The exhibits here are not silly slips — they are mistakes that clever people make too. For each one, find the flaw yourself first, then open the diagnosis. 208 exhibits so far.
Vol. 1 Sets hall
Exhibit
Solving for the parameter and forgetting to check distinctness
✗
The wrong step: 4∈{a+2,a2}, so a+2=4 or a2=4, giving a=2 or a=−2.
What went wrong? Think first, then open
DiagnosisWhen a=2, both elements a+2 and a2 equal 4, and the set reads {4,4}. The elements of a set must be distinct, so this is not a two-element set at all.
✓
The right step: After solving, substitute each candidate back. a=2 makes the two elements equal, so reject it; a=−2 gives the set {0,4}, which works. So a=−2.
“When a set contains a parameter, always substitute back and check that the elements are distinct”
The wrong step: ∅={∅}, because both sides are “empty”.
What went wrong? Think first, then open
Diagnosis∅ has no elements at all; {∅} has one element, which happens to be the empty set. Think of an empty box versus a box containing one empty box.
✓
The right step: ∅∈{∅} and ∅⊆{∅}, but ∅={∅}: one has 0 elements, the other has 1.
“Count the elements first, then ask whether the sets are equal”
DiagnosisTake A=B=C={1}. On the left, B∖C=∅, so the left side is {1}; on the right, A∖B=∅, and removing more still leaves ∅. The two sides disagree on the region A∩B∩C.
✓
The right step: Check element by element, or rewrite the difference as an intersection: A∖(B∖C)=A∩(∁UB∪C)=(A∖B)∪(A∩C).
“Only union and intersection let you move brackets freely; difference does not”
Taking the complement inside the brackets but not flipping the operation
✗
The wrong step: ∁U(A∩B)=∁UA∩∁UB
What went wrong? Think first, then open
DiagnosisAn element outside A∩B is only missing from at least one of them: it may be in A but not in B. For example, with U={1,2}, A={1}, B={2}, the left side is U and the right side is ∅.
✓
The right step: De Morgan's laws do two things at once: complement each set, and swap intersection for union: ∁U(A∩B)=∁UA∪∁UB.
“When the complement goes inside the brackets, intersection and union swap”
Counting through a correspondence that is not one-to-one
✗
The wrong step: Match each subset of {1,2,3,4} with the sum of its elements. The sum can only be 0,1,…,10, so there are 11 subsets.
What went wrong? Think first, then open
DiagnosisDifferent subsets can have the same sum: {1,4} and {2,3} both sum to 5. This correspondence is “many-to-one”, so the two sides do not have the same size (there are actually 16 subsets).
✓
The right step: Before counting through a correspondence, check two things: different objects go to different images (injective), and every image is hit by some object (surjective). Subsets ↔ 0-1 strings passes both checks.
“Only a one-to-one correspondence lets you count the other side instead”
The wrong step: Split {1,2,3,4} into two classes: each element picks a class, 24=16 ways; the classes are unordered, so divide by 2 to get 8.
What went wrong? Think first, then open
DiagnosisTwo of the 16 colorings use a single color for everything, and then the other class is the empty set. Every piece of a partition must be non-empty, so these two do not count.
✓
The right step: (24−2)÷2=7 ways. In general, an n-element set can be split into two pieces in 2n−1−1 ways.
The wrong step: 20 students in the class are in the math club and 15 in the physics club, so 35 are in at least one club.
What went wrong? Think first, then open
DiagnosisSome students may be in both clubs. “Math club” and “physics club” are not a partition: the two pieces intersect, and anyone in both is counted twice.
✓
The right step: The addition principle only works for pairwise disjoint pieces. With overlap you must subtract what was counted twice: ∣A∪B∣=∣A∣+∣B∣−∣A∩B∣. That is the inclusion–exclusion principle of station 9.
“Check that the pieces are pairwise disjoint before you use the addition principle”
Thinking a largest intersecting family must all share one element
✗
The wrong step: For an intersecting family of subsets of {1,2,3,4} to reach 8 members, it must be “all subsets containing some element i”.
What went wrong? Think first, then open
DiagnosisCounterexample: {1,2},{1,3},{2,3} together with all 4 three-element subsets and {1,2,3,4}. That is 8 members, every two of which intersect (a three-element subset misses only one element, so it meets every two-element subset), yet no element is shared by all of them.
✓
The right step: The proof of the bound 2n−1 only says “at most one subset from each complementary pair”. Any family that takes exactly one from each pair and is pairwise intersecting reaches the bound, and there is more than one such family.
“A construction that reaches the bound is often not unique”
The wrong step: Subsets in the same layer never contain one another, so all 1-element subsets of {1,2,3,4,5} form the largest family in which no member contains another.
What went wrong? Think first, then open
DiagnosisSubsets in one layer really do not contain one another, but the layers have different sizes: there are only 5 one-element subsets, but C52=10 two-element subsets.
✓
The right step: The layer sizes Cnk are largest in the middle. Sperner's theorem says the maximum is exactly Cn[n/2], reached by the middle layer.
Saying “equals” when you have only shown “must contain”
✗
The wrong step: S is closed under subtraction, 3∈S and 5∈S, so S=Z.
What went wrong? Think first, then open
DiagnosisWhat follows is 5−3=2, 3−2=1, and from there every integer is in S, that is, S⊇Z. But the problem never said S contains only integers: S=R is also closed under subtraction and also contains 3 and 5.
✓
The right step: The conclusion should be Z⊆S. To claim equality you would also have to prove the other direction: S contains nothing else.
“What you derive is a lower bound, not the whole set”
The wrong step: S satisfies a∈S⇒1−a1∈S. Take S={a}; all we need is a=1−a1.
What went wrong? Think first, then open
Diagnosisa=1−a1 becomes a2−a+1=0, whose discriminant is 1−4<0: no real solution. So in a set S of real numbers, a and 1−a1 are always two different elements.
✓
The right step: First check whether the fixed-point equation has a solution. Here it has none, so S has at least two elements; pushing further, it actually has at least three.
“Before you make the rule “stand still”, solve the equation”
The wrong step: Solve ∣x∣+∣x−2∣=2. When x<0: …; when x>2: …; when 0<x<2: …
What went wrong? Think first, then open
Diagnosisx=0 and x=2 belong to no case: something is left out. And this equation holds at every point of 0≤x≤2, including both endpoints, so leaving them out gives a wrong answer.
✓
The right step: The cases must cover everything: x<0, 0≤x≤2, x>2. A boundary point may go on either side, but it must go on one side, and only one.
“Nothing left out (the union is everything); nothing counted twice (the classes are pairwise disjoint)”
The wrong step: Let S be the set of all positive real numbers, and let m be the smallest element of S…
What went wrong? Think first, then open
DiagnosisS has no smallest element: for any m>0, 2m is smaller. The well-ordering principle only promises a smallest element for a finite non-empty set of reals, or a non-empty set of natural numbers.
✓
The right step: Before using the extremal principle, check that there are finitely many objects, or that the values are non-negative integers. Otherwise, first find a way to make the problem finite.
“Make sure the extreme object exists before you pick it”
The wrong step: Let A have the most wins. If B beat A, then B has more wins than A.
What went wrong? Think first, then open
DiagnosisB beating A only means B won one game that A did not; A may also have won many games that B did not. And several players may be tied for the most wins.
✓
The right step: The extremal principle gives you just one sentence: “nobody is larger”. Your argument has to produce a concrete reason why someone has strictly more, for example B beat Aand everyone A beat.
“Use “nobody is larger”; do not swap it for “it is larger than everyone else””
Stopping after subtracting the pairwise intersections
✗
The wrong step: ∣A∪B∪C∣=∣A∣+∣B∣+∣C∣−∣A∩B∣−∣A∩C∣−∣B∩C∣
What went wrong? Think first, then open
DiagnosisLook at the middle region A∩B∩C: the first three terms count it 3 times, the last three subtract it 3 times, leaving 0 times. It has been left out.
✓
The right step: Add ∣A∩B∩C∣ back. On the workbench, switch off the last term and the middle region immediately turns amber and is marked “0×”.
“If you subtract too much, add it back, until every region is counted exactly once”
The wrong step: ∣A∣=2 and ∣B∣=3, so there are 23=8 mappings from A to B.
What went wrong? Think first, then open
DiagnosisThe elements of A are the ones making choices: each picks an image from the 3 elements of B, and 2 choices are made. That is 3×3, not 2×2×2.
✓
The right step: The number of mappings A→B is ∣B∣∣A∣=32=9. To remember it: the exponent is “how many choices are made” (the size of A), the base is “how many options each choice has” (the size of B).
Checking “is there an arrow?” but not “how many arrows?”
✗
The wrong step: A={1,2}, B={3,4}, with 1↦3, 1↦4, 2↦4. Every element of A has an image, so this is a mapping.
What went wrong? Think first, then open
DiagnosisA mapping needs every element of A to have a unique image. Here 1 goes to both 3 and 4, so its image is not unique.
✓
The right step: Check both: every element of A has at least one arrow (none left out) and at most one arrow (none doubled). How often an element of B is hit, or whether it is hit at all, has nothing to do with being a mapping.
“On the A side, none missing and none doubled; on the B side, anything goes”
The wrong step: The graph of y=(x+1)2 is the graph of y=x2 shifted 1 unit to the right.
What went wrong? Think first, then open
DiagnosisThe point of the new graph at x has the same height as the point of the old graph at x+1. So what the old graph looked like at x+1 now appears at x: everything has moved 1 to the left.
✓
The right step: y=f(x−h) is y=f(x) shifted right by h (left if h<0). (x+1)2=(x−(−1))2, so h=−1: shift left by 1.
The wrong step: f(x)=x2−2x−3; the graph of y=f(∣x∣) is the graph of y=f(x) with the part below the x-axis folded up.
What went wrong? Think first, then open
DiagnosisThat is y=∣f(x)∣: the absolute value sits on the function value, so it acts up and down. In f(∣x∣) the absolute value sits on the variable: x and −x give the same value, so it is an even function. Keep the part to the right of the y-axis and replace the left part by the mirror image of the right.
✓
The right step: ∣f(x)∣: fold the bottom up, nothing moves sideways. f(∣x∣): copy the right onto the left, nothing moves up or down. Toggle between the two buttons on the workbench to compare.
The wrong step: f(x)=x−1x2(x−1)=x2, so f(x) is even.
What went wrong? Think first, then open
DiagnosisSimplifying threw away a restriction on the domain: the original function needs x=1. So x=−1 is in the domain but 1 is not. The domain is not symmetric about the origin, so the function can be neither odd nor even.
✓
The right step: The first step in deciding parity is always to check whether the domain is symmetric about the origin. Here it is not, so f(x) is neither odd nor even.
Decreasing on two pieces, so decreasing on their union
✗
The wrong step: f(x)=x1 is decreasing on (−∞,0)∪(0,+∞).
What went wrong? Think first, then open
DiagnosisTake x1=−1<x2=1: f(−1)=−1<f(1)=1, so the right-hand point is higher. Monotonicity must hold for any two points of the interval, and a pair straddling 0 breaks it.
✓
The right step: Write “f(x) is decreasing on (−∞,0) and on (0,+∞)”, separating the intervals with “and” or a comma. Do not join them with ∪.
“Do not glue monotone intervals together carelessly”
The wrong step: f(1)<f(2)<f(3), so f(x) is increasing on [1,3].
What went wrong? Think first, then open
DiagnosisThree points say nothing about what happens in between: f(x)=(x−1.2)2 has f(1)=0.04<f(2)=0.64<f(3)=3.24, yet it is decreasing on [1,1.2].
✓
The right step: A proof must work for anyx1<x2: form the difference f(x2)−f(x1), turn it into a product, and decide the sign of each factor. Special values can only disprove monotonicity.
The wrong step: f(x+2)=−f(x), so f(x) is periodic with period 2.
What went wrong? Think first, then open
DiagnosisAfter a shift by 2 the graph does not land on itself: it lands upside down. A period needs f(x+T)=f(x), and here there is an extra minus sign.
✓
The right step: Shift once more: f(x+4)=−f(x+2)=−(−f(x))=f(x). The period is 4 (and 2 is not a period unless f is identically 0).
Every periodic function has a least positive period
✗
The wrong step: Every periodic function has a least positive period; just work it out.
What went wrong? Think first, then open
DiagnosisThe constant function f(x)=c has every positive number as a period, so there is no smallest one. The Dirichlet function (1 when x is rational, 0 when x is irrational) has every positive rational as a period, and also has no least positive period.
✓
The right step: Before you talk about “the least positive period”, make sure it exists. When we say “sinx has period 2π”, we mean its least positive period.
“Having a period does not mean having a least positive period”
Writing only $f(1)>0$ for “both roots greater than 1”
✗
The wrong step: Both roots of x2+4x+3=0 are greater than 1, because f(1)=8>0.
What went wrong? Think first, then open
Diagnosisf(1)>0 only says that 1 does not lie between the roots: both roots could be to the right of 1, or both to the left. Here the roots are −1 and −3, both on the left.
✓
The right step: You also need the axis: both roots greater than 1⟺Δ≥0, −2b>1, f(1)>0. Here −2b=−2<1, so it fails.
“The sign at an end point only tells “same side” or “opposite sides”; which side comes from the axis”
Checking the axis and the end points, forgetting the discriminant
✗
The wrong step: x2−3x+3: the axis x=23 lies in (0,3), f(0)=3>0 and f(3)=3>0, so both roots lie in (0,3).
What went wrong? Think first, then open
DiagnosisΔ=9−12<0: there are no real roots at all. The whole parabola floats above the x-axis; the end-point and axis conditions hold, but there are no roots to place.
✓
The right step: “Both roots in (m,n)” needs all four conditions: Δ≥0, m<−2b<n, f(m)>0, f(n)>0.
“First ask whether there are roots, then ask where they are”
Finding where a composite log function increases, forgetting the domain
✗
The wrong step: y=log21(x2−2x−3): the inner x2−2x−3 is decreasing on (−∞,1) and the outer function is decreasing, so y is increasing on (−∞,1).
What went wrong? Think first, then open
DiagnosisThe argument of a logarithm must be positive: x2−2x−3>0, that is, x<−1 or x>3. Points of [−1,1) are not in the domain at all; at x=0, for instance, the argument is −3.
✓
The right step: Find the domain first, then study the inner function's monotonicity inside it: y is increasing on (−∞,−1) and decreasing on (3,+∞).
“With a logarithm, settle the argument first, then talk about monotonicity”
The discriminant method, forgetting the leading coefficient can be 0
✗
The wrong step: y=x2+x+12x2+2x+5: from Δ≥0 for (2−y)x2+(2−y)x+(5−y)=0 we get 2≤y≤6, so the minimum is 2.
What went wrong? Think first, then open
DiagnosisWhen y=2 the equation becomes 0⋅x2+0⋅x+3=0, which has no solution at all. Δ only means something for a quadratic equation; here Δ=0 tells you nothing.
✓
The right step: Check y=2 separately by substituting it back: it is not reached. So the range is (2,6]: there is no minimum, and the maximum is 6 (reached at x=−21).
“Last step of the discriminant method: test separately the value that makes the leading coefficient 0”
Checking only the endpoints for a quadratic as well
✗
The wrong step: h(t)=4t2−4t+21, and h(0)=h(1)=21>0, so h(t)>0 on [0,1].
What went wrong? Think first, then open
Diagnosis“The two ends are enough” is true only for a linear function, whose segment lies between its two ends. h is a parabola opening upward, and it dips in the middle: h(21)=−21<0.
✓
The right step: First make sure the function you built is linear (or monotonic) in the chosen variable. For a quadratic, look at the vertex and the discriminant.
“The endpoint method needs “linear” or “monotonic””
Concluding $u=v$ from $f(u)=f(v)$ when $f$ is not monotonic
✗
The wrong step: f(t)=t3−t is odd and f(2x+1)+f(x)=0, so f(2x+1)=f(−x), 2x+1=−x, and the only solution is x=−31.
What went wrong? Think first, then open
Diagnosisf(t)=t3−t is not monotonic on R: f(0)=f(1)=f(−1)=0, so different inputs can give the same value. For example x=0 gives f(1)+f(0)=0, so it is a solution too.
✓
The right step: To get u=v from f(u)=f(v) you must first prove that f is strictly monotonic (t3+t is, t3−t is not). Otherwise, expand and solve the equation the honest way.
“Oddness makes the two sides match; monotonicity makes the inputs equal”
Solving the fixed-point equation without proving the limit exists
✗
The wrong step: a1=1, an+1=2an+1. Let the limit be L; then L=2L+1, so an→−1.
What went wrong? Think first, then open
Diagnosisan is 1,3,7,15,…, which grows without bound and has no limit at all. Solving L=f(L) only says: if there is a limit, it must be a fixed point.
✓
The right step: First prove that the limit exists (for example, monotonic and bounded, or, as on this page, the distance to the fixed point shrinks by a fixed ratio), then solve L=f(L). Here f(x)+1=2(x+1): the distance doubles at every step, and the orbit moves away from −1.
“A fixed point is only a candidate; whether the orbit goes there needs its own proof”
The wrong step: f(x2)=x+1 for every real x. Let t=x2, so x=t and f(t)=t+1.
What went wrong? Think first, then open
Diagnosisx can be positive or negative, but t=x2 can only be non-negative, and both x=±t must satisfy the equation. x=2 gives f(4)=3, while x=−2 gives f(4)=−1: a contradiction.
✓
The right step: This equation has no solution: no such function exists. When you substitute, state the range of the new variable clearly, and check with every value of x that can occur.
“The substitution must be reversible, and the range must be stated”
Stopping once you have a formula, without substituting back
✗
The wrong step: Elimination in the system gives a formula for f(x), so that is the function we want.
What went wrong? Think first, then open
DiagnosisElimination only proves: “iff satisfies the original equation, thenf can only be this formula.” That is a necessary condition. Sometimes the formula is undefined at some points, or does not satisfy the original equation at all.
✓
The right step: Substitute the f you found back into the original equation (the “Substituted back” meter on the workbench does exactly that), and state the domain, for example x=0,1.
“Elimination gives a candidate; substituting back makes it an answer”
The wrong step: f(x+y)+f(x−y)=2f(x)+2f(y). x=y=0 gives f(0)=0, x=0 shows f is even, so f(x)=x2.
What went wrong? Think first, then open
Diagnosisf(0)=0 and “even” are only necessary conditions. Plenty of functions satisfy them (for example f(x)=x4), and x4 does not satisfy the original equation; on the other hand f(x)=3x2 does satisfy it too. Plugging in values by itself cannot give “f(x)=x2”.
✓
The right step: Plugging in gives properties and special values. Either keep going until you derive a formula for f (with suitable extra conditions), or guess a solution, substitute it back into the original equation to check, and explain why there are no others.
Proving it for the rationals, then claiming it for all reals
✗
The wrong step: From f(x+y)=f(x)+f(y) we get f(r)=rf(1) for every rational r, so f(x)=cx for every real x.
What went wrong? Think first, then open
DiagnosisThe equation only “strings together” the rational points. An irrational number like 2 can never be reached by finitely many additions, subtractions and integer multiples. Without an extra condition, there really are solutions that are not cx (it is just impossible to write one down).
✓
The right step: The last rung must use the condition given in the problem (monotonic, continuous or bounded) to squeeze each irrational number between rationals (see the last two steps of the proof at the start of this page).
Still writing $c_1x_1^n+c_2x_2^n$ for a double root
✗
The wrong step: f(n+2)=4f(n+1)−4f(n) has characteristic roots x1=x2=2, so f(n)=c1⋅2n+c2⋅2n.
What went wrong? Think first, then open
Diagnosisc1⋅2n+c2⋅2n=(c1+c2)2n really has only one unknown constant, which cannot match two initial values f(1) and f(2). For example, with f(1)=2 and f(2)=12, 2c=2 contradicts 4c=12.
✓
The right step: For a double root the general term is f(n)=(c1+c2n)x1n. The example above gives f(n)=(2n−1)⋅2n; check it term by term: 2,12,40,112,…
“Two initial values need two independent constants”
Taking a square root without checking the quadrant
✗
The wrong step: sinα=135, so cosα=1−16925=1312.
What went wrong? Think first, then open
Diagnosissin2α+cos2α=1 only tells you that cos2α=169144, and that has two square roots, one positive and one negative. An angle with sinα>0 can be in quadrant I or in quadrant II: the line y=135 meets the unit circle at two points.
✓
The right step: If no quadrant is given, split into cases: in quadrant I, cosα=1312; in quadrant II, cosα=−1312. If the quadrant is given, take the sign from the direction of the cosine line OM.
“The Pythagorean identity gives the size; ask the unit circle for the sign”
DiagnosisAt 90∘ the terminal side OP points straight up and is parallel to the tangent x=1, so it never meets it: tan90∘ is undefined. It is not “the number infinity”. And if you approach from the 91∘ side, the tangent line points down and tan heads off in the negative direction.
✓
The right step: Write down the domain: tanx needs x=2π+kπ(k∈Z). Only compare values inside one interval where the function is monotonic.
The wrong step: The graph of y=sin3x is the graph of y=sinx with every x-coordinate stretched to 3 times its value.
What went wrong? Think first, then open
Diagnosisy=sin3x already reaches its maximum at x=6π, while y=sinx only gets there at 2π: the same height comes earlier, so the graph has been squeezed. The period 32π is shorter than 2π.
✓
The right step: Every x-coordinate is shrunk to 31 of its value (y-coordinates unchanged). Think of ω as “how far it turns per unit of x”: the faster it turns, the denser the waves.
The wrong step: y=sin∣x∣ is periodic with least positive period 2π.
What went wrong? Think first, then open
DiagnosisFor x⩾0 it is just sinx, but for x<0 the graph is the mirror image of the right half. Take x=−2π: f(−2π)=1, while f(−2π+2π)=sin23π=−1. One counterexample is enough.
✓
The right step: y=sin∣x∣ has just one axis of symmetry, x=0, and it is not periodic. To show that “T is not a period”, it is enough to find one x with f(x+T)=f(x).
“A period must work for every x; one x is enough to break it”
Copying the increasing-interval recipe when ω is negative
✗
The wrong step: The increasing intervals of y=sin(4π−2x) come from solving −2π+2kπ⩽4π−2x⩽2π+2kπ.
What went wrong? Think first, then open
DiagnosisThe inner function u=4π−2x is decreasing in x. While sinu increases, y actually decreases as x grows, so this gives exactly the decreasing intervals.
✓
The right step: First rewrite it as y=−sin(2x−4π). The increasing intervals of y are the decreasing intervals of sin(2x−4π): 2π+2kπ⩽2x−4π⩽23π+2kπ.
“Make ω positive first, then substitute the whole block”
The wrong step: sin(α+β)=sinα+sinβ, so sin75∘=sin45∘+sin30∘=22+1.
What went wrong? Think first, then open
DiagnosisThe right-hand side is about 1.207, which is bigger than 1: it cannot be a sine. The picture shows it clearly: the two pieces that make up the height of P are not sinα and sinβ; each one is multiplied by a cosine, giving cosβsinα and sinβcosα.
✓
The right step: sin75∘=sin45∘cos30∘+cos45∘sin30∘=46+2≈0.966.
“sin is not multiplication, so you cannot just open the brackets”
Taking only the positive root in a half-angle formula
✗
The wrong step: cosα=21 and 270∘<α<360∘, so cos2α=21+cosα=23.
What went wrong? Think first, then open
DiagnosisIf α is in (270∘,360∘), then 2α is in (135∘,180∘), an angle in quadrant II, whose cosine is negative. The formula only gives the size; the sign depends on where 2α lands.
✓
The right step: cos2α=−21+cosα=−23. (In fact α=300∘ and 2α=150∘.)
“Before taking the root in a half-angle formula, work out the range of the half angle”
DiagnosisCosine is decreasing on (0∘,90∘), so cos40∘>cos80∘ and the difference must be positive. The formula is cosα−cosβ=−2sin2α+βsin2α−β, with a minus sign in front. In the picture: u−v is perpendicular to the angle bisector, and turning the bisector direction by 90∘ turns its cos into −sin.
✓
The right step: cos40∘−cos80∘=−2sin60∘sin(−20∘)=2sin60∘sin20∘=3sin20∘>0. When you are done, use monotonicity to sanity-check the sign.
“Of the four formulas, only “cos minus cos” carries a minus sign”
Turning a sum into a product when the functions differ
✗
The wrong step: sinα+cosβ=2sin2α+βcos2α−β.
What went wrong? Think first, then open
DiagnosisSum-to-product needs the same function in both terms: both vectors must be projected onto the same axis. sinα is a projection onto the y-direction and cosβ onto the x-direction, so they do not combine into one diagonal.
✓
The right step: Make the names match first: cosβ=sin(90∘−β), so sinα+cosβ=2sin2α−β+90∘cos2α+β−90∘.
Finding the auxiliary angle from tan φ = b/a alone
✗
The wrong step: −sinx+3cosx: tanφ=−13=−3, take φ=−3π, so the expression is 2sin(x−3π).
What went wrong? Think first, then open
DiagnosisExpand: 2sin(x−3π)=sinx−3cosx, off by exactly a minus sign. tanφ only fixes the direction of a line, so φ could still be off by π. The auxiliary angle must satisfy two conditions, cosφ=Ra and sinφ=Rb: the point (a,b)=(−1,3) is in quadrant II.
✓
The right step: cosφ=−21 and sinφ=23, so φ=32π and the expression is 2sin(x+32π). Try (a,b)=(−1,1) on the workbench: the φ reading is 43π, not −4π.
“φ is the direction of the point (a, b), not just a slope”
The wrong step: For x∈[0,2π], the minimum of y=sinx+3cosx=2sin(x+3π) is −2.
What went wrong? Think first, then open
Diagnosis−2 would need x+3π=23π, that is x=67π, which is not in the interval. The L may only turn through 90∘, so the hypotenuse never gets to point straight down.
✓
The right step: x+3π∈[3π,65π], where the smallest value of sine is sin65π=21. So the minimum is 1 (at x=2π) and the maximum is 2 (at x=6π).
Finding an angle from its sine and keeping only the acute one
✗
The wrong step: In △ABC, A=30∘, a=2, b=23. By the law of sines sinB=23, so B=60∘.
What went wrong? Think first, then open
DiagnosisIn (0∘,180∘) there are two angles whose sine is 23: 60∘ and 120∘. The law of sines only tells you sinB, not whether B is acute or obtuse. Here b>a, so B>A, and neither angle clashes with A=30∘.
✓
The right step: B=60∘ or B=120∘, and both triangles exist: the first has C=90∘, the second C=30∘. To get everything in one go, use the law of cosines instead and write a quadratic in c: its two positive roots are exactly the two triangles.
“When you get an angle from its sine, ask: does the obtuse one work too?”
The wrong step: In △ABC, sin2A=sin2B, so 2A=2B and the triangle is isosceles.
What went wrong? Think first, then open
DiagnosisTwo angles in (0,2π) with the same sine are either equal or supplementary. Both 2A and 2B lie in (0,2π), and 2A+2B=π is perfectly possible, for example A=30∘, B=60∘.
✓
The right step: sin2A−sin2B=2cos(A+B)sin(A−B)=0, so A=B or A+B=2π: the triangle is isosceles or right-angled.
“Equal sines? Think “equal or supplementary” first”
The wrong step: tan(arctan2+arctan3)=1−2×32+3=−1, so arctan2+arctan3=−4π.
What went wrong? Think first, then open
DiagnosisAngles with the same tangent differ by an integer multiple of π, so the value of the tangent alone does not pin down the angle. Both arctan2 and arctan3 lie in (4π,2π), so their sum lies in (2π,π): it cannot possibly be negative.
✓
The right step: Estimate the range first: the sum lies in (2π,π), and the only angle in that interval with tangent −1 is 43π. So arctan2+arctan3=43π.
“Estimate the range of the angle first, then use the trig value to pin it down”
The wrong step: Solve 2sinxcosx=sinx: divide both sides by sinx to get cosx=21, so x=2kπ±3π.
What went wrong? Think first, then open
DiagnosisDividing by sinx silently assumes sinx=0. But when sinx=0 both sides are 0, so x=kπ are solutions too, and they have been quietly thrown away.
✓
The right step: Bring everything to one side and factor: sinx(2cosx−1)=0, so sinx=0 or cosx=21. The solutions are x=kπ or x=2kπ±3π (k∈Z).
“A factor containing the unknown may be factored out, never cancelled”
Solving sin x > ½ as if it were a linear inequality
✗
The wrong step: sinx>21 and sin6π=21, so x>6π.
What went wrong? Think first, then open
DiagnosisSine is not monotonic. x=π satisfies x>6π, yet sinπ=0; x=−611π does not satisfy x>6π, yet sinx=21… Nothing matches. The solution set comes in separate pieces, and it repeats every 2π.
✓
The right step: Look at the graph over one period: the arc above the horizontal line y=21 is (6π,65π). Then add the period: 2kπ+6π<x<2kπ+65π (k∈Z).
Forgetting to restrict the angle after substituting
✗
The wrong step: Range of y=x+1−x2: let x=cost; then 1−x2=sint and y=2sin(t+4π)∈[−2,2].
What went wrong? Think first, then open
Diagnosis1−x2 is non-negative, but sint can be negative. Only with t∈[0,π] is 1−x2=sint true; without that restriction the equation is wrong whenever t is on the lower half of the circle, and the minimum −2 is never actually reached.
✓
The right step: Let x=cost with t∈[0,π] (then x takes every value in [−1,1] exactly). So t+4π∈[4π,45π] and y∈[−1,2].
“A substitution must match both ways: x covers its whole domain, and t ranges over exactly enough”
Is “real part 0” enough for a purely imaginary number?
✗
The wrong step: z=(m2−1)+(m−1)i is purely imaginary, so m2−1=0 and m=±1.
What went wrong? Think first, then open
DiagnosisFor m=1 we get z=0+0i=0, which is a real number, not a purely imaginary one. A purely imaginary number needs real part 0 and imaginary part not 0; the second condition was dropped.
✓
The right step: Solve {m2−1=0m−1=0 to get m=−1, and then z=−2i.
“Purely imaginary = real part zero + imaginary part nonzero: you need both conditions”
The wrong step: −2(cos30∘+isin30∘) is in trigonometric form, with modulus −2 and argument 30∘.
What went wrong? Think first, then open
DiagnosisTrigonometric form requires r>0. The modulus is a length and can never be negative. The minus sign in front amounts to a further turn of 180∘.
✓
The right step: −2(cos30∘+isin30∘)=2(cos210∘+isin210∘): modulus 2, principal argument 210∘.
“Trigonometric form has three requirements: r positive, cos first, a plus sign in the middle”
DiagnosisThere is a minus sign in the middle, so this is not trigonometric form. The point (cos40∘,−sin40∘) is in the fourth quadrant; its argument is −40∘ plus a full turn.
✓
The right step: cos40∘−isin40∘=cos(−40∘)+isin(−40∘)=cos320∘+isin320∘, so argz=320∘.
“Rewrite in standard trigonometric form first, then read off the argument”
The wrong step: ∣z−1∣+∣z+1∣=2 is the ellipse with foci ±1.
What went wrong? Think first, then open
DiagnosisThe distance between the two fixed points is 2, and the sum of the distances is also 2. Two sides of a triangle add up to more than the third side, with equality only when P lies on the segment, so the locus is just a segment.
✓
The right step: Compare 2k with ∣Z1Z2∣: if it is larger, an ellipse; if equal, the segment Z1Z2 (here the part −1≤x≤1 of the real axis); if smaller, the empty set.
“Compare the constant with the distance between the two fixed points before you conclude anything”
Using the discriminant to test for real roots when the coefficients are not real
✗
The wrong step: The equation x2−(2+i)x+2i=0 has discriminant Δ=3−4i, which is not a non-negative real number, so it has no real root.
What went wrong? Think first, then open
Diagnosis“If Δ≥0 there are real roots” was derived for equations with real coefficients. When the coefficients contain i, the discriminant cannot even be “greater than zero”. In fact, substitute x=2: 4−4−2i+2i=0, so there is a real root.
✓
The right step: Substitute a real root x and separate real and imaginary parts: (x2−2x)+(2−x)i=0, so x2−2x=0 and x=2. The real root is 2 (the other root is i).
“The discriminant is only for real coefficients; with complex coefficients, turn the complex into the real first”
DiagnosisFor two vectors with the same starting point, the difference points from the head of the vector being subtracted to the head of the vector it is subtracted from: AB−AC goes from C to B. Check: AC+CB=AB.
✓
The right step: AB−AC=CB. Rule of thumb: same tail, join the heads, point at the first vector.
DiagnosisAdding vectors combines directions. You get 7 only when a,b point the same way; 1 when they point opposite ways, and 5 when they are perpendicular. In general there is only the triangle inequality: ∣a∣−∣b∣≤∣a+b∣≤∣a∣+∣b∣.
✓
The right step: There is not enough information: all we know is that ∣a+b∣ lies in [1,7]. We also need the angle between them.
Diagnosisa⋅b only sees the shadow of b in the direction of a, and infinitely many vectors cast the same shadow: with a=(1,0), both b=(2,1) and c=(2,−5) have dot product 2.
✓
The right step: The correct conclusion is a⋅(b−c)=0, that is, a⊥(b−c) (or b=c).
“A dot product is a number: the direction information is lost, so you cannot cancel”
The wrong step: a⋅b<0, so the angle between a and b is obtuse.
What went wrong? Think first, then open
Diagnosisa⋅b<0 only says cosθ<0, that is, 90∘<θ≤180∘. When θ=180∘ (the vectors point in opposite directions) the dot product is negative too, but that is not an obtuse angle.
✓
The right step: “The angle is obtuse” ⟺a⋅b<0 and a,b do not point in opposite directions. When you find a range for a parameter, remove the value that makes them opposite.
“Once you know the sign of the dot product, check the collinear case as well”
The wrong step: u⋅v=−4 and ∣u∣∣v∣=8, so the angle between the skew lines is 120∘.
What went wrong? Think first, then open
DiagnosisThe angle between two lines lies in (0∘,90∘]. The angle between two vectors can be obtuse, but a line has no direction, so an obtuse angle has to be replaced by its supplement.
✓
The right step: Use cosθ=∣u∣∣v∣∣u⋅v∣=21: the angle between the lines is 60∘.
“For the angle between lines, put an absolute value in the cosine formula”
Treating the modulus of a sum as the sum of the moduli
✗
The wrong step: ∣z∣=2, so ∣z2+1∣=∣z∣2+1=5.
What went wrong? Think first, then open
Diagnosis∣z2+1∣ is the distance from the point z2 to the point −1. ∣z2∣=4 is fixed, so z2 runs around the circle of radius 4, and this distance ranges from 3 to 5.
✓
The right step: The triangle inequality gives ∣z∣2−1≤∣z2+1∣≤∣z∣2+1, that is, 3≤∣z2+1∣≤5; the value is 5 at z=±2 and 3 at z=±2i.
“The triangle inequality gives a range, not an equation”
The wrong step: Construct the equilateral triangle ABT outward on AB, so t=a+(b−a)(cos60∘+isin60∘).
What went wrong? Think first, then open
DiagnosisMultiplying by cos60∘+isin60∘ turns counterclockwise by 60∘. Whether T lands to the left or the right of AB depends on whether the vertices of the triangle go counterclockwise or clockwise; get it wrong and you construct the equilateral triangle pointing inward.
✓
The right step: First agree that A,B,C go counterclockwise. Then the triangle lies to the left of AB, and outward means turning clockwise: t=a+(b−a)(cos60∘−isin60∘).
“Before you express a rotation as a product, fix the orientation of the vertices”
Calling the points concyclic because a ratio is real
✗
The wrong step: z4−z1z3−z1/z4−z2z3−z2 is real, so the four points Z1,Z2,Z3,Z4 are concyclic.
What went wrong? Think first, then open
DiagnosisWhen the four points lie on one line, both ratios are real and so is their quotient, yet the four points are not concyclic.
✓
The right step: The conclusion should be “the four points are concyclic or collinear”. When a problem asks you to prove they are concyclic, also show that three of them are not collinear.
“A line is a “circle of infinite radius”; do not forget to rule it out”
The wrong step: Claim: n2+n+1 is even. Proof: suppose k2+k+1 is even. Then (k+1)2+(k+1)+1=(k2+k+1)+2(k+1) is even too. By induction, the claim holds for every n.
What went wrong? Think first, then open
DiagnosisThe inductive step really is fine, but for n=1 we get 1+1+1=3, which is odd. The first domino never falls, so it does not matter that each one “would knock over the next”. In fact n2+n=n(n+1) is always even, so n2+n+1 is always odd: the claim is false for every n.
✓
The right step: Induction needs both parts: first check P(1), then prove P(k)⇒P(k+1). Choose “n2+n+1 is even” in the domino chain above: every gap shows ✓, yet not a single domino falls.
“Without the first push, even the strongest chain stays still”
The wrong step: “Any n horses all have the same color.” For n=1 it is obvious. Suppose any k horses have the same color. Take k+1 horses. Remove the first: the remaining k have the same color. Remove the last: those k have the same color too. The two groups share a horse, so all k+1 have the same color.
What went wrong? Think first, then open
Diagnosis“The two groups share a horse” needs k+1≥3. Going from k=1 to k=2, removing the first horse leaves {2} and removing the last leaves {1}: the two groups have no element in common, so the argument breaks at the step 1→2. One broken gap is enough to stop everything after it.
✓
The right step: When you write the inductive step, check that the argument also works for the smallestk. If it only works for k≥k0, you must check P(1),…,P(k0) separately (and here P(2) is exactly the one that is false).
“Break the chain in one place and nothing after it falls”
The step uses two earlier terms, but only one base case is checked
✗
The wrong step: The sequence a1=3,a2=6,an=3an−1−2an−2. “Proof” that an=2n+1: for n=1, a1=3=2+1. Suppose ak−1=2k−1+1 and ak=2k+1. Then ak+1=3(2k+1)−2(2k−1+1)=2k+1+1.
What went wrong? Think first, then open
DiagnosisThe inductive step uses two earlier terms. The first time it can be used is for k+1=3, and then it needs P(1)andP(2). But P(2) was never checked. In fact a2=6=5, then a3=18−6=12=9, and the conclusion is wrong all the way.
✓
The right step: If the step uses several earlier terms, check that many base cases: here you must check both P(1) and P(2). As soon as you do, you find that P(2) is false.
The wrong step: Proof that x=2y has no solution in positive rationals: if (x,y) is a solution, then so is (2x,2y), and it is smaller. This can go on getting smaller forever, a contradiction.
What went wrong? Think first, then open
Diagnosisx=2,y=1 is obviously a solution. The mistake: “getting smaller forever” is no contradiction at all among the positive rationals, since 1,21,41,… is an infinite decreasing sequence of positive rationals. The well-ordering principle only holds for the positive integers (or any set of integers that is bounded below).
✓
The right step: The quantity that descends must be a positive integer (or a non-negative integer): a denominator q, the value of one variable, ∣x∣+∣y∣+∣z∣, and so on. First make sure that the descending quantity takes positive integer values; only then say “it cannot keep getting smaller forever”.
“Descent needs positive integers, so that there is a floor”
Using Sₙ − Sₙ₋₁ for the general term and forgetting n = 1
✗
The wrong step: Given Sn=n2+2n+1, we have Sn=(n+1)2, so an=Sn−Sn−1=(n+1)2−n2=2n+1.
What went wrong? Think first, then open
Diagnosisan=Sn−Sn−1 only makes sense for n≥2 (S0 is not defined). a1 must be found separately: a1=S1=4, while the formula 2n+1 gives 3 at n=1. Check: if an=2n+1 held for every n, then S1 would be 3, not 4.
✓
The right step: a1=S1=4, and for n≥2, an=Sn−Sn−1=2n+1. So an={4,2n+1,n=1n≥2. The two pieces merge into one formula only when the constant term of Sn is 0.
The wrong step: In the arithmetic sequence 2,5,8,11,14,… we have a2+a3=a5, because 2+3=5.
What went wrong? Think first, then open
Diagnosisa2+a3=5+8=13, but a5=14. The property says that “two pairs of terms with the same index sum have the same sum”: am+an=ap+aq when m+n=p+q. Both sides must have two terms, and a5 is only one.
✓
The right step: a2+a3=a1+a4=2+11=13. If you insist on comparing with a5, the right statement is a2+a3=a5+a0, and a0=a1−d=−1: exactly the difference of 1.
Seeing 1, 2, 4, 8, 16 and deciding the next term is 32
✗
The wrong step: Take n points on a circle and join every pair by a chord (no three chords through one point). The numbers of regions are 1,2,4,8,16. So for n=6 there are 32 regions.
What went wrong? Think first, then open
DiagnosisThere are actually 31. The number of regions is Cn4+Cn2+1, a polynomial of degree 4, which happens to agree with 2n−1 for the first five terms. Finitely many terms never determine the next one: any number at all can come next, if you allow polynomials of high enough degree.
✓
The right step: The difference method comes with a condition: you must know that the sequence is arithmetic of order p (for example, because you can prove from the problem that its general term is a polynomial of degree p), and then use at least p+1 terms to pin it down. Type 1,2,4,8,16,31 into the difference machine: only the fourth differences are constant.
The initial values are a₁, a₂, but the constants are solved as if they were a₀, a₁
✗
The wrong step: a1=1,a2=4,an+2=5an+1−6an. The characteristic roots are 2,3. Put an=A⋅3n+B⋅2n; from A+B=1,3A+2B=4 we get an=2⋅3n−2n.
What went wrong? Think first, then open
DiagnosisThe system of equations uses n=0 and n=1, but the problem gives a1 and a2. Check: the formula gives a1=6−2=4, while the problem says a1=1. The whole sequence has been shifted by one place.
✓
The right step: Set up the equations with n=1,2: 3A+2B=1 and 9A+4B=4, which give A=32,B=−21, that is, an=2⋅3n−1−2n−1. Check that a1=1,a2=4.
“Once you have the constants, substitute the initial values back to check”
Combining two periods and thinking only of the least common multiple
✗
The wrong step: A sequence has period 4 and also period 6, so its least positive period is lcm(4,6)=12.
What went wrong? Think first, then open
Diagnosis12 is certainly a period, but it is not the least one. From an+6=an and an+4=an we get an+2=an+6−4=an (go forward 6, then back 4). So 2 is a period too. In general, the greatest common divisor(T1,T2) of two periods T1,T2 is also a period (by Bézout's identity, (T1,T2)=uT1−vT2).
✓
The right step: Periods can be “subtracted”: gcd(4,6)=2 is a period. Whether the least positive period is 2 or 1 depends on the sequence itself. Also, every period is a multiple of the least positive period.
“Periods can be subtracted, so look at the greatest common divisor”
The number of tilings is off from the Fibonacci index by one
✗
The wrong step: A 2×n strip has Fn tilings, so a 2×5 strip has F5=5 tilings.
What went wrong? Think first, then open
DiagnosisIn this book's notation F1=F2=1. A 2×1 strip has 1 tiling and a 2×2 strip has 2 (two upright or two lying), so the tiling counts start 1,2,3,5,… — one place ahead of Fn, which starts 1,1,2,3,….
✓
The right step: A 2×n strip has Fn+1 tilings; a 2×5 strip has F6=8. Line up the indices with the two or three smallest n before using a formula.
Treating a necessary condition as a sufficient one
✗
The wrong step: We know: if Fn is prime (n=4), then n is prime. 19 is prime, so F19 is prime.
What went wrong? Think first, then open
DiagnosisThe first sentence says: if n is composite (n=4), then Fn has a proper factor Fd (d a proper divisor of n), so it is not prime. It does not say that Fn is prime whenever n is. In fact F19=4181=37×113.
✓
The right step: “n is prime” is only a necessary condition for Fn to be prime. To decide about F19 you have to factor it separately.
Proving it for powers of 2 only, then claiming it for every n
✗
The wrong step: For n=2, 2a+b≥ab; if it holds for n, it holds for 2n. By mathematical induction the AM–GM inequality holds for every n.
What went wrong? Think first, then open
DiagnosisStarting from 2, “n⇒2n” only reaches 2,4,8,16,…. It never touches n=3,5,6,7: only a few rungs of the ladder are lit, and the ones in between stay dark.
✓
The right step: Add one more step, “n+1⇒n” (or in general m⇒n for m>n): above every n there is a proved 2k, and you walk down from there one rung at a time.
“If you jumped up, you must also be able to climb down”
The wrong step: To prove the case of 3 numbers, append a 1 to a,b,c and use the result for 4 numbers: 4a+b+c+1≥4abc.
What went wrong? Think first, then open
DiagnosisThat inequality is true, but it does not give 3a+b+c≥3abc: the added 1 changes the mean, and both sides now contain something you do not want.
✓
The right step: Pad with their own mean A=3a+b+c. The mean of the 4 numbers is still A, so A≥4abcA, that is A4≥abcA, so A3≥abc.
“The number you add must leave the mean unchanged”
The wrong step: We want P(n) for every n≥1. We checked P(1) and proved P(n)⇒P(n+2), so the statement holds.
What went wrong? Think first, then open
DiagnosisP(1) knocks over P(3), P(3) knocks over P(5), … only the odd numbers fall. Nobody pushes P(2), so not one domino in the even chain falls.
✓
The right step: With step size k you must check k consecutive base cases P(1),…,P(k). Each base starts one chain, and together they cover every remainder on division by k.
A “convenient” step size whose step does not actually work
✗
The wrong step: If n cents can be paid, so can n+1 cents (n≥8): just pay 1 cent more.
What went wrong? Think first, then open
DiagnosisThere are only 3-cent and 5-cent stamps, no 1-cent ones. The inductive step must really be doable: from a way of paying n, build a way of paying n+1.
✓
The right step: Either change the step size — add one 3-cent stamp, n⇒n+3 — or do step size 1 properly: if the payment uses a 5, swap one 5 for two 3s; if it has no 5 (so at least three 3s), swap three 3s for two 5s.
“Choose the step size so that the step really works”
The wrong step: Induct on m to prove D(m,n)=n⋅2m. Induction hypothesis: D(m−1,n)=n⋅2m−1. Then D(m,n)=D(m−1,2n)=…
What went wrong? Think first, then open
DiagnosisThe step needs D(m−1,2n): the second variable has changed from n to 2n. If the hypothesis only says “true for thisn”, it is of no use.
✓
The right step: Put the other variable into the hypothesis, and make it arbitrary: “for every n, D(m−1,n)=n⋅2m−1”. Then use it at 2n: D(m,n)=D(m−1,2n)=2n⋅2m−1=n⋅2m.
“Induct on one variable, and let the hypothesis cover every value of the other”
Inducting on m+n when the step does not make the sum smaller
✗
The wrong step: The inductive step for P(m,n) uses P(m−1,n+1). Induct on m+n: P(m−1,n+1) is a smaller case, so it holds by the induction hypothesis.
What went wrong? Think first, then open
Diagnosis(m−1)+(n+1)=m+n: the sum has not decreased, so this case is not in the hypothesis “m′+n′<m+n”. Inducting on m+n only works if every case the step uses has a smaller sum.
✓
The right step: Look for the quantity that really decreases in this step: here it is m. Induct on m, assuming the statement for m−1 and everyn.
“Whatever you induct on must get smaller at every step”
The wrong step: Suppose Sn=1+221+⋯+n21≤2. Then Sn+1=Sn+(n+1)21, and (n+1)21 is tiny, so Sn+1≤2.
What went wrong? Think first, then open
DiagnosisThe induction hypothesis only tells you Sn≤2; it says nothing about how far Sn is from 2. If Sn were exactly 2 (the hypothesis allows that), adding any positive number would go over 2. “Tiny” is not a proof.
✓
The right step: Write “how far from 2” into the statement: prove Sn≤2−n1. The inductive step then needs exactly (n+1)21≤n1−n+11=n(n+1)1, which is clearly true.
“Room that is not in the hypothesis cannot be used in the conclusion”
The wrong step: To be safe, prove the stronger Sn≤2−2n3. The inductive step (n+1)21≤2n(n+1)3 holds, so the original statement is proved.
What went wrong? Think first, then open
DiagnosisThe inductive step does hold, but the base case fails: for n=1, S1=1 while 2−23=21. The strengthened statement is simply false, so of course it cannot be proved.
✓
The right step: A strengthening must satisfy both ends: the step must go through (enough room), and the base case must stand (not too much room). For 2−na, the step needs a≥1 and the base needs a≤1: only a=1 works.
“The strengthened statement must, first of all, be true”
The wrong step: Join n points on a circle by all possible chords: the disc is cut into 1,2,4,8,16 regions (n=1,…,5). So n points give 2n−1 regions.
What went wrong? Think first, then open
DiagnosisFor n=6 there are only 31 regions, not 32. The first five agreeing is a coincidence: the true answer is 1+Cn2+Cn4, a polynomial of degree four, while 2n−1 is exponential; the two agree only for the first few terms.
✓
The right step: Small cases are only for guessing. A guessed formula must be confirmed by induction (or some other proof). Try to prove 2n−1 and you will find you cannot give any reason for the step “one more point doubles the number of regions”.
The wrong step: Guess an=n+1n. Proof: an+1=n+2n+1; substitute into the recurrence an+1=2−an1, check that both sides are equal, so the guess is right.
What went wrong? Think first, then open
DiagnosisThis uses the very conclusion to be proved for an+1. The right direction is: assume onlyan=n+1n, computean+1 from the recurrence, and see whether it comes out as n+2n+1.
✓
The right step: an+1=2−n+1n1=2(n+1)−nn+1=n+2n+1. Work from the hypothesis to the conclusion, not backward from the conclusion.
The wrong step: Arrange 1 to 8 by “always place the smallest number allowed”: 1,2,4,5,…, and halfway through there is nothing left to place. So no such arrangement exists.
What went wrong? Think first, then open
DiagnosisOne failed attempt only shows that this way of arranging does not work. “Does not exist” is a claim about every arrangement: either check them all, or give a reason that applies to every arrangement. In fact 1,5,3,7,2,6,4,8 works.
✓
The right step: To prove existence, look for a construction; to prove non-existence, find a reason that “no arrangement can escape” (counting, pigeonhole, parity, an invariant).
The pigeonhole principle says “there is one”, so you think you know which
✗
The wrong step: Any sequence of 10 distinct numbers has a monotone subsequence of length 4, so the first 4 terms (or the last 4) must be monotone.
What went wrong? Think first, then open
DiagnosisThe pigeonhole principle only guarantees that such a subsequence exists; it does not say where. In 3,2,1,6,5,4,9,8,7,10 neither the first 4 nor the last 4 terms are monotone; a monotone subsequence is 1,4,7,10, taken with gaps.
✓
The right step: An indirect existence proof only tells you “there is one”; to point to it you must search separately (for example by following the labels used in the proof).
“An indirect proof guarantees there is one, but does not point the way”
Vol. 6 The Mean Inequalities and the Cauchy Inequality hall
Exhibit
Forgetting the “non-negative” condition
✗
The wrong step: For every x=0, x+x1≥2x⋅x1=2.
What went wrong? Think first, then open
DiagnosisThe mean inequality only holds for non-negative numbers. At x=−1 the left side is −2, far below 2. The arithmetic mean of two negative numbers is negative, while the geometric mean ab is positive.
✓
The right step: Split into cases: for x>0, x+x1≥2; for x<0, apply the inequality to −x>0 and get x+x1≤−2.
“Before using AM–GM, make sure every term is non-negative”
The wrong step: a,b>0 and a+b=4, so ab≤2 and the maximum of ab is 4. (No word on when it is reached.)
What went wrong? Think first, then open
DiagnosisHere it happens to be right: it is reached at a=b=2. But “≤” only says that 4 is an upper bound. Whether it is the maximum depends on whether equality can actually hold. The next stop and Chapter 2 show many bounds that cannot be reached.
✓
The right step: Every time you use the mean inequality, add a sentence “equality holds if and only if …”, and check that this condition is compatible with the other conditions of the problem.
“An inequality is only half the story; the equality condition is the other half”
Averaging pairs and assuming everything becomes equal in finitely many steps
✗
The wrong step: For 1,2,6, keep replacing two numbers by their mean. After finitely many steps all three become 3, and the product keeps growing, so 1⋅2⋅6≤33.
What went wrong? Think first, then open
DiagnosisThe conclusion is right, the argument is not. After averaging, the new number equals 3 if and only if the two averaged numbers add up to 6, which means the third number is already 3. There is no 3 at the start, so a 3 can never appear: the process does not stop after finitely many steps.
✓
The right step: Use the adjustment method: at each step make one number exactly equal to the mean A, and let the other one absorb the difference. Each step adds one more A, so it ends after at most n−1 steps. (Or take a limit of the infinite process, but that needs more argument.)
“For an argument by “finitely many adjustments”, every step must make definite progress on something”
Taking “half and half” as the weights without thinking
✗
The wrong step: a2b≤2a3+b3; adding the cyclic versions gives a2b+b2c+c2a≤a3+b3+c3.
What went wrong? Think first, then open
DiagnosisThe conclusion happens to be true, but the first step is false: with a=1, b=0.8 the left side is 0.8 and the right side is 21+0.512=0.756. On the right, the average exponent of a is 23+0=23, not 2: the two sides do not match up.
✓
The right step: Choose the weights so that the exponents match: the exponent of a is 2=3⋅32+0⋅31, so a2b≤32a3+31b3, that is, a3+a3+b3≥3a2b.
“Match the exponents first, then write the inequality”
Using AM–GM twice with clashing equality conditions
✗
The wrong step: a,b>0 and a+b=1. Since a+a1≥2 and b+b1≥2, the minimum of (a+a1)(b+b1) is 4.
What went wrong? Think first, then open
DiagnosisThe two inequalities need a=1 and b=1 for equality, but then a+b=2=1. Equality cannot be reached, so 4 is only a lower bound, not the minimum. In fact at a=b=21 the product is (25)2=425, and that is the minimum.
✓
The right step: Expand: ab+ab1+ba+ab. Since ab≤41 and t+t1 is decreasing on 0<t≤41, we get ab+ab1≥417; also ba+ab≥2. Both equalities hold at a=b=21, so they are compatible, and the minimum is 425.
A chain of inequalities, each with its own equality case
✗
The wrong step: For a triangle with perimeter 6: S=21absinC≤21ab≤21(2a+b)2<21⋅32=29, so the maximum area is 29.
What went wrong? Think first, then open
DiagnosisEvery step is true, but the three equalities need C=90∘, a=b and a+b=6 (that is, c=0) respectively. They cannot all hold at once, so 29 is never reached. The true maximum is 3≈1.73, less than half of that.
✓
The right step: Switch to the tangent lengths x,y,z: x+y+z=p=3 and S=3xyz≤3⋅13=3, with equality at x=y=z=1, the equilateral triangle with side 2. This one is reached.
“In a chain of bounds, all the equality cases must hold at once, in the same figure”
Minimizing 1/a + 4/b under a + b = 1 with two AM–GMs in a row
✗
The wrong step: a1+b4≥2ab4=ab4≥2a+b4=8, so the minimum is 8.
What went wrong? Think first, then open
DiagnosisThe first equality needs a1=b4, that is, b=4a; the second needs a=b. Both hold together only at a=b=0, which is impossible. 8 is never reached; it is only a lower bound.
✓
The right step: Multiply by 1=a+b: (a+b)(a1+b4)=5+ab+b4a≥5+4=9, with equality at b=2a, that is, a=31, b=32, which is allowed. The minimum is 9.
“Get there with one application; do not let two equality cases fight”
The wrong step: xy≤2x2+y2 and 2yz≤y2+z2; adding, xy+2yz≤21x2+23y2+z2≤23(x2+y2+z2), so the maximum is 23.
What went wrong? Think first, then open
DiagnosisEvery step is true, but equality cannot be reached. The first two steps need x=y=z, and the last step enlarges 21x2 and z2 to 23x2 and 23z2, so equality there needs x=z=0, and then y is 0 as well. The true maximum is 25≈1.118.
✓
The right step: Split y2 in an unknown ratio t:(1−t), and choose t so that the three square terms get equal coefficients: t=51.
“Wherever the coefficients are unbalanced, that is where the slack hides”
The wrong step: By the Cauchy inequality, (a+b+c)2≤a2+b2+c2.
What went wrong? Think first, then open
DiagnosisThe right-hand side of the Cauchy inequality is the product of two sums of squares. Here the second list is 1,1,1; its sum of squares is 3, and you cannot drop it. Take a=b=c=1: the left side is 9, the right side 3, so the inequality as written is simply false.
✓
The right step: (a⋅1+b⋅1+c⋅1)2≤(a2+b2+c2)(12+12+12), that is, (a+b+c)2≤3(a2+b2+c2).
“Write down both lists first, then apply the formula”
The wrong step: (x+2y)2≤5(x2+y2), with equality when x=y.
What went wrong? Think first, then open
DiagnosisTake x=y=1: the left side is 9 and the right side 10, not equal. The two lists are (x,y) and (1,2); equality needs them to be proportional, not x equal to y.
✓
The right step: Equality holds if and only if 1x=2y, i.e. y=2x. Check: at x=1,y=2 both sides are 25.
“Equality is about the ratio between the two lists, not about the numbers inside one list”
DiagnosisThe left side is 1−2=−1, far below 8. The fraction form comes from the Cauchy inequality applied to (yixi) and (yi), and yi needs yi>0.
✓
The right step: Before using ∑yixi2≥∑yi(∑xi)2, make sure every denominator is positive.
DiagnosisThe left side is only 4. The merging formula needs each numerator to be a square: yx2. When the numerator is 3, the matching x is 3, not 3.
✓
The right step: Write 112+1(3)2≥2(1+3)2≈3.73; that is correct. Or first multiply top and bottom by the same number to make the numerator a square.
“Read each numerator as a square first, then add up the things being squared”
Carrying on while the products still contain variables
✗
The wrong step: Given a+b+c=3, prove (a2+b2+c2)(a1+b1+c1)≥9: by Cauchy, the left side is ≥(a+b+c)2≥9.
What went wrong? Think first, then open
DiagnosisThe first step is fine, but the right side (a+b+c)2 still contains variables. And it is in fact at most3(a+b+c)=9 (Cauchy again), so the second step goes the wrong way.
✓
The right step: Here bound each factor on its own: a2+b2+c2≥3(a+b+c)2=3 and a1+b1+c1≥a+b+c9=3; multiplying gives ≥9.
“A good construction makes the products constant; if variables are left, check the direction first”
Checking proportionality for equality but forgetting the sign
✗
The wrong step: x2+y2=5, x+2y=−5. From the equality condition 1x=2y, take x=1,y=2.
What went wrong? Think first, then open
DiagnosisSubstituting (1,2) into the second equation gives 5, not −5. In the proportionality (x,y)=k(1,2) the factor k can be positive or negative; x+2y=5k=−5 needs k=−1.
✓
The right step: Set (x,y)=k(1,2); substituting into x+2y=−5 gives k=−1, so the solution is (−1,−2). Then check x2+y2=5.
“Write the equality condition as $(x,y)=k(a,b)$, then let the equations decide $k$”
The wrong step: To prove ∑b+ca≥23: ∑b+c(a)2≥2(a+b+c)(a+b+c)2≥23.
What went wrong? Think first, then open
DiagnosisThe first step is right, the second is wrong: by Cauchy, (a+b+c)2≤3(a+b+c), so this lower bound is at most23 and only reaches it when a=b=c. The merge threw away too much.
✓
The right step: Use b+ca=a(b+c)a2 instead. Merging gives 2(ab+bc+ca)(a+b+c)2, and (a+b+c)2≥3(ab+bc+ca) always holds.
“Before merging, ask whether the step that remains will still be true”
The wrong step: A 5×5 board without rectangles has T filled cells. On average a row has 5T cells and uses CT/52 column pairs, so 5CT/52≤10; and from this we conclude that every row has exactly 5T cells.
What went wrong? Think first, then open
DiagnosisDifferent rows can have different numbers of cells. What lets you replace ∑Cri2 by “the case of the average” is the Cauchy inequality ∑ri2≥5T2, and that only gives an inequality; it does not say the rows really are equal.
✓
The right step: Write it out: ∑ri2≥5(∑ri)2 (Cauchy), so 5T2−T≤∑ri2−∑ri≤20. Equal rows are only the equality condition of Cauchy.
“When Cauchy replaces “every row” by “the average”, that is an inequality, not an assumption”
Using AM–GM twice with clashing equality conditions
✗
The wrong step: a+b=1, a,b>0. a1+b8≥2ab8≥232=82, so the minimum is 82.
What went wrong? Think first, then open
DiagnosisEquality in the first step needs a1=b8, i.e. b=8a; equality in the second step, ab≤41, needs a=b. The two conditions cannot hold together, so 82≈11.3 is never reached. The true minimum is about 14.66.
✓
The right step: Use Cauchy (fraction form), which gets there in one step: a12+b(22)2≥a+b(1+22)2=9+42, with equality when b=22a.
“When you chain several inequalities, their equality conditions must hold together”
The wrong step: On the number line A has coordinate 3 and B has coordinate −2, so AB=5.
What went wrong? Think first, then open
Diagnosis5 is the length∣AB∣. The signed value AB also carries a direction: going from A to B means moving in the negative direction, so AB=−2−3=−5. Length only asks “how far”; the signed value also asks “which way”.
✓
The right step: Always take “end coordinate minus start coordinate”: AB=xB−xA=−2−3=−5, and ∣AB∣=5.
“Signed values have a sign, lengths do not; write the end first, then subtract the start”
The wrong step: On the line P1P2 there is always a point P with λ=PP2P1P=−1.
What went wrong? Think first, then open
Diagnosisλ=−1 means P1P=−PP2=P2P, that is x−x1=x−x2, so x1=x2: the two endpoints coincide, which contradicts having a segment. The farther P runs, the closer λ gets to −1, but it never gets there.
✓
The right step: The denominator 1+λ in the division formula x=1+λx1+λx2 already warns you: λ=−1.
“A denominator in a formula marks the value the quantity can never take”
The wrong step: P is the trisection point of P1P2 nearer to P1, so λ=P1P2P1P=31.
What went wrong? Think first, then open
DiagnosisFor a dividing point, λ is PP2P1P: the piece before the point over the piece after it, not over the whole segment. The trisection point nearer to P1 splits the segment as 1:2, so λ=21.
✓
The right step: Substitute λ=21 into x=1+λx1+λx2 to get x=32x1+x2, which is exactly “one third of the way from P1”.
“From the start to the dividing point, then from the dividing point to the end”
The wrong step: A(0,0), B(0,2), C(3,0): S=21003020111=−3. An area cannot be negative, so something must have gone wrong.
What went wrong? Think first, then open
DiagnosisThe determinant gives a signed area: positive when the vertices go counterclockwise, negative when they go clockwise. A→B→C is clockwise (up first, then down to the right), so −3 is correct.
✓
The right step: Take the absolute value: S=3. To add the areas of several triangles, list all their vertices in the same direction, or take each absolute value before adding.
“The minus sign tells you the direction, not that you made a mistake”
Choosing coordinates “too well” and losing the general case
✗
The wrong step: To prove a property of a triangle, take A(0,a), B(−a,0), C(a,0).
What went wrong? Think first, then open
DiagnosisWith this choice, △ABC has quietly been restricted to an isosceles right triangle (∣AO∣=∣BO∣=∣CO∣). You may choose the coordinate system freely, but the coordinate system must not decide the shape of the figure “on the side”.
✓
The right step: Choosing the axes only uses up the freedom of the figure’s position (translation and rotation). Once the triangle is placed, it still has 3 independent quantities: A(0,a), B(b,0), C(c,0), with a, b, c unrelated.
The wrong step: The locus of points at distance 3 from the y-axis has equation x=3.
What went wrong? Think first, then open
DiagnosisThe distance from a point to the y-axis is ∣x∣, not x. The point (−3,0) is also at distance 3 from the y-axis but does not satisfy x=3: condition ① (every point of the curve satisfies the equation) is broken.
✓
The right step: ∣x∣=3, i.e. x=3 or x=−3: two lines. It can also be written as x2−9=0.
“Distances need absolute values, and removing an absolute value means splitting into cases”
The wrong step: Square both sides of the curve y=9−x2 to get its equation x2+y2=9.
What went wrong? Think first, then open
Diagnosisy=9−x2 requires y≥0: it is the upper half of the circle, while x2+y2=9 is the whole circle. The point (0,−3) satisfies the second but is not on the original curve: squaring was not an equivalent transformation.
✓
The right step: Write x2+y2=9(y≥0). At every step of simplifying an equation, ask: could this step add solutions, or lose solutions?
“Simplifying only works if every step is an equivalence”
“Equal intercepts”, but only the intercept form was tried
✗
The wrong step: The line through (4,1) with equal intercepts on the two axes: let ax+ay=1; substituting gives a=5, so the only answer is x+y−5=0.
What went wrong? Think first, then open
DiagnosisThe intercept form requires both intercepts to be nonzero. The line y=41x through the origin has both intercepts equal to 0, so they are “equal” too, yet the intercept form rules it out.
✓
The right step: Split into two cases. Nonzero intercepts give x+y−5=0; zero intercepts mean the line passes through the origin, giving x−4y=0. There are two answers.
“Whatever form you use, first ask which lines it leaves out”
The wrong step: The line x=3 is perpendicular to the x-axis, its angle of inclination is 90∘, so its slope is k=0.
What went wrong? Think first, then open
DiagnosisSlope 0 belongs to a horizontal line (angle of inclination 0∘). tan90∘ is undefined, so a vertical line has no slope; in the formula x2−x1y2−y1 the denominator is 0.
✓
The right step: x=3 has angle of inclination 90∘ and no slope. It cannot be written in point–slope or slope–intercept form, but it can be written in general form as x−3=0.
The wrong step: To put 3x−4y+15=0 in normal form, divide both sides by 5: 53x−54y+3=0, so p=−3.
What went wrong? Think first, then open
DiagnosisThe constant term of the normal form is −p, which must be ≤0. After multiplying by +51 the constant term is +3, so this is not the normal form. The sign of μ must be opposite to that of C: C=15>0, so take μ=−51.
✓
The right step: −53x+54y−3=0: cosθ=−53, sinθ=54, p=3.
“Give μ the sign opposite to C, and p will never be negative”
The wrong step: The line in normal form xcos30∘+ysin30∘−2=0 has angle of inclination 30∘.
What went wrong? Think first, then open
Diagnosisθ is the direction of the normalON, and the line is perpendicular to it. The angle of inclination of the line is θ+90∘ or θ−90∘ (whichever lies in [0∘,180∘)).
✓
The right step: Here the angle of inclination is α=120∘. Check: the slope is −sin30∘cos30∘=−3=tan120∘.
Distance between parallel lines without matching the coefficients
✗
The wrong step: The distance between 3x+4y−7=0 and 6x+8y+1=0 is 32+42∣−7−1∣=58.
What went wrong? Think first, then open
DiagnosisThe formula A2+B2∣C1−C2∣ needs the two equations to have exactly the sameA and B. The coefficients of x and y in the second equation are twice those in the first, so its constant term 1 must be divided by 2 as well.
✓
The right step: Write the second line as 3x+4y+21=0; the distance is 5∣−7−21∣=23. Or take a point (1,1) on the first line and find its distance to the second: 10∣6+8+1∣=23.
The wrong step: Find the line through the intersection of l1:x−y+1=0 and l2:x+y−3=0 that also passes through (4,−1). Let (x−y+1)+λ(x+y−3)=0; substituting gives 6+0⋅λ=0, which has no solution, so no such line exists.
What went wrong? Think first, then open
Diagnosis(4,−1) happens to lie on l2 (4−1−3=0). The family l1+λl2=0 contains every line through the intersection point except l2 itself, so “no solution” is telling you precisely that the answer is l2.
✓
The right step: The answer is x+y−3=0. To miss nothing, you can use μl1+λl2=0 (μ, λ not both 0) instead.
“When λ has no solution, first check whether l₂ is the answer”
The wrong step: x2+y2−2x+4y+6=0 is a circle with center (1,−2).
What went wrong? Think first, then open
DiagnosisCompleting the square gives (x−1)2+(y+2)2=−1. The left side is a sum of two squares and can never be negative, so no point at all satisfies the equation. D2+E2−4F=4+16−24=−4<0 already told you so.
✓
The right step: Compute D2+E2−4F first: greater than 0 gives a circle, equal to 0 a single point, less than 0 no points at all.
“After completing the square, check the sign of the right side first”
Reading off the center when the square terms do not have coefficient 1
✗
The wrong step: The center of 2x2+2y2−8x+4y+1=0 is (4,−2).
What went wrong? Think first, then open
DiagnosisThe formula (−2D,−2E) is for an equation in which x2 and y2 both have coefficient 1. Here the coefficient is 2, so plugging in directly doubles the center.
✓
The right step: Divide both sides by 2 first: x2+y2−4x+2y+21=0, so the center is (2,−1) and r2=4+1−21=29.
The wrong step: The foci of 9x2+25y2=1 are (±4,0).
What went wrong? Think first, then open
DiagnosisThe major axis follows the larger denominator. Here 25 is under y2, so the major axis is the y-axis, with a2=25 and b2=9. The points (±4,0) are not on the major axis at all; they are even outside the ellipse (916>1).
✓
The right step: c2=25−9=16, so the foci are (0,±4).
“Compare the denominators first; the larger one tells you which axis is the major axis”
The wrong step: For 16x2+9y2=1 the half focal distance is c=16+9=5.
What went wrong? Think first, then open
DiagnosisIn an ellipse, a is the hypotenuse: the distance from an end of the minor axis to a focus is exactly a (the string folded in half), so b2+c2=a2. With c=5 larger than a=4, the foci would lie outside the ellipse.
✓
The right step: c2=a2−b2=7, so c=7. (c2=a2+b2 is the relation for the hyperbola.)
The wrong step: The asymptotes of 4y2−9x2=1 are y=±23x.
What went wrong? Think first, then open
DiagnosisThe transverse axis of this hyperbola is the y-axis, and a and b got mixed up when applying y=±abx. Check: for large x, 4y2≈9x2, so y≈±32x.
✓
The right step: Replace the 1 on the right side by 0: 4y2−9x2=0, that is, y=±32x. This way you never have to remember which is a and which is b.
Finding a hyperbola's foci with the ellipse relation
✗
The wrong step: The foci of the hyperbola 16x2−9y2=1 are (±7,0).
What went wrong? Think first, then open
Diagnosisc2=a2−b2 is the relation for the ellipse. The vertices of the hyperbola are at (±4,0), yet these foci at (±7,0) would lie between the two branches, while each focus of a hyperbola should be “wrapped” inside its own branch.
✓
The right step: For a hyperbola c2=a2+b2=25, so the foci are (±5,0). How to remember: c is half the diagonal of the box.
“In an ellipse a is the largest; in a hyperbola c is the largest”
DiagnosisIn y=2x2 the squared term is x2, so the axis of symmetry is the y-axis and the focus must lie on the y-axis. Also, 2 is not 2p: first write the equation in the form x2=2py.
✓
The right step: x2=21y, so 2p=21, p=41; the focus is (0,2p)=(0,81) and the directrix is y=−81.
“First write it as x²=2py or y²=2px, then read off p”
Using the replacement rule at a point off the curve
✗
The wrong step: Tangents from the point (0,2) to the ellipse 4x2+y2=1: the replacement rule gives 40⋅x+2y=1, that is, y=21.
What went wrong? Think first, then open
DiagnosisThe replacement rule gives the tangent only at a point on the curve. (0,2) is not on the ellipse, and the line y=21 does not even pass through (0,2); it is actually the line joining the two points of tangency (the chord of contact).
✓
The right step: Let the tangent be y=kx+2, substitute into the ellipse and set the discriminant to 0: (1+4k2)x2+16kx+12=0, Δ=256k2−48(1+4k2)=0, k=±23. Both points of tangency have y-coordinate 21: they lie exactly on that “wrong answer”.
Getting the signs of the translation formulas backwards
✗
The wrong step: (x−2)2+4(y+2)2=16: put x′=x+2, y′=y−2 and get x′2+4y′2=16.
What went wrong? Think first, then open
DiagnosisYou need x′=x−2 to write (x−2)2 as x′2. The new origin O′ is (2,−2) in the old coordinates, and the formulas are x=x′+h, y=y′+k, that is, x′=x−h, y′=y−k.
✓
The right step: h=2, k=−2: x′=x−2, y′=y+2. The equation becomes x′2+4y′2=16, with center O′(2,−2).
“The old coordinates of the new origin are (h, k), and the brackets hold a minus sign”
The wrong step: To remove the xy-term of 5x2+4xy+2y2−6=0: cot2θ=BC−A=−43, θ≈63.4∘.
What went wrong? Think first, then open
DiagnosisSubstitute back into B′=Bcos2θ−(A−C)sin2θ: cos2θ=−0.6, sin2θ=0.8, so B′=4×(−0.6)−3×0.8=−4.8=0. With the sign flipped, the xy-term actually got bigger.
✓
The right step: B′=0 gives cot2θ=BA−C=43, θ≈26.6∘. If in doubt, check with the formula for B′.
The wrong step: x2+4xy+4y2+x=0: B=2, B2−4AC=4−16<0, so it is of elliptic type.
What went wrong? Think first, then open
DiagnosisHere B is the whole coefficient of the xy-term, B=4. (Some books write the equation as Ax2+2Bxy+⋯; the discriminant is then B2−AC. Do not mix the two conventions.)
✓
The right step: B2−4AC=16−16=0: parabolic type. In fact (x+2y)2=−x, which is a parabola.
“First make sure which convention the equation is written in”
The wrong step: Find a circle or line through the intersection points of the circle x2+y2=4 and the line x=1 that also passes through (1,0): let x2+y2−4+λ(x−1)=0; substituting (1,0) gives −3=0, so there is none.
What went wrong? Think first, then open
DiagnosisS1+λS2=0 can never be S2=0 itself (that would be λ=∞). The point (1,0) lies on the line x=1, and that very line is the answer.
✓
The right step: Write μS1+λS2=0 (μ,λ not both 0). Substituting gives μ=0: the answer is the line x−1=0. The three points are collinear, so a circle through them really does not exist.
Using $\theta=\arctan\dfrac yx$ without looking at the quadrant
✗
The wrong step: The point (−1,3): ρ=2, θ=arctan(−3)=−3π, so its polar coordinates are (2,−3π).
What went wrong? Think first, then open
Diagnosisarctan only takes values in (−2π,2π), so it only ever points into the first or fourth quadrant. (−1,3) is in the second quadrant, but (2,−3π) is the fourth-quadrant point (1,−3): exactly half a turn away.
✓
The right step: tanθ=−3 only tells you the line the terminal side lies on. Now look at x<0,y>0 and take the angle in the second quadrant: θ=32π, so the polar coordinates are (2,32π).
“The tangent only sees the line; the quadrant is your job”
Seeing $\theta$ in the first quadrant and assuming the point is there too
✗
The wrong step: The point (−3,4π) has polar angle 4π, so it lies in the first quadrant.
What went wrong? Think first, then open
DiagnosisWhen ρ<0 the point is not on the terminal side of θ but on its extension backward through the pole. (−3,4π) and (3,45π) are the same point, in the third quadrant.
✓
The right step: The formulas are the safest way: x=−3cos4π=−232, y=−3sin4π=−232. Both are negative.
One substitution fails, so you conclude there is no symmetry
✗
The wrong step: ρ=2sin2θ: replacing θ by −θ gives ρ=−2sin2θ. The equation changed, so the curve is not symmetric about the polar axis.
What went wrong? Think first, then open
DiagnosisThe mirror image of (ρ,θ) in the polar axis can be written as (ρ,−θ), but also as (−ρ,π−θ), and a point of the curve may satisfy the equation under only one of its names. Substitute the second form: −ρ=2sin(2π−2θ)=−2sin2θ. The equation is unchanged, so the curve is in fact symmetric.
✓
The right step: Equation unchanged after a substitution ⇒ definitely symmetric. Equation changed ⇒ no conclusion yet: try the other form, or look at the graph and test points.
“A substitution test is sufficient, not necessary”
Letting θ run only from 0 to 2π and thinking the curve is complete
✗
The wrong step: ρ=2cos2θ has period 4π, but one full turn of the polar angle is 2π, so letting θ go from 0 to 2π draws the whole curve.
What went wrong? Think first, then open
Diagnosisθ and θ+2π point the same way, but cos2θ+2π=−cos2θ: ρ changes sign and the point lands on the other side. Drawing only [0,2π] loses half the curve.
✓
The right step: To graph ρ=f(θ), let θ run through a full period of f (here [0,4π]), or first prove that the second half is a mirror image of the first.
“One turn of the ray is not one period of the equation”
Taking $p$ to be the distance from the directrix to the center
✗
The wrong step: For the ellipse 25x2+16y2=1, e=53 and the directrix is x=325, so p=325 and the polar equation is ρ=1−53cosθ5.
What went wrong? Think first, then open
DiagnosisIn ρ=1−ecosθep, p is the distance from the focus to its directrix, not from the center to the directrix. 325 is the distance from the directrix to the center O; you still have to subtract c=3.
✓
The right step: p=ca2−c=325−3=316, and ep=516=ab2 (half the latus rectum). Check: at θ=2π, ρ=ep, which is exactly half the latus rectum.
“ep is half the latus rectum; check it with θ = π/2”
Copying the ellipse formula for a focal chord of a hyperbola, with absolute values
✗
The wrong step: A focal chord of a hyperbola also satisfies ∣FA∣1+∣FB∣1=ep2.
What went wrong? Think first, then open
DiagnosisWhen the chord meets both branches of the hyperbola, A and B are on the same side of the focus and one of the radii is negative. The identity ρ11+ρ21=ep2 always holds for the signedρ; with absolute values it becomes ∣FA∣1−∣FB∣1=ep2.
✓
The right step: In the widget, set e to 1.3 and turn α: watch when the ∣FA∣ readout turns yellow with “ρ<0”. First decide whether the chord meets one branch or two, then decide between sum and difference.
“The formula holds for signed ρ; absolute values need cases”
Taking the $\varphi$ in the ellipse’s parametric equations to be the inclination of $OP$
✗
The wrong step: On the ellipse x=4cosφ,y=2sinφ, the point where OP makes a 60∘ angle with the x-axis is the point with φ=60∘, namely (2,3).
What went wrong? Think first, then open
Diagnosisφ is the eccentric angle: it is the angle of the point M on the auxiliary large circle, not the angle of P. The line from O to (2,3) has inclination with tanθ=23, about 40.9∘, not 60∘.
✓
The right step: You need xy=tan60∘, that is 4cosφ2sinφ=3, so tanφ=23. The two angles agree only for φ=0,2π,π,23π.
“The eccentric angle lives on the auxiliary circle, not on OP”
Handing in as soon as the parameter is gone, forgetting the ranges of x and y
✗
The wrong step: x=t,y=2t−1: from t=x2 we get y=2x2−1, so the curve is a parabola.
What went wrong? Think first, then open
Diagnosisx=t can never be negative. The relation t=x2 used in the elimination also holds for x<0, so the Cartesian equation “invites in” the left half as well, although no t ever gets there.
✓
The right step: The answer must be written y=2x2−1(x≥0): the right half of the parabola (including the vertex (0,−1)).
“Was the elimination reversible? Take a look at the ranges of x and y”
The wrong step: The circle x2+y2=4 can be written x=1+t22(1−t2),y=1+t24t(t∈R), and the two describe the same curve.
What went wrong? Think first, then open
DiagnosisThese come from setting t=tan2θ, and tan2θ is undefined at θ=π. The corresponding point (−2,0) is never reached, whatever t is: x=−2 would need 1−t2=−(1+t2), that is 1=−1.
✓
The right step: Either add “together with the point (−2,0)”, or use x=2cosθ,y=2sinθ instead.
“The half-angle substitution loses the point θ = π”
Finding the “chord with a given midpoint” from $t_1+t_2=0$ without checking that the chord exists
✗
The wrong step: The chord of the ellipse 4x2+y2=1 with midpoint P0(2,1): from t1+t2=0 the slope is −a2y0b2x0=−21, so the chord lies on the line y=−21x+2.
What went wrong? Think first, then open
DiagnosisP0(2,1) is outside the ellipse (44+1=2>1). Substituting y=−21x+2 into the ellipse gives x2−4x+6=0 with discriminant 16−24<0: this line does not meet the ellipse at all, so where is the chord? t1+t2=0 only means something when there are two intersection points.
✓
The right step: First check that P0 is inside the curve (substituting gives a value less than 0), or, after finding the line, check that the discriminant is positive. When P0 is outside the ellipse, no chord has it as midpoint.
Taking $t_1t_2$ to be $|P_0A|\cdot|P_0B|$ directly
✗
The wrong step: P0(1,1) is inside the circle x2+y2=9, and a chord through it is cut into two pieces whose product is ∣P0A∣⋅∣P0B∣=t1t2=1+1−9=−7.
What went wrong? Think first, then open
Diagnosist is a signed distance: when P0 lies between A and B, t1 and t2 have opposite signs and their product is negative. A product of lengths needs an absolute value.
✓
The right step: ∣P0A∣⋅∣P0B∣=∣t1t2∣=7. Conversely, the sign of t1t2 itself tells you whether P0 lies between A and B.
The wrong step: 2x−y+1>0 is the region above the line y=2x+1.
What went wrong? Think first, then open
DiagnosisTest a point: the origin (0,0) gives 1>0, but the origin is below the line y=2x+1. The coefficient of y is −1, so the larger y is, the smaller F is: above the line, F<0. “Above means >0” is true only when the coefficient of y is positive.
✓
The right step: Rewrite the inequality as y<2x+1 before reading it, or simply test a point not on the line: 2x−y+1>0 is the region below the line (not including the line).
Testing with the origin when the origin is on the curve
✗
The wrong step: Which part is y2−4x>0? Substituting the origin gives 0, so… no idea. Maybe the whole plane?
What went wrong? Think first, then open
DiagnosisThe origin lies on the parabola y2=4x, and F=0 cannot stand for either side. The test point must not be on the curve.
✓
The right step: Use another point, for example (1,0): 0−4<0, and it is “inside” the parabola (on the side it opens toward). So y2−4x>0 is the piece outside the parabola.
Comparing only the vertices when the objective is not linear
✗
The wrong step: Minimum of x2+y2 on the triangle with vertices (1,3), (3,1), (4,4): the values at the three vertices are 10, 10, 32, so the minimum is 10.
What went wrong? Think first, then open
Diagnosis“The extremum is at a vertex” holds only for a linear objective: its level curves are lines, and a line sliding off a convex polygon touches a vertex last. The level curves of x2+y2=c are circles centered at the origin, and a growing circle may first touch the middle of an edge.
✓
The right step: The foot of the perpendicular from the origin to the edge x+y=4 is (2,2), which lies between (1,3) and (3,1), so the minimum is 22+22=8.
“Look at the shape of the level curves before deciding where to search”
The wrong step: Maximum of z=x−2y: slide the line x−2y=z upward; the bigger the intercept, the bigger z, so take the highest point of the region.
What went wrong? Think first, then open
Diagnosisx−2y=z becomes y=21x−2z, whose y-intercept is −2z: the bigger the intercept, the smallerz. When the coefficient of y is negative, every direction is reversed.
✓
The right step: Write down how the intercept is related to z before deciding the direction; or, easiest of all, substitute the candidate vertices one by one and compare.
“Whether intercept and z move together depends on the sign of y’s coefficient”
Removing an absolute value and keeping only one case
✗
The wrong step: The locus of points equidistant from the x-axis and the y-axis: ∣y∣=∣x∣, so y=x.
What went wrong? Think first, then open
Diagnosis∣y∣=∣x∣ is equivalent to y=xory=−x. Writing only y=x leaves out the points in the second and fourth quadrants (such as (2,−2)), which satisfy the condition but are not on the curve: completeness fails.
✓
The right step: The locus is the two lines y=x and y=−x, which can be written together as x2−y2=0.
Counting A and B in the locus of a right-angle vertex
✗
The wrong step: With A(−3,0) and B(3,0), the locus of points P with PA⊥PB has equation x2+y2=9.
What went wrong? Think first, then open
DiagnosisWhen P coincides with A, the “line PA” does not even exist, let alone being perpendicular to anything. A and B lie on the circle but do not satisfy the condition: purity fails.
✓
The right step: The locus is the circle x2+y2=9 without the two points (±3,0), written x2+y2=9(x=±3).
“Where the figure in the condition degenerates, that point does not count”
Substituting the wrong way round in the substitution method
✗
The wrong step: P is on the circle x2+y2=16 and Q is the midpoint of P and A(6,0). Let P(x,y); then Q(2x+6,2y), and substituting into the circle’s equation gives (2x+6)2+(2y)2=16.
What went wrong? Think first, then open
DiagnosisWe want the locus of Q, so (x,y) should be the coordinates of Q; and what gets substituted into the circle’s equation must be the point P, which is the one satisfying it. The version above pushes Q’s coordinates into P’s equation, and the resulting curve has nothing to do with Q.
✓
The right step: Let Q(x,y) and express P through it: P(2x−6,2y). P is on the circle, so (2x−6)2+(2y)2=16, that is (x−3)2+y2=4.
“Name the point you want, express the known point, substitute into the known equation”
Forgetting the excluded points after eliminating the parameter in the intersecting-loci method
✗
The wrong step: The line through A(−2,0) with slope k meets the line through B(2,0) with slope k1 at P. Multiplying the two equations gives y2=x2−4, so the locus is the whole hyperbola x2−y2=4.
What went wrong? Think first, then open
Diagnosisk=0 (otherwise k1 makes no sense), and the vertices (±2,0) of the hyperbola could only come from k=0 or an “infinite” slope. At k=±1 the two lines are parallel and there is no intersection: those are the directions of the asymptotes. The vertices (±2,0) are extra points.
✓
The right step: The locus is x2−y2=4 without the two points (±2,0). In the widget, pull θ to 0 and to ±45∘ and see.
“After eliminating the parameter, go back to 9.1 and check purity”
Substituting, then forgetting the range of the new variable
✗
The wrong step: Find the minimum of y=x+1−x2. Let x=cost. Then y=cost+sint=2sin(t+4π), so the minimum is −2.
What went wrong? Think first, then open
Diagnosis1−x2=∣sint∣, which equals sint only for t∈[0,π]. In that range t+4π∈[4π,45π], and sin only gets down to −22.
✓
The right step: Whenever you substitute, write down the range of the new variable as well: t∈[0,π]. Then y∈[−1,2], and the minimum is −1 (reached at x=−1).
The wrong step: From f(x)+2f(x1)=3x we get f(x)=3x−2f(x1), and that is the answer.
What went wrong? Think first, then open
DiagnosisThe right-hand side still contains the unknown f(x1), so nothing has been solved. There are two unknowns in one equation.
✓
The right step: Replace x by x1 to get a second equation. Two equations in two unknowns f(x) and f(x1): the problem has been reduced to a system of two linear equations.
The wrong step: To prove “at least one of a,b,c is greater than 1”, assume: “none of a,b,c is greater than 1” … that is fine. But to prove “a,b,c are all positive”, the assumption is written as “none of a,b,c is positive”.
What went wrong? Think first, then open
DiagnosisThe opposite of “all are” is “not all are”, that is, “at least one is not”. Writing “none is” leaves out the case where some are and some are not, so the contradiction only rules out part of the possibilities.
✓
The right step: Learn the common negations by heart: all are ↔ not all are; at least one ↔ none at all; at most one ↔ at least two; there exists ↔ every one fails. After writing the assumption, ask yourself: is exactly one of the original conclusion and the assumption true?
“The assumption must be the whole opposite of the conclusion”
The wrong step: Assume 3 is rational. Since 1<3<4, we have 1<3<2, so 3 is not an integer: contradiction.
What went wrong? Think first, then open
Diagnosis“Not an integer” does not contradict “rational”: 23 is not an integer either. The assumption never appeared in the argument, so it has not ruled anything out.
✓
The right step: Write the assumption as something you can compute with: 3=qp (in lowest terms), p2=3q2, so 3∣p … The contradiction has to come out of this equation.
The wrong step: Prove n=n+1. Suppose k=k+1. Adding 1 to both sides gives k+1=k+2, so P(k+1) holds. Hence it holds for all n.
What went wrong? Think first, then open
DiagnosisThe inductive step really is fine, but P(1) says 1=2, which is false. Every domino would knock over the next one, but nobody pushes the first, so not a single one falls.
✓
The right step: Induction needs both halves: the base case (P(n0) holds) and the inductive step (P(k)⇒P(k+1)).
The wrong step: Prove that every postage of n≥8 cents can be made with 3-cent and 5-cent stamps: 8=3+5 works; if k can be made, add one 3-cent stamp to get k+3. Done.
What went wrong? Think first, then open
DiagnosisStarting from 8 and stepping by 3 only proves 8,11,14,…. Nothing ever reaches 9, 10 or the numbers that follow them.
✓
The right step: With step size 3 you need three base cases: 8=3+5, 9=3+3+3, 10=5+5. Only the three chains together cover every n≥8.
The boxes are built, but sharing a box proves nothing
✗
The wrong step: Take 11 numbers from 1,…,20 and prove that one is a multiple of another: build 10 boxes from “consecutive pairs” {1,2},{3,4},…. Two of the 11 numbers share a box, so one is a multiple of the other.
What went wrong? Think first, then open
Diagnosis3 and 4 share a box, but neither is a multiple of the other. The number of boxes is right, but the link “same box ⇒ conclusion” is broken. Consecutive pairs guarantee coprime numbers.
✓
The right step: Build the boxes by odd part: {1,2,4,8,16},{3,6,12},{5,10,20},…, 10 of them. Two numbers in the same box differ only by some factors of 2, so one is a multiple of the other.
“First check that two in the same box are the pair you want”
The wrong step: What is the smallest number of numbers that guarantees two coprime ones? 51 numbers go into the 50 boxes {1,2},…,{99,100}, so the answer is 51.
What went wrong? Think first, then open
DiagnosisThe pigeonhole principle only shows that 51 numbers are enough. To call 51 the smallest, you must also show that 50 are not enough.
✓
The right step: Add an example: the 50 even numbers 2,4,6,…,100 are pairwise not coprime. So 50 are not enough, 51 are, and the answer is 51.
Stopping after subtracting the pairwise intersections
✗
The wrong step: Among 1∼100 there are 50+33+20−16−10−6=71 multiples of 2, 3 or 5.
What went wrong? Think first, then open
DiagnosisWhen you add the three sets, the numbers that are multiples of all of 2,3,5 (30,60,90) are counted 3 times; when you subtract the three pairwise intersections they are subtracted 3 times. In the end they are not counted at all.
✓
The right step: Add back the intersection of all three sets: 71+3=74. The formula for three sets has seven terms, with signs alternating +,−,+.
“Added three times and subtracted three times means never counted”
The wrong step: Among 1∼100 there are [24100]=4 numbers that are multiples of both 4 and 6.
What went wrong? Think first, then open
DiagnosisBeing a multiple of both 4 and 6 means being a multiple of their least common multiple12, not of their product 24. The number 12 itself was missed.
✓
The right step: There are [12100]=8. The lcm equals the product only when the two numbers are coprime.
Taking the “smallest solution” among positive reals
✗
The wrong step: Suppose x2+y2=3z2 has a solution in positive reals. Take the one with the smallest z, and then …
What went wrong? Think first, then open
DiagnosisThere is no smallest positive real: if (x,y,z) is a solution, so is (2x,2y,2z). Besides, this equation really does have solutions in positive reals (for example x=y=23, z=1).
✓
The right step: The well-ordering principle holds only for positive integers (or for finite sets). Infinite descent has to run among the positive integer solutions: take the positive integer solution with the smallest z.
The wrong step: In a round robin, take the player K with the most wins. Since K has the most wins, K beat everyone.
What went wrong? Think first, then open
DiagnosisMost wins is not the same as winning every game. In a five-player round robin everyone can have exactly 2 wins and 2 losses, and then nobody is unbeaten. The extreme element is only “the most”; what it guarantees has to be proved.
✓
The right step: What you can prove is this: for every other player, K either beat them directly or beat someone who beat them. The proof is by contradiction: if some v beat K and also beat everyone K beat, then v has more wins than K.
“The extreme element has an extreme property, but do not exaggerate it”
The wrong step: a+b is even, so a and b are both even. Let a=2s, b=2t…
What went wrong? Think first, then open
DiagnosisAn even a+b only says that a and b have the same parity. 3+5=8 is a sum of two odd numbers. Miss the “both odd” case and you have done only half of the argument.
✓
The right step: Write: a+b is even ⟺a≡b(mod2). Then treat the two cases “both even” and “both odd”.
The wrong step: The board shows 1∼10; each move erases two numbers and writes their difference. The sum is 55 (odd), so the last number can be any odd number, for example 11.
What went wrong? Think first, then open
DiagnosisA parity invariant only tells you that an even number cannot be reached. It does not promise that every odd number can. In fact each difference written is no bigger than the largest number on the board, so the last number is at most 10: 11 is out of reach.
✓
The right step: Parity can only rule things out. To say “it can be done”, you must give an actual sequence of moves.
Different bases, yet the area ratio is taken as the ratio of heights
✗
The wrong step: The areas of △ABD and △ACE are in the ratio 2:3, so their heights are in the ratio 2:3.
What went wrong? Think first, then open
DiagnosisArea =21× base × height. The area ratio equals the ratio of heights only when the bases are the same (or equal in length), and equals the ratio of bases only when the heights are the same.
✓
The right step: Before using an area ratio, find “what is shared”: a common side (ratio of heights), a common height (ratio of bases), or a common angle (ratio of the products of the two sides around it).
“First find a common side, a common height or a common angle”
Forgetting the absolute value when finding an area from coordinates
✗
The wrong step: A(0,0), B(1,3), C(4,1), S=21(xByC−xCyB)=21(1−12)=−211.
What went wrong? Think first, then open
Diagnosis21(x1y2−x2y1) is a signed area: positive when the vertices go counterclockwise, negative when they go clockwise. Here A→B→C is clockwise.
✓
The right step: Take the absolute value for the area: S=21∣1×1−4×3∣=211. The sign of the signed area is useful in itself: it tells you which way the three points turn.
The wrong step: f(x)=4x+24x, find ∑k=19f(10k): the 9 terms make 29 pairs, each adding up to 1… but 29 is not a whole number, so pairing fails.
What went wrong? Think first, then open
Diagnosisn−1=9 is odd, so the middle term k=5 is paired with itself: f(21)=21. Four pairs plus half a pair is exactly 29. Pairing has not failed.
✓
The right step: A safer way to write it is “adding in reverse order”: add S=∑f(10k) to the same S written backwards. Every position gives f(x)+f(1−x)=1, there are 9 positions, so 2S=9 and S=29.
“Add in reverse order, and odd or even no longer matters”
The wrong step: ∑k=1nk(k+2)1=21∑(k1−k+21)=21(1−n+21).
What went wrong? Think first, then open
DiagnosisThe two terms in each difference are two apart, so after cancelling, two terms survive at the front, 11,21, and two at the back, −n+11,−n+21. Keeping only one at each end loses some.
✓
The right step: 21(1+21−n+11−n+21). Write out the first three and the last three terms and see with your own eyes what cancels and what is left.
“Terms that cancel k apart leave k terms at each end”
Joining two points on the same side with a straight line
✗
The wrong step: Minimum of x2+4+(x−3)2+1: it is the sum of the distances from P(x,0) to A(0,2) and B(3,1), so the minimum is ∣AB∣=10.
What went wrong? Think first, then open
DiagnosisA and B are both above the x-axis. The segment AB never meets the x-axis, so P can never lie on it, and 10 is never reached. “A straight segment is the shortest path” only helps when the broken line can actually be pulled straight.
✓
The right step: First reflect B in the x-axis to B′(3,−1), so that ∣PB∣=∣PB′∣. Then join AB′: it crosses the x-axis, and the minimum is ∣AB′∣=32.
“Same side? Reflect first. Only points on opposite sides can be joined straight”
The wrong step: Squaring both sides of 4−x2=x+m gives x2+(x+m)2=4: a circle and a line. Whenever they meet there are two roots, so there are two roots for −22<m<22.
What went wrong? Think first, then open
Diagnosisy=4−x2 requires y≥0, so it is only the upper half of the circle. Squaring invites the lower half in as well, and the extra intersection points are not roots of the original equation.
✓
The right step: Draw the semicircle y=4−x2 and the family of lines y=x+m of slope 1, and count only the intersections with the semicircle: there are two roots for 2≤m<22.
“See exactly which part of the curve you have before counting crossings”
The wrong step: The non-negative integer solutions of x+y+z=8 correspond to “8 balls in a row, with 2 dividers placed in the 7 gaps”, so there are (27)=21 of them.
What went wrong? Think first, then open
DiagnosisDividers go only into gaps, so every part gets at least one ball: this matches the positive integer solutions. Solutions with x=0 (two dividers side by side, or one at the far left) are never counted. This “correspondence” is not a surjection.
✓
The right step: First set x′=x+1 and so on, which turns it into positive integer solutions of x′+y′+z′=11: (210)=45. Or line up the 8 balls and 2 dividers together and choose where the dividers go: (210)=45.
“Check that a correspondence has no repeats and also no gaps”
The wrong step: Count the paths from (0,0) to (n,n) that never go above y=x: reflect each “bad” path in y=x, and its end point becomes… still (n,n). Reflection gets nowhere.
What went wrong? Think first, then open
DiagnosisPaths may touchy=x, so “bad” means touching y=x+1. The reflection must be in the line the path touches at its “first foul”, which is y=x+1.
✓
The right step: Reflect the part of a bad path after it first touches y=x+1 in the line y=x+1. The end point (n,n) becomes (n−1,n+1), so there are (n+12n) bad paths.
The right recurrence with the wrong initial values
✗
The wrong step: The number of subsets of {1,…,n} with no two consecutive numbers satisfies an=an−1+an−2, and a1=1,a2=2, so a5=8.
What went wrong? Think first, then open
DiagnosisThe empty set also qualifies. a1 should be 2 (∅ and {1}), and a2=3 (∅,{1},{2}). Once an initial value is wrong, every later term is wrong too: the correct value is a5=13.
✓
The right step: Always count the initial values by hand from the definition, and say clearly whether the empty set counts. A recurrence that uses the two previous terms needs two initial values.
“Initial values are the foundation of a recursion; count them by hand”
The wrong step: Derangements: n goes to some position k (n−1 choices); then k has to go to position n, and the remaining n−2 numbers are deranged, so Dn=(n−1)Dn−2.
What went wrong? Think first, then open
Diagnosiskdoes not have to go to position n. “n and k swap” is only one case. In the other case k does not go to position n; treat position n as k's “forbidden” place and you get a derangement of n−1 numbers, contributing (n−1)Dn−1.
✓
The right step: When splitting by the last element, check that the cases do not overlap and miss nothing, so together they make up everything. Dn=(n−1)(Dn−1+Dn−2).
The color counts are equal, so it must be possible
✗
The wrong step: Color a 6×6 board black and white by (r+c)mod2: 18 cells of each. Each 1×4 strip covers two black and two white cells, and 9 strips cover exactly 18 black and 18 white, so it can certainly be tiled.
What went wrong? Think first, then open
DiagnosisA coloring can only prove “impossible”. If the color counts do not match, it certainly cannot be done; if they match, it only means that this one coloring raises no objection. Switch to the four-color diagonals (r+c)mod4: on a 6×6 board the four colors have 9,10,9,8 cells, so it cannot be tiled.
✓
The right step: To say “possible”, show the tiling. To say “impossible”, find a coloring that really gets stuck.
“Coloring only ever says no; “possible” needs a construction”
The wrong step: Analyze 1×3 strips with a checkerboard coloring: there is one more black cell than white left, which shows no contradiction, so it can probably be tiled.
What went wrong? Think first, then open
DiagnosisA 1×3 strip may cover two black and one white, or one black and two white. The numbers of black and white cells it covers are not fixed, so this coloring puts no constraint on it.
✓
The right step: The first rule of coloring: every tile (every move) must cover the same number of cells of each color. For 1×k strips, use the k-color diagonals (r+c)modk.
“Look at the shape of the tile first, then decide how to color”
The wrong step: (2x−1)6=a0+a1x+⋯+a6x6. Setting x=1 gives a0+a1+⋯+a6=1, so the sum of the even-numbered terms, a0+a2+a4+a6, is about half of that.
What went wrong? Think first, then open
Diagnosisx=1 only gives the sum of all the coefficients; it cannot separate the odd and even terms. To “cancel” the odd terms you need a value that changes the sign of odd powers and leaves even powers alone: x=−1.
✓
The right step: Set x=1: ∑ak=1. Set x=−1: ∑(−1)kak=36=729. Add and divide by 2: a0+a2+a4+a6=365.
“To separate some terms, choose a value that changes their sign”
Plugging in values gives only a necessary condition
✗
The wrong step: f(x)f(y)−f(xy)=x+y: setting y=0 gives f(x)f(0)−f(0)=x, so f(x)=f(0)x+1, and the answer is this family of functions.
What went wrong? Think first, then open
DiagnosisPlugging in values only shows “if there is a solution, it must look like this”. Here f(0) has not been found yet, and you must substitute back into the original equation before you can say it really is a solution.
✓
The right step: First set x=y=0: f(0)∈{0,1}. If f(0)=0 we get 0=x, a contradiction, so f(0)=1 and f(x)=x+1. Substitute back: (x+1)(y+1)−(xy+1)=x+y. It holds.
“Plugging in finds candidates; substituting back makes them answers”
Each object is counted twice, but you forget to divide
✗
The wrong step: 9 people each shook hands with exactly 4 others, so there were 9×4=36 handshakes in all.
What went wrong? Think first, then open
DiagnosisCounting person by person, each handshake is counted once by each of its two people. Here the “incidence table” is people × handshakes, and each column (one handshake) has two marks.
✓
The right step: 9×4=2×(number of handshakes), so there were 18 handshakes. And by the way: if 9 people each shook hands 3 times, 27 would be odd, which is impossible. That is the handshake lemma.
Taking the bound from double counting as the answer
✗
The wrong step: A family of 3-element subsets of {1,…,8}, any two sharing at most one element. Count “(pair of elements, set)”: 3∣F∣≤(28)=28, so ∣F∣ is at most 9.
What went wrong? Think first, then open
Diagnosis∣F∣≤9 is only an upper bound. For n=8, count once more by elements: the sets containing a given element meet pairwise only in that element, so there are at most ⌊7/2⌋=3 of them. Then 3∣F∣≤8×3=24, so ∣F∣≤8, which is sharper.
✓
The right step: Once you have an upper bound, either give a construction that reaches it, or count in another way to improve the bound.
“A bound becomes an answer only with a construction”
The wrong step: To prove that n positive numbers with a fixed sum have the largest product when they are all equal: each time, replace the largest and the smallest by their average. The product grows, and if you keep going they all become equal.
What went wrong? Think first, then open
DiagnosisWith real numbers, “replace by the average” may never finish. For example 1,2,3 becomes 2,2,2 in one step, but 1,1,4 becomes 2.5,1,2.5, then again… and every step leaves new unequal numbers. Whether they are equal after infinitely many steps needs a limit argument; you cannot just claim it.
✓
The right step: Instead, make one number exactly equal to the average A at each step: replace the largest M and the smallest m by A and M+m−A. Each step adds at least one more number equal to A, so it ends after at most n−1 steps.
“An adjustment must be sure to end after finitely many steps”
The wrong step: Write 20 as a sum of 3 positive integers. The sum of squares is smallest when all three equal 320, so the minimum is 3⋅(320)2=3400.
What went wrong? Think first, then open
DiagnosisThe variables must be integers, so 320 is not allowed. The end point of the adjustment is “any two numbers differ by at most 1”, not “all equal”.
✓
The right step: Adjust to 7,7,6: any two differ by at most 1, and moving them closer only swaps them. The minimum is 49+49+36=134.
“The constraints decide the end point; for integers it is “differ by at most 1””
The wrong step: Choose n numbers from {1,…,2n}, none dividing another: take all the odd numbers 1,3,5,…,2n−1, since odd numbers do not divide each other.
What went wrong? Think first, then open
Diagnosis1 divides everything, 3 divides 9… Odd numbers can certainly divide each other. A constructed example must be checked against every condition; “looks right” does not count.
✓
The right step: Take n+1,n+2,…,2n: the ratio of any two of them is greater than 1 and less than 2, so neither can divide the other.
The invariant does not object, so it must be possible
✗
The wrong step: The board shows 1,2,3,4; each time, erase a,b and write ∣a−b∣. The parity of the sum does not change (the sum is 10, even), so the last number can be any even number, such as 4 or 6.
What went wrong? Think first, then open
DiagnosisAn invariant gives only a necessary condition. There are other restrictions: the largest number never increases, so the last number is at most 4, and 6 is out of the question. Whether 4 can be reached has to be shown by actually doing it (1,3→2, 2,2→0, 0,4→4).
✓
The right step: Use invariants (and monovariants) to rule out impossible end points, then give a sequence of moves for each remaining candidate.
“The invariant says “no”; only a construction says “yes””
The wrong step: Erase a,b and write ∣a−b∣: the total does not change, so starting from 1,2,…,10, the number left at the end is 55.
What went wrong? Think first, then open
Diagnosisa+b becomes ∣a−b∣, so the total drops by 2min(a,b): the total is a decreasing monovariant, not an invariant. What stays the same is its parity, because the drop 2min(a,b) is even.
✓
The right step: Check first: by how much does the quantity change in one move? Only if the change is always 0 is it an invariant; if the change is always even, the parity is invariant.
The wrong step: Can each of 7 people shake hands with exactly 3 of the others? The number of handshakes is 7×3=21, a positive integer, so yes.
What went wrong? Think first, then open
DiagnosisEach handshake involves two people and is counted once by each of them: 7×3 is the sum of the degrees, which must equal twice the number of handshakes. 21 is odd, so it cannot be twice an integer.
✓
The right step: Draw it as a graph: the sum of the degrees =2× the number of edges, which is always even. 7×3=21 is odd, so it cannot be done.
“The sum of the degrees is twice the number of edges”
The wrong step: Among any 5 people there must be 3 who all know each other or all do not know each other, by the same argument as for 6 people.
What went wrong? Think first, then open
DiagnosisThe proof for 6 people uses “one person has 5 edges, so 3 of them have the same color”. With 5 people, each has only 4 edges, which can be 2 red and 2 blue. And there really is a counterexample: color the sides of a pentagon red and the edges of the pentagram blue; there is no monochromatic triangle.
✓
The right step: 6 is the smallest: K5 has a coloring that avoids it, K6 does not. See it for yourself on the K5 on this page.