Here's a question you already know how to answer with arithmetic: if water flows into a tank at a steady 5 liters per minute for 3 minutes, how much went in? You multiplied — 5 × 3 = 15 liters. Rate times time equals total.
But what if the flow rate changes every instant — fast at first, then trickling, then surging again? You can't just multiply, because there's no single rate to multiply by. This is the exact problem the integral was built to solve: it's what "rate times time" becomes when the rate won't hold still. Slice the time into tiny pieces so small the rate is nearly constant across each one, multiply rate by the tiny time on each slice, and add up all those tiny totals. That sum — taken to the limit — is the integral. So the integral of a rate isn't some abstract symbol; it's the grown-up version of rate × time.
This is why the idea is so powerful: you never need a formula for the quantity itself, only for its rate. You don't track the tank's water level moment by moment — you take the flow rate, integrate it, and out comes the total added. A capacitor charging, a population growing, a car's odometer climbing: in every case, total accumulated = integral of the rate. Everything else on this page is this one idea wearing a different costume.
Now give that rate a specific job. Velocity is the rate your position changes — so by the idea above, position is just the accumulation of velocity. Integrate velocity over some time and you get how much position changed. Simple enough. But motion hides a fork in the road that costs students points on the exam constantly, and it's worth slowing all the way down for.
The fork is this: which "how far" are you asking about? There are two, and they're different numbers. Imagine walking 3 steps forward, then 3 steps back. How far did you travel? Six steps — your feet did real work. But where did you end up? Right where you started — net zero. Both answers are correct; they're answering different questions. Velocity handles them with one small distinction.
They're equal only when the particle never reverses. The moment it turns around, they split — and the visual below lets you watch exactly that happen. Sweep the slider: the two numbers start identical and stay locked while the particle moves forward, then pull apart the instant it reverses.
So far we've integrated over a fixed interval to get one number. But what if you let the finish line move? Fix the start at a, and let the upper limit slide. Now the integral isn't a single value — it's a function of where you stop. That's the accumulation function: a running total that grows (or shrinks) as you sweep to the right.
The classic picture is a bank balance, or water in a tank. The rate is money flowing in and out each day; the accumulation function is your balance over time. On days money flows in (positive rate), the balance climbs. On days more flows out than in (negative rate), the balance drops. The rate drives the total — and that relationship is the whole point.
Watch the two graphs together below. The top is the rate f; the bottom is the accumulation F being drawn as you sweep. Three things to notice — they're the entire concept:
Think about how you average test scores: add them up, divide by how many there are. But a function has infinitely many values across an interval — you can't "add them up and divide by how many." So what could the average of a function even mean? Here's the move that rescues it: instead of averaging points, ask about area.
Picture the area under a wavy curve. Now ask: what single flat height would trap that exact same area over the same interval? If you could press the wave down into a level rectangle without spilling any area, the height of that rectangle is the average value. It's the "fair" constant — the steady level that accounts for the same total as all the ups and downs combined. (Same logic as average speed: the one steady speed that would've covered your trip's distance in the same time.)
See it directly: the area under the curve and the dashed rectangle at the average height enclose exactly the same area. The wave has been pressed flat into a level of equal size — that height is favg.
One guarantee worth knowing: the Mean Value Theorem for Integrals says a continuous curve must actually hit its average height at some point on the interval — the wave genuinely crosses the flat line, it isn't just an abstract average.
Six questions across net change, the displacement-vs-distance trap, accumulation, and average value. Each shows a worked explanation after you answer.