A wrong answer is rarely a mystery. It is almost always one of a short list of familiar mistakes, made at a familiar place, for a familiar reason: a sign dropped when a term moved, a second root forgotten, an inequality divided without flipping, a decimal rounded two steps too early. This guide names those mistakes, shows each one next to its correction on the same example, and then builds the checking toolkit that catches them all: substitution, test values, estimation, inverse operations, and the second method. Every example uses only the equation types the Math Solver handles — linear, quadratic, systems, inequalities, and monic cubics — so you can reproduce each mistake and each check on the tool itself, and watch the solver’s own check lines do exactly what this guide recommends.

Mistake 1: sign errors when moving terms

The most common mistake in all of school algebra is a sign error when a term moves across an equals sign. For 3x + 5 = 20, the correct move is to subtract 5 from both sides: 3x = 20 − 5 = 15, so x = 5. The mistaken version adds 5 instead — 3x = 20 + 5 = 25, so x = 25/3 — or moves the 5 without changing its sign at all. Both errors come from treating “move across and change the sign” as a magic rule rather than as shorthand for an operation performed on both sides. The defence is to write the operation, not just the result: write “−5 both sides” next to the line, and the sign change becomes a consequence you can see, not a rule you must remember. The solver’s steps model this: each step shows the equation in its new form, so a sign error can be located to a single line.

Sign errors concentrate even more densely in quadratics, where b appears twice with different sign behaviour. For x² − 5x + 6 = 0, b = −5, so b² = 25 (the square removes the sign) but −b = 5 (the formula flips it). A student who uses −5 for −b gets x = (−5 ± 1) ÷ 2, giving x = −2 and x = −3 — the negatives of the true roots — and the error is invisible until substitution: (−2)² − 5(−2) + 6 = 4 + 10 + 6 = 20, not 0. That check is the point. Sign errors are not a character flaw; they are a predictable failure mode of working with signed numbers, and the professional response is not “be more careful” but “build a check that catches them”. Substitution catches sign errors every time, because a wrong sign changes the value of the expression, and the equation will not balance.

Mistake 2: the forgotten second root

A quadratic with a positive discriminant has two roots, and answering with only one of them is half an answer. For x² − 5x + 6 = 0, the formula gives x = (5 ± 1) ÷ 2, and the ± is doing real work: the plus branch gives 3, the minus branch gives 2. Students drop the second root in two ways: by computing only the plus branch, or by factoring and then setting only one factor to zero — from (x − 2)(x − 3) = 0, concluding x = 2 and stopping. The defence is a counting habit: before writing the answer, state how many roots you expect. The discriminant tells you: positive means two, zero means one double root, negative means a complex pair. If you expected two and wrote one, the answer is incomplete by your own forecast, and no check is needed to know it.

The same mistake appears in disguise in systems and cubics. A system has two unknowns, and an answer that gives x but not y is not an answer; the solver prints both, with a check in each equation, and your working should do the same. A cubic has up to three roots, and stopping after the first integer root — finding x = 1 for x³ − 6x² + 11x − 6 = 0 and declaring victory — leaves 2 and 3 unfound. The general principle is: the degree of an equation tells you how many roots to expect (counting multiplicity and complex roots), and an answer with fewer entries than that count is unfinished. For a quadratic, always write both branches; for a cubic, always factor out the found root and solve what remains; for a system, always give the point, not a coordinate.

Mistake 3: the inequality that did not flip

Dividing or multiplying an inequality by a negative number without reversing its direction inverts the entire solution set. For −2x + 4 > 10, the correct working is: −2x > 6, then flip on dividing by −2, giving x < −3. The mistaken working keeps the direction, giving x > −3 — which is not merely slightly wrong but exactly backwards: it claims that values like 0 satisfy the inequality, when 4 > 10 is false, and rejects values like −4, when 12 > 10 is true. The mistake is dangerous precisely because the working looks identical to correct linear solving in every other respect; the only difference is one symbol, and it changes every value in the answer.

The defence has two layers. First, flag the flip in writing: put the word “flip” next to the line where you divide by a negative, so the step is visible in your own working just as the solver prints a note when it flips. Second, test both sides of the boundary. A single test value only confirms one side: testing x = −4 confirms that values below −3 work, but it does not confirm that values above −3 fail, and an unflipped answer passes its own one-sided test. Test x = −4 and x = 0, and the unflipped answer fails immediately. This two-sided test is the inequality equivalent of substitution, and it should be as automatic: no inequality is finished until a value on each side of the boundary has been tried in the original inequality.

Mistake 4: solving before rewriting

Applying a method to an equation that has not been put into that method’s standard form produces a wrong answer with correct-looking steps. For 2x² = 8x − 6, the quadratic formula cannot be applied until the equation reads 2x² − 8x + 6 = 0, with a = 2, b = −8, c = 6. A student who reads a = 2, b = 8, c = −6 off the unrewritten form — or worse, a = 2, b = 0, c = −6 — gets roots that fail substitution, and the failure is confusing because every subsequent step was performed correctly. The error was upstream of all of them. The same trap appears in linear equations: 5x = 2x + 9 must become 3x = 9 before the divide step, and in systems, where Cramer’s rule expects both equations in the form ax + by = c. The solver enforces this discipline structurally — its boxes hold coefficients, so an unrewritten equation cannot be entered correctly — but your handwritten working has no such guard, so you must supply it yourself: rewrite first, then solve, every time.

A related rewriting error is misreading which coefficient is which. In ax² + bx + c = 0, a is the coefficient of x², b of x, and c the constant — but only after rewriting. For x² − 5x + 6 = 0, a = 1, not 0 or “nothing”: a missing coefficient is 1, not 0, and a missing term is 0, not 1. For 0x² + 4x − 8 = 0, a = 0, which means the equation is linear, not quadratic, and the solver handles it by solving linearly rather than forcing the formula. Writing a, b, and c explicitly before computing — “a = 1, b = −5, c = 6” — takes five seconds and prevents an entire class of errors, including the discriminant computed from the wrong b, which then forecasts the wrong root count and misleads everything after it.

Mistake 5: rounding too early

Rounding an intermediate value turns a correct method into a wrong answer, and the error grows with every subsequent step. The system’s solution x = 2.2, y = 1.2 is exact in this case, but consider a quadratic whose roots are irrational, such as x² − 5x + 3 = 0, with roots (5 ± √13) ÷ 2. Rounding √13 to 3.61 early gives roots 4.305 and 0.695, and substituting the rounded values gives residuals, not zeros — small, but nonzero, and enough to fail an exact check. Worse, if those rounded roots are then used in further work, the error compounds. The rule is simple: keep fractions and surds exact through every intermediate step, and convert to a decimal only on the final line, to the precision asked for. The solver follows this rule itself, reporting exact fractions such as x = 1/2 where they exist, and its check lines use exact values so that they give exactly 0.

Decimals entered as coefficients deserve the same care in reverse. The solver accepts 0.5 for 1/2 exactly, but 1/3 entered as 0.333 is an approximation, and an answer computed from it inherits that approximation. When a problem gives fractions, the exact approach is to keep them as fractions through your own working and only use the solver’s decimal entry when the fraction converts exactly (halves, quarters, fifths, tenths) or when an approximate answer is acceptable. And when a problem asks for a decimal answer, round only at the end, and say how many places you rounded to — “x = 2.2 (1 decimal place)” is an answer; “x = 2.2” with the rounding hidden is a claim of exactness that is not true.

Mistake 6: dividing by zero, or by an unknown

Division by zero is not a large number or an error to gloss over; it is undefined, and any step that divides by zero invalidates everything after it. The a = 0 quadratic is the classic case: applying the formula x = (−b ± √D) ÷ (2a) with a = 0 divides by zero, which is why the solver refuses and solves the linear equation instead. The system’s determinant is the same story: when det = 0, Cramer’s rule divides by zero, and the correct answer is “no unique solution” — parallel lines or identical lines — not a number. Students who force a division by zero usually get an error from a calculator and then guess; the correct response is to stop and reinterpret: a zero divisor is information, telling you the equation has changed type or the system has no unique point.

Dividing by an unknown is the subtler version of the same mistake. From 3x = 2x, dividing both sides by x gives 3 = 2, which is false — because the division assumed x ≠ 0, while x = 0 is in fact the solution. The correct move is subtraction: 3x − 2x = 0, so x = 0. The general rule is: never divide by an expression containing the unknown unless you have separately considered the case where that expression is zero. In the methods of this guide the risk appears when simplifying: cancelling a common factor of x from both sides of an equation loses the root x = 0. Factor instead of cancelling — x(x − 2) = 0 gives both x = 0 and x = 2 — and the lost root stays found.

Mistake 7: evaluating in the wrong order

Many equation errors are not equation errors at all; they are expression evaluations done in the wrong order, usually during a substitution check. Checking x = 3 in x² − 5x + 6 = 0 means evaluating 9 − 15 + 6: power first (9), multiplication (15), then left to right (0). A reader who works left to right as a chain — 9 − 5 = 4, 4 × 3 = 12, 12 + 6 = 18 — concludes that a correct root has failed, which is worse than no check, because it teaches distrust of right answers. The same failure appears in the discriminant: D = b² − 4ac for a = 1, b = −5, c = 6 is 25 − 24 = 1, where 4ac is a single block evaluated before the subtraction; a reader who subtracts first gets a different, wrong discriminant and then a wrong forecast. Order of operations is the grammar underneath every check in this guide, and fluency in it is a prerequisite, not a separate topic.

The checking toolkit: five checks that catch everything

No single check catches every mistake, but five checks together catch essentially all of them, and they take less time than redoing a problem. First, substitution: put every root back into the original equation and require an exact balance — 0 for a quadratic in standard form, the right side for a linear equation, both equations for a system, zero for each cubic root. Second, test values for inequalities: one value inside the solution set and one outside it, both tried in the original inequality. Third, estimation: before computing, predict the rough size of the answer, and be suspicious of anything wildly different — a root of 25/3 for 3x + 5 = 20 should feel wrong before it is checked, because 3x cannot exceed 20. Fourth, inverse operations: undo your last step and confirm you return to the previous line; if dividing gave x = 5, multiplying back should give 3x = 15. Fifth, the second method: solve the same problem another way — factoring versus the formula, elimination versus substitution — and require agreement. A disagreement between two methods locates the error to the method that differs, which is far more useful than merely knowing an error exists.

Make checking the last line of every solution, not an optional extra done when time permits. Write “Check:” and then the substitution, in full, underneath the answer. The solver’s own output models this: every answer arrives with its check lines attached, and the result tiles report whether the check passed. Students who adopt the same format find that their marks improve without their mathematics getting harder — because most of the marks they were losing were never about hard mathematics. They were about sign errors, dropped roots, and flipped inequalities, made in haste and never caught. A check is not what you do when you are unsure. It is what you do so that you can be sure, and it is the habit that separates a student who gets the right answer from a student who knows they have the right answer.

Frequently asked questions

What is the most common math mistake?

A sign error when moving a term across an equals sign, or a sign error in b when using the quadratic formula. Both are caught by substitution: put the answer back into the original equation and require an exact balance. Writing the operation (“subtract 5 from both sides”) next to each line, rather than just the result, prevents most sign errors before they happen.

How do you check the answer to a quadratic equation?

Substitute each root into the original equation and confirm it gives exactly 0. For x² − 5x + 6 = 0, the roots 3 and 2 give 9 − 15 + 6 = 0 and 4 − 10 + 6 = 0. Both roots must be checked, and the check must use the original equation, not a rewritten version, so that rewriting errors are caught too.

How do you check the answer to an inequality?

Test one value inside the solution set and one outside it, both in the original inequality. For x < 5 from 3x + 5 < 20, x = 0 satisfies it and x = 6 does not. A single test value only confirms one side, so an unflipped answer — the most common inequality mistake — passes a one-sided test and fails a two-sided one.

Why do I keep forgetting the second root of a quadratic?

Because the ± in the formula is easy to read as decoration. Before writing the answer, state how many roots the discriminant predicts: positive means two different real roots, zero means one double root, negative means a complex pair. If you expected two and wrote one, the answer is incomplete by your own forecast.

Should I round my answer during the working?

No. Keep fractions and surds exact through every intermediate step and convert to a decimal only on the final line, to the precision asked for. Rounding early makes substitution checks give small nonzero residuals instead of exact zeros, and the error compounds if the rounded value feeds further steps.

What should I do when my answer fails its check?

Do not start over. Compare your steps line by line against the solver’s steps for the same equation until you find the first line where they differ; the error is there or earlier. The usual suspects are a sign error, a dropped root, a missed flip, an equation solved before rewriting, or an expression evaluated in the wrong order.