Identities That Do the Work
— the most-used identity in competition algebra.
In plain terms. A difference of two squares always splits into a product, often collapsing huge arithmetic.
Example. — no squaring needed.
By Vieta's formulas, everything symmetric in two roots can be written using and : for instance .
In plain terms. You rarely need the roots themselves — just their sum and product.
Example. If and , then .
From : and .
In plain terms. Expressions built from and collapse into a polynomial in .
Example. If , then .
Why it works. Square : the cross term is , a constant, so . Cubing works the same way: . Because the products keep collapsing to , every expression is a polynomial in . One caution: for real , , since needs .
What is the largest prime factor of ?
Spotting it. An even power minus is a difference of squares, and it keeps splitting.
Step 1. .
Step 2. and , so and the largest prime factor is . Computing first and then factoring it is the slow way to the same place.
If , find and .
Spotting it. Everything is symmetric in and , so use and the identities; never solve for .
Step 1. .
Step 2. , and squaring AGAIN, .
Find all real solutions of .
Spotting it. The coefficients read the same forwards and backwards (). That symmetry means dividing by produces and .
Step 1. is not a root, so divide by : .
Step 2. With : , i.e. , so or .
Step 3. : , so . : , so . All four solutions are real: a quartic became two quadratics.
Dropping the constant: is , not , because squaring produces the cross term . And watch for impossible values: if a problem gives for real , there is no such (the equation has no real roots), which a correct formula will not warn you about.
What is the largest prime factor of ?
Show the answerHide the answer
Split three times: . Since is prime (no prime up to divides it), the largest prime factor is .
Inequalities That Decide Problems
For nonnegative : , with equality exactly when all are equal.
The equality case is the useful half. To minimise for : AM–GM gives , attained when , i.e. . Any optimisation whose answer "looks symmetric" is usually AM–GM in disguise, and the equality condition tells you where the optimum sits without calculus.
.
With it gives the power-mean bound , which turns "the sum is fixed" into "the sum of squares is at least this much" — exactly the step needed for extremal problems on contests. Equality needs the vectors proportional, which again names the optimiser.
For , what is the least value of , and where is it attained?
Step 1 — AM–GM on the two terms: .
Step 2 — equality when , i.e. , .
Step 3 — check: ✓
For , minimise .
Show the answerHide the answer
, attained at .
For , find the least value of .
Spotting it. AM–GM on the two terms gives , which still contains : useless. Split into two equal halves so that the PRODUCT of the three terms is constant.
Step 1. .
Step 2. Equality needs , i.e. , and then ✓. The least value is .
How to choose the split. The powers of in the pieces must cancel: .
Positive reals and satisfy . Find the least value of .
Spotting it. A sum of fractions against a fixed sum of denominators is Cauchy–Schwarz in the form .
Step 1. With , : .
Step 2. Equality needs , so ; with that is , , and indeed ✓. The least value is .
Using an equality case that is not allowed. For , AM–GM says , but equality needs , which is outside the range, so is never reached. The function is increasing for , so the true minimum is at : . Always check that the equality case is actually attainable.
Radicals: Conjugates and Denesting
If and with , then and . To clear from a denominator, multiply by .
Why it works. , so a nested radical of the form denests exactly when two numbers have sum and product , which is a Vieta question. Conjugates work because has no radicals left: a difference of squares again.
Simplify .
Spotting it. First make the inner coefficient a : . Now look for two numbers with sum and product .
Step 1. and .
Step 2. So . Check: ✓
Compute .
Spotting it. Each denominator is a sum of square roots of consecutive integers, and its conjugate has product .
Step 1. .
Step 2. The sum telescopes: .
Simplify .
Spotting it. The inner radical has no factor in front, and leads to fractions with no nice sum-product pair. Multiply inside and outside by instead.
Step 1. .
Step 2. and , so .
Step 3. .
Splitting a root over a sum: . For instance , not . Square roots split over PRODUCTS ( for ), and that is the only reason denesting needs the sum-and-product search at all.
Simplify .
Show the answerHide the answer
Two numbers with sum and product are and , so (the larger one first, since a square root is positive). Check: ✓