A Rule That Lets Engineers Integrate the Unintegrable
Most real-world functions don't have closed-form antiderivatives. Simpson's rule is how engineers and physicists evaluate them anyway.
The Simpson's rule formula (Simpson's $\frac{1}{3}$ rule) says:
$$\int_a^b f(x),dx \approx \frac{h}{3}\left[f(x_0) + 4f(x_1) + 2f(x_2) + 4f(x_3) + \ldots + 4f(x_{n-1}) + f(x_n)\right].$$
The endpoints get weight $1$, odd-indexed points get weight $4$, even-indexed interior points get weight $2$ — a 1, 4, 2, 4, 2, …, 4, 1 pattern. The number of strips $n$ must be even.
The Formula
For a function $f(x)$ on $[a, b]$ divided into $n$ equal strips (with $n$ even) of width $h = (b-a)/n$, define ordinates $x_k = a + kh$ for $k = 0, 1, \ldots, n$. Then:
$$\boxed{;\int_a^b f(x),dx \approx \frac{h}{3}\left[f(x_0) + 4\sum_{k\text{ odd}} f(x_k) + 2\sum_{k\text{ even},, k\neq 0,n} f(x_k) + f(x_n)\right];}$$
In the simplest case ($n = 2$, just two strips):
$$\int_a^b f(x),dx \approx \frac{h}{3}[f(a) + 4f((a+b)/2) + f(b)] = \frac{b-a}{6}[f(a) + 4f((a+b)/2) + f(b)].$$
Quick facts.
Named after: Thomas Simpson (1710–1761), English mathematician.
Grade introduced: CCSS-M HSF.IF.B advanced — typically AP Calculus AB/BC or Class 12; NCERT Class 12 Chapter 7 — Integrals (numerical-integration extension topic).
Constraint: $n$ must be even (strips come in pairs, since each parabola needs three ordinates).
Error bound: $|E_S| \leq \frac{(b-a)^5}{180 n^4} \max|f^{(4)}(x)|$ — fourth-order accuracy, much better than the trapezoidal rule's second-order.
Exact for polynomials of degree ≤ 3 — even though it uses only quadratic fits, by symmetry it integrates cubics exactly.
How the Formula Is Derived — A Quadratic Fit
Take three equally-spaced points $x_0, x_1, x_2$ with spacing $h$, where $x_1 = (x_0 + x_2)/2$. Fit a unique quadratic $p(x)$ through $(x_0, f(x_0)), (x_1, f(x_1)), (x_2, f(x_2))$.
Integrating $p(x)$ from $x_0$ to $x_2$ (a calculation using Lagrange interpolation) gives:
$$\int_{x_0}^{x_2} p(x),dx = \frac{h}{3}\left[f(x_0) + 4f(x_1) + f(x_2)\right].$$
This is Simpson's rule for one pair of strips. For $n$ strips (with $n$ even), concatenate $n/2$ such formulas. Interior endpoints get counted twice — once as the right end of one pair, once as the left end of the next — so they pick up weight $2$. Odd-indexed interior points keep weight $4$ (each appears in exactly one parabola). Endpoints $x_0$ and $x_n$ keep weight $1$.
The 1-4-2-4-2-…-4-1 weight pattern follows from this concatenation logic.
Three Worked Examples, From Quick to Stretch
Quick. Approximate $\int_0^2 x^2,dx$ using Simpson's rule with $n = 2$.
$h = (2 - 0)/2 = 1$. Ordinates: $x_0 = 0, x_1 = 1, x_2 = 2$. Values: $f(x_0) = 0, f(x_1) = 1, f(x_2) = 4$.
$$\int_0^2 x^2,dx \approx \frac{1}{3}[0 + 4 \cdot 1 + 4] = \frac{8}{3} \approx 2.667.$$
Exact value: $\int_0^2 x^2,dx = \frac{8}{3}$. Final answer: $\frac{8}{3} \approx 2.667$ — Simpson's rule is exact for polynomials of degree $\leq 3$, so it nails this.
Standard (Wrong-Path-First). Approximate $\int_0^1 e^{-x^2},dx$ using Simpson's rule with $n = 4$.
Wrong path. A student in our McKinney TX Grade 12 cohort once wrote: "$n = 4$, so the weights are $1, 4, 4, 4, 1$." Look at the weight pattern again. Odd-indexed points get weight $4$; even-indexed interior points get weight $2$. The correct pattern for $n = 4$ is $1, 4, 2, 4, 1$ — not $1, 4, 4, 4, 1$.
Correct. $h = 1/4 = 0.25$. Ordinates: $x_0 = 0, x_1 = 0.25, x_2 = 0.5, x_3 = 0.75, x_4 = 1$. Function values: $f(0) = 1$, $f(0.25) = e^{-0.0625} \approx 0.9394$, $f(0.5) = e^{-0.25} \approx 0.7788$, $f(0.75) = e^{-0.5625} \approx 0.5698$, $f(1) = e^{-1} \approx 0.3679$.
$$\int_0^1 e^{-x^2},dx \approx \frac{0.25}{3}[1 + 4(0.9394) + 2(0.7788) + 4(0.5698) + 0.3679].$$
Inside the bracket: $1 + 3.7576 + 1.5576 + 2.2792 + 0.3679 = 8.9623$.
$$\approx \frac{0.25}{3} \cdot 8.9623 \approx 0.7469.$$
Final answer: $\int_0^1 e^{-x^2},dx \approx 0.7469$.
True value (to 4 dp, from the error function): $\approx 0.7468$. Simpson's rule with $n = 4$ gets the answer to 4 significant figures — a huge improvement over the trapezoidal rule at the same resolution.
Stretch. A surveyor records the width of a curving river at 7 equally-spaced cross-sections, each $20$ m apart: widths $25, 38, 45, 50, 47, 42, 32$ metres. Estimate the surface area of the river along this $120$-metre stretch using Simpson's rule.
$n = 6$ (six strips), $h = 20$. Apply the rule:
$$\text{Area} \approx \frac{20}{3}[25 + 4(38) + 2(45) + 4(50) + 2(47) + 4(42) + 32].$$
Inside: $25 + 152 + 90 + 200 + 94 + 168 + 32 = 761$.
$$\text{Area} \approx \frac{20}{3} \cdot 761 = \frac{15220}{3} \approx 5073.33 \text{ m}^2.$$
Final answer: $\approx 5073.33$ m².
Where Simpson's Rule Lives — Beyond the Textbook
The rule is everywhere computation meets a curve.
Naval architecture. Ship hull cross-section areas are computed using Simpson's rule from offset measurements — the displacement integral that determines buoyancy and stability is approximated this way in every shipyard.
Pharmacokinetics. Drug concentration $C(t)$ in blood plasma is measured at discrete times. The area under the curve (AUC), a regulatory-required quantity for FDA submissions, is typically computed using Simpson's rule.
Engineering CAD. Software like ANSYS and Abaqus uses Simpson's rule (and its higher-order cousins) for moment-of-inertia integration over irregular cross-sections.
Physics — wave propagation. The total energy carried by a wave packet is the integral of intensity over time. Most experimental setups produce discrete samples; Simpson's rule integrates them.
Statistics. Computing the area under a non-standard probability density function — e.g., a histogram-fitted continuous distribution — often uses Simpson's rule when an analytic integral is unavailable.
The rule is one line. Its real-world reach is everything that integrates a function known only at sample points.
Three Errors That Cost the Most Marks on Simpson's Rule
1. Wrong weight pattern.
Where it slips in: Composite Simpson's rule with $n > 2$.
Don't do this: Apply weights $1, 4, 4, 4, \ldots, 4, 1$.
The correct way: The pattern is $1, 4, 2, 4, 2, \ldots, 4, 1$ — alternating $4$s and $2$s in the interior. Odd-indexed points (counting from $0$) get $4$; even-indexed interior points get $2$. Roughly five out of every ten Grade 12 students in our McKinney TX cohort write the wrong weight pattern on the first composite-Simpson's problem.
2. Forgetting that $n$ must be even.
Where it slips in: Problems where the number of ordinates is provided directly.
Don't do this: Apply the formula with $n = 3$ or $n = 5$.
The correct way: Simpson's rule needs an even number of strips (an odd number of ordinates $\geq 3$). If $n$ is odd, either use $n+1$ strips, switch to Simpson's $\frac{3}{8}$ rule for the last three strips, or use a different method.
3. Confusing Simpson's rule with the trapezoidal rule.
Where it slips in: First-week numerical methods homework.
Don't do this: Use $\frac{h}{2}$ instead of $\frac{h}{3}$ in front, or write weights $1, 2, 2, 2, 1$ (the trapezoidal pattern).
The correct way: Trapezoidal: weights $1, 2, 2, \ldots, 2, 1$, prefactor $\frac{h}{2}$. Simpson's: weights $1, 4, 2, 4, 2, \ldots, 4, 1$, prefactor $\frac{h}{3}$. Different orders of accuracy ($O(h^2)$ vs $O(h^4)$).
4. Treating the error bound as the actual error — the 1986 Challenger lesson on misreading numerical bounds.
Where it slips in: Engineering calculations where the error bound is treated as a hard limit.
Don't do this: Use the error bound $\frac{(b-a)^5}{180 n^4} \max|f^{(4)}|$ as if it were the exact error.
The correct way: The error bound is an upper bound — the actual error is typically far smaller. But — and this is the Challenger lesson — when engineers ignore the worst-case bound and rely on typical behaviour, the worst case can still happen.
The Challenger disaster in 1986 — seven astronauts killed — traced in part to O-ring engineers extrapolating outside the temperature range where data existed; the Rogers Commission found this reasoning failure decisive. The classroom-scale lesson: always check whether the function's fourth derivative is well-behaved over the integration interval before quoting an error figure.
The Mathematician Behind Simpson's Rule
Thomas Simpson (1710–1761, England) was a self-taught weaver-turned-mathematician who never attended university. His father was a weaver in Leicestershire; Simpson followed the trade until a fortune-telling almanac caught his eye, which led him to algebra, which led him to teach mathematics in his spare time.
By 1737 he had published A New Treatise on Fluxions — a textbook on calculus — and by 1745 he was elected a Fellow of the Royal Society and appointed to the Royal Military Academy at Woolwich. His 1743 textbook Mathematical Dissertations on Physical and Analytical Subjects contains the formula that now carries his name, though variants of the rule had been known to Johannes Kepler (1571–1630) — Kepler's barrel rule used the same parabolic-fit logic for measuring wine-barrel volumes a century earlier.
The story Simpson's rule tells is that of a mathematician without formal training contributing to mainstream calculus and getting named recognition — at a time when self-taught mathematicians were the rule, not the exception.
Conclusion
The Simpson's rule formula for an integer-strip approximation is $\frac{h}{3}[f(x_0) + 4f(x_1) + 2f(x_2) + 4f(x_3) + \ldots + f(x_n)]$ with $n$ even.
It fits a quadratic through each pair of consecutive strips — fourth-order accurate.
The 1-4-2-4-2-…-4-1 weight pattern follows from how the per-pair formulas concatenate.
Simpson's rule is exact for polynomials of degree $\leq 3$ — even though only quadratics are fitted.
The most common mistake is writing the wrong weight pattern; the second is forgetting that $n$ must be even.
Five Minutes of Practice
Try these three before moving on. If you slip on the weight pattern, come back to Mistake 1.
Approximate $\int_0^4 \sqrt{x},dx$ using Simpson's rule with $n = 4$.
Approximate $\int_1^3 \frac{1}{x},dx$ using Simpson's rule with $n = 4$, and compare to the exact value $\ln 3 \approx 1.0986$.
The cross-sections of a vase (in cm²) at heights $0, 2, 4, 6, 8, 10, 12$ cm are $20, 28, 32, 34, 32, 28, 22$. Estimate the volume.
Want a live Bhanzu trainer to walk through more Simpson's rule formula problems with your child? Book a free demo class — online globally.
Was this article helpful?
Your feedback helps us write better content