Linear vs Quadratic vs Cubic Functions: Key Differences

What actually makes a parabola curve, and a straight line not?

They're both just equations. Plug in $x$, get $y$. But a quadratic feels bent in a way a line never is, and a cubic feels bent in yet another way, with an S-shaped wiggle a parabola can't do.

Here's the thing you should stare at: not the shape, but the rate of change of the rate of change. And how many times you have to ask that question before the answer goes flat.

That number is the degree. Once you know it, you know almost everything about the curve, before you've even graphed it.

Picture a car, not a curve

Imagine three cars leaving the same starting line, each recording distance traveled against time.

Car L cruises at a fixed speed, cruise control locked. Car Q floors the gas and holds it there, so it keeps accelerating at a steady rate. Car C is trickier: it presses the gas harder and harder as it goes, so even its acceleration keeps climbing.

Car L's distance is $d = 5t$. That's linear. Car Q's is $d = 0.4t^2$. That's quadratic. Car C's is $d = 0.045t^3$. That's cubic.

Now forget the odometer. Check each car's speedometer instead.

Drag the marker along the timeline. Car L's speed reading never budges, it's degree one, a single constant number. Car Q's speed climbs at a steady rate, degree two, its rate of change is itself a straight line. Car C's speed rockets, and not steadily, it climbs faster and faster, degree three, the rate of change of the rate of change is the constant thing here.

That's the whole idea behind degree: it counts how many layers of "rate of change" you have to peel back before you hit something constant.

We'll lean on this car analogy for the rest of the post.

Naming the shapes

Now that we've felt the difference, let's name it properly.

A linear function looks like $y = mx + b$. One power of $x$, straight line, constant slope $m$.

A quadratic function looks like $y = ax^2 + bx + c$. Two powers of $x$ at most. Its graph bends into a bowl or a dome, called a parabola.

A cubic function looks like $y = ax^3 + bx^2 + cx + d$. Three powers of $x$ at most. Its graph can bend twice, which gives it that signature S-shaped wiggle.

The highest power of $x$ that shows up is the function's degree. Linear is degree 1. Quadratic is degree 2. Cubic is degree 3. Everything else, how many times the curve can turn, how many times it can cross zero, who wins as $x$ grows huge, falls out of that single number.

How many times can it turn?

Here's a question you can answer without touching any algebra: how many times can a curve turn around, switching from climbing to falling or back again?

A line never turns. It just keeps climbing, or keeps falling, forever, one direction, no U-turns.

A parabola turns exactly once, at its vertex. Climb to the top, or fall to the bottom, then reverse. That's it.

A cubic can turn twice: up, then down, then up again, or the mirror image. That's the S-wiggle.

Notice the pattern? Degree 1 turns 0 times. Degree 2 turns 1 time. Degree 3 turns 2 times. A degree-$n$ curve turns at most $n - 1$ times.

That same limit governs something else too, how many times a curve can cross any given horizontal line.

Drag the checkpoint line up and down. Watch the crossing count for each car's path. It only ever hits Car L's path once, no matter where you put it. It can hit Car Q's path zero, one, or two times. And it can hit Car C's path zero, one, two, or even three times.

A degree-$n$ curve crosses any horizontal line at most $n$ times. That's not a separate fact from the turning-point rule, it's the same fact, just counted in terms of crossings instead of turns.

Spotting the degree from raw numbers alone

Here's a party trick. Give us a table of numbers, evenly spaced inputs and their outputs, no formula attached, and we can often name the degree without ever seeing the equation.

The trick: subtract each number from the one after it. Do it again. And again... and watch for when the row of differences goes flat.

Step through the differencing order and watch each car's numbers. Car L's numbers go flat after just one round of subtracting, first differences constant. Car Q needs two rounds. Car C needs three.

The number of times you have to difference a sequence before it goes flat is exactly its degree. This is how spreadsheets and physics labs quietly detect linear, quadratic, or cubic patterns hiding in raw data, no curve-fitting software required.

Who wins in the long run?

One more question, and it's the one that trips people up the most: which function grows fastest?

Your gut says "whichever has the bigger number out front." That's only true for a while.

Press play. Car L jumps out to an early lead, its coefficient is bigger, so it looks fastest at first. But keep watching... Car C, the cubic, quietly closes the gap and blows past everyone. Car Q catches up to Car L not long after that.

Given enough time, the higher-degree function always wins, no matter how small its coefficient is. A cubic with a tiny leading coefficient will eventually outrun a linear function with a massive one. It just needs enough runway in $x$ to do it.

We can write this more formally. As $x \to \infty$, the term with the highest power of $x$ swallows every other term in the function:

Everything but the leading term becomes irrelevant by comparison.

The short version

Degree is the single number that decides almost everything about a polynomial's shape. It counts how many times you have to ask "how is the rate of change changing" before the answer goes flat. It caps how many times the curve can turn. It caps how many times the curve can cross any horizontal line. It's exactly how many rounds of differencing flatten its raw numbers. And it decides who wins the race as $x$ grows without bound.

Linear: one flat rate of change, no turns, wins early. Quadratic: a steadily changing rate, one turn, catches up. Cubic: an accelerating rate of change, two turns, wins in the end.

Five different questions, same answer every time: count the degree.

All visualizations are interactive React components running entirely in your browser, built with plain SVG and no external libraries. The race simulator uses requestAnimationFrame to animate; everything else responds instantly to drags.