript

You've probably seen this formula before. x equals negative b, plus or minus the square root of b squared minus four a c, all over two a. Most people memorize it without ever knowing where it comes from. Today we're going to build it, step by step — because once you've watched it come together, you don't have to trust it. You'll know why it works.

Here's our starting point: a x squared, plus b x, plus c, equals zero. The letters a, b, and c stand for any numbers you like. That's the whole idea — if we can solve this equation once, with letters, we've solved every quadratic equation at the same time. Our tool is a technique called completing the square. If you've met it before, this will feel familiar — and if you haven't, no problem at all: we're going to do every single step together.

Step one. Completing the square works best when x squared stands alone, with no coefficient in front. Right now it has an a. So we divide both sides — every single term — by a. a x squared divided by a is x squared. b x divided by a is b over a, times x. c divided by a is c over a. And zero divided by a is still zero. Same equation, same solutions — just easier to work with.

We're ready to complete the square, and the recipe starts with the coefficient in front of x — here, that's b over a. Take half of it. Half of b over a is b over two a. Keep an eye on that expression — it's about to do all the work.

The second step is to square that half: b over two a, squared, is b squared over four a squared. Here's why we want it: if we add this exact amount, the first terms become a perfect square. But we can't just add something to one side of an equation — that would change the equation. So we do it honestly: add it, and subtract it, in the same line. Adding and subtracting the same amount changes nothing at all.

The payoff is sitting in the first three terms: x squared, plus b over a x, plus b squared over four a squared. Together, they are exactly x plus b over two a, squared. If you're not sure, multiply it out and check — square the first, two times the product, square the last. The left side has become a perfect square, with two leftover terms trailing behind it.

Time to tidy up. The two leftovers — minus b squared over four a squared, and plus c over a — don't belong with the square. Move each one to the right side; each flips its sign as it crosses the equals sign. Now we have: x plus b over two a, squared, equals b squared over four a squared, minus c over a.

The right side is two fractions, so let's combine them into one. They need a common denominator. Multiply c over a by four a over four a — that's multiplying by one, so its value doesn't change — and it becomes four a c over four a squared. Now subtract the numerators: b squared minus four a c, all over four a squared. Take a good look at that top — b squared minus four a c. It's important enough to have its own name, and it can tell you how many solutions an equation has before you even solve it. We'll come back to it.

Now we take the square root of both sides — and this is where most mistakes happen, so let's make the mistake on purpose. You might write: x plus b over two a equals the square root of b squared minus four a c, over two a. It looks finished. But remember: if something squared equals nine, that something could be three — or negative three. Both work. A square has two square roots, one positive and one negative, and this line only kept one of them. It's missing half the answers — so it comes off the board.

The correct line carries a plus-or-minus sign: x plus b over two a equals plus or minus the square root of b squared minus four a c, over two a. One small symbol — and it's the difference between finding both solutions and silently losing one.

We have one step left, and it's the easiest one: get x by itself. Subtract b over two a from both sides. Both pieces already share the denominator two a, so they combine into a single fraction. And there it is: x equals negative b, plus or minus the square root of b squared minus four a c, all over two a. That's the quadratic formula — and this time you didn't memorize it, you built it.

Step back and look at what we actually did. We took the general equation and completed the square, once. That's the whole secret — the formula is completing the square, done in advance, for every equation at the same time. The plus-or-minus is the square root keeping both answers. And b squared minus four a c, sitting under the root, is the part that decides how many solutions there are — more on that soon. From here on, solving a quadratic is reading off a, b, and c.

The quadratic formula, built from scratch · 6:35

Every piece has a birthplace — the b-b and 2a2a come from the “half of the middle” fold, and b24acb^2 - 4ac is the square’s leftover. Use it on the trinomial M2 proved unfactorable, x2+3x+1=0x^2 + 3x + 1 = 0: here a=1a = 1, b=3b = 3, c=1c = 1, so

x=3±324(1)(1)2(1)=3±52x = \dfrac{-3 \pm \sqrt{3^2 - 4(1)(1)}}{2(1)} = \dfrac{-3 \pm \sqrt{5}}{2}

The dead end from M2’s slider — cracked in two lines.

Simplify the answer — F7 was training for this exact moment

Solve x2+4x8=0x^2 + 4x - 8 = 0. The formula gives b24ac=16+32=48b^2 - 4ac = 16 + 32 = 48, so

x=4±482x = \dfrac{-4 \pm \sqrt{48}}{2}

F7 taught you to never leave 48\sqrt{48} unopened — it deliberately stopped at 48=43\sqrt{48} = 4\sqrt{3} and called the next move M3’s. This is that move:

x=4±432=2(2±23)2=2±23x = \dfrac{-4 \pm 4\sqrt{3}}{2} = \dfrac{2\left(-2 \pm 2\sqrt{3}\right)}{2} = -2 \pm 2\sqrt{3}

Look hard at the middle step. Cancelling a fraction divides factors of the whole top — and 4±43-4 \pm 4\sqrt{3} is a sum, not a product, so you must factor the 22 out of every term before it may cancel. The tempting shortcut — cancel the 4-4 against the 22, leave the 434\sqrt{3} untouched, write 2±43-2 \pm 4\sqrt{3} — feels natural precisely because it works on single-term fractions. Here it changes the answer, and the SAT lists it among the choices every time.

Try the move yourself — pick divisors and watch which terms cooperate:

x² +x += 0

The root cleans up first (F7's move): .

The formula hands you this — is it done?

divide every term by…
Reduce the formula's answer by hand

Predict before you click: on the x2+4x8x^2+4x-8 preset, will ÷3 go through? Decide which term refuses, then click it and see. Then load 4x24x14x^2 - 4x - 1 — the top and bottom share a 44: does it matter whether you take it in one ÷4 bite or two ÷2 bites? And x2+3x+1x^2 + 3x + 1 ends somewhere important: with nothing to divide. 3±52\tfrac{-3 \pm \sqrt{5}}{2} is finished — a denominator is not dirt, and recognizing “done” is worth points.

The discriminant: count the answers without finding them

Here’s the SAT’s favorite twist: “How many real solutions does 2x25x+7=02x^2 - 5x + 7 = 0 have?” — and solving it would be falling for the trap. Look at the formula’s anatomy. The b-b and the 2a2a are fixed scaffolding; the only part that can change how many answers exist is the piece under the root, b24acb^2 - 4ac, the discriminant:

  • b24ac>0b^2 - 4ac > 0: the root is a real, positive number — the ±\pm spreads two different answers.
  • b24ac=0b^2 - 4ac = 0: the root is 00 — the ±\pm adds and subtracts nothing — exactly one answer.
  • b24ac<0b^2 - 4ac < 0: you’d need negative\sqrt{\text{negative}}, and F7 said it straight: no real number squares to a negative. No real solution — that’s a complete, correct answer, not a failure.

So: (5)24(2)(7)=2556=31<0(-5)^2 - 4(2)(7) = 25 - 56 = -31 < 0. Zero real solutions, ten seconds, no solving. (When you meet parabolas in M4, you’ll see these three cases as a curve crossing, touching, or missing the xx-axis. For now the number line tells the whole story.)

x² +x + c = 0
drag c and watch the twins
−1212c = -3

the deciding number:

2 two real solutions

-4-3-2-1012midpoint-31

Positive fuel under the root — the ± spreads two answers apart.

Wherever the twins sit, they average to — the ± spreads them evenly around . That symmetry is why the sum of solutions never needs solving.

Watch the answers appear and vanish

Predict before you drag: the opening preset is x2+2x3=0x^2 + 2x - 3 = 0, twins at 3-3 and 11. Drag cc upward one step at a time — each step drains the discriminant by 44. At what cc do the twins kiss? Check with algebra: b24ac=44c=0b^2 - 4ac = 4 - 4c = 0 at c=1c = 1. That drag was the SAT’s “find kk so the equation has exactly one solution” question — your hand solved b24ac=0b^2 - 4ac = 0 before your pencil did. Then load x26x+7x^2 - 6x + 7 and read the twins’ exact labels: 3±23 \pm \sqrt{2}, the formula’s answers living on the line. And notice, through every drag: the twins never leave their balance point b2a-\tfrac{b}{2a} — the ±\pm spreads them evenly around it.

Sum and product: answers about the answers

That symmetry has a cash value. Add the formula’s two answers and the ±\pm halves kill each other:

b+D2a+bD2a=2b2a=ba\dfrac{-b + \sqrt{D}}{2a} + \dfrac{-b - \sqrt{D}}{2a} = \dfrac{-2b}{2a} = -\dfrac{b}{a}

So when the SAT asks “what is the sum of the solutions of 2x28x+3=02x^2 - 8x + 3 = 0?”, it is testing whether you’ll spend ninety seconds solving — or read ba=82=4-\tfrac{b}{a} = \tfrac{8}{2} = 4 off the coefficients. (The product collapses the same way, to ca\tfrac{c}{a}.) The answers were never hiding; they were encoded in the equation the whole time — the same lesson the pair hunt taught in M2, one floor up.

The one thing to remember

Get zero alone on one side — only zero confesses. If the left side factors, each factor hands you a root. If it doesn’t, the formula is completing the square already done for you: read aa, bb, cc with their signs, simplify the root the F7 way, and reduce every term or none. And when the question is only how many, don’t solve — the discriminant b24acb^2 - 4ac knew before you did.

The decision tree

SeeDoExample
factored =0= 0read each root (flip the sign)(x3)(x+5)=0x=3,5(x-3)(x+5)=0 \to x = 3,\, -5
it factorszero-product: set each factor =0= 0x2+5x+6=0x=2,3x^2+5x+6=0 \to x = -2,\, -3
(x+h)2=k(x+h)^2 = ksquare root both sides, ±\pm appears(x+3)2=7x=3±7(x+3)^2 = 7 \to x = -3 \pm \sqrt{7}
nothing factorsthe quadratic formulax2+3x+1=0x=3±52x^2+3x+1=0 \to x = \tfrac{-3 \pm \sqrt{5}}{2}
”how many solutions?“discriminant only — don’t solveb24acb^2 - 4ac, read the sign
”sum/product of solutions?”Vieta — don’t solvesum ba-\tfrac{b}{a}, product ca\tfrac{c}{a}

The discriminant D=b24acD = b^2 - 4ac:   D>0\;D > 0 → two real solutions.   D=0\;D = 0 → exactly one.   D<0\;D < 0 → no real solution. “Exactly one solution” in a question is code for set b24ac=0b^2 - 4ac = 0 and solve for the unknown coefficient.

Simplest radical form: simplify the root (48=43\sqrt{48} = 4\sqrt{3}), then divide all three of b-b, the root’s coefficient, and 2a2a by their common factor.

x² +x + c = 0
drag c and watch the twins
−1212c = -3

the deciding number:

2 two real solutions

-4-3-2-1012midpoint-31

Positive fuel under the root — the ± spreads two answers apart.

Wherever the twins sit, they average to — the ± spreads them evenly around . That symmetry is why the sum of solutions never needs solving.

x² +x += 0

The root cleans up first (F7's move): .

The formula hands you this — is it done?

divide every term by…
Solve: . Give every solution.

Factor the left side with M2’s hunt, then set each factor to zero.

Correct: 0Attempts: 0Streak: 0Best: 0