Solving 2×2 systems using Cramer's rule: Visual walkthrough

Step 1 — Two rules about two secret numbers

WHAT. We are hunting two hidden numbers. Call the first one $x$ and the second one $y$ . Two "rules" tie them together:

$how much x counts a_{1} x + how much y counts b_{1} y = the total we’re told c_{1}$ $a_{2} x + b_{2} y = c_{2}$

Here $a_{1}, b_{1}, c_{1}, a_{2}, b_{2}, c_{2}$ are just known numbers the problem gives us. The letters $x, y$ are the unknowns. The symbol $=$ means "the left side and right side are the same amount."

WHY. Each rule alone allows many pairs $(x, y)$ . Only together do they pin down one answer — so we must use both.

PICTURE. Each rule draws a straight line on the flat plane whose across-axis is $x$ and up-axis is $y$ . A point $(x, y)$ sits on a line exactly when it obeys that line's rule. The pair we want lies on both lines — their crossing point.

Figure — Solving 2×2 systems using Cramer's rule

Step 2 — Package the numbers into a grid (a matrix)

WHAT. Stack the coefficients into a $2 \times 2$ grid and the unknowns/constants into columns:

$A = coefficient matrix [a_{1} a_{2} b_{1} b_{2}] the unknowns [x y] = the constants [c_{1} c_{2}]$

A matrix is just a rectangle of numbers. This one is $2 \times 2$ : two rows, two columns.

WHY. Writing $A x = c$ lets us talk about the whole system as one object — and, crucially, it makes the columns visible. The left column of $A$ is " $x$ 's home"; the right column is " $y$ 's home." Remember that — the whole trick will be swapping a column.

PICTURE. Colour the first column (the $x$ -column) cyan, the second (the $y$ -column) amber.

See Matrix form of linear equations Ax=b for how this multiplication rebuilds the two original rules.

Step 3 — Kill $y$ so $x$ stands alone

WHAT. To isolate $x$ , we make the $y$ -terms cancel. Multiply the first rule by $b_{2}$ and the second by $b_{1}$ :

$a_{1} b_{2} x + now equal b_{1} b_{2} y = c_{1} b_{2}$ $a_{2} b_{1} x + now equal b_{1} b_{2} y = c_{2} b_{1}$

WHY these multipliers? We want both $y$ -coefficients to become the same number, $b_{1} b_{2}$ . Multiplying rule 1 by $b_{2}$ turns its $b_{1}$ into $b_{1} b_{2}$ ; multiplying rule 2 by $b_{1}$ turns its $b_{2}$ into $b_{1} b_{2}$ . Same number on both — ready to subtract.

PICTURE. Scaling a rule by a constant does not move its line (both sides grew equally); it only relabels it. So the crossing point is untouched — we are just choosing convenient labels.

Step 4 — Subtract, and meet the determinant

WHAT. Subtract the second scaled rule from the first. The identical $b_{1} b_{2} y$ terms vanish:

$call this D (a_{1} b_{2} - a_{2} b_{1}) x = call this D_{x} (c_{1} b_{2} - c_{2} b_{1})$

WHY. With $y$ gone, $x$ is finally alone — divide and we have it. But look at the two clumps: each is a main-diagonal product minus anti-diagonal product. That pattern is exactly a $2 \times 2$ determinant.

So: $D = a_{1} a_{2} b_{1} b_{2}, D_{x} = c_{1} c_{2} b_{1} b_{2}$

PICTURE. Watch $D_{x}$ : it is the matrix $A$ with its first column (the $x$ -column) swapped for the constant column $c$ . The $b$ 's stayed put; the $a$ 's were kicked out for the $c$ 's.

Step 5 — Do the same dance for $y$

WHAT. Repeat, but now cancel $x$ instead. Multiply rule 1 by $a_{2}$ , rule 2 by $a_{1}$ , subtract:

$same D (a_{1} b_{2} - a_{2} b_{1}) y = D_{y} (a_{1} c_{2} - a_{2} c_{1}), D_{y} = a_{1} a_{2} c_{1} c_{2}$

WHY. By symmetry, the bottom is the same $D$ (the geometry that decides "do the lines cross" does not care which unknown we chase). The top, $D_{y}$ , is $A$ with its second column (the $y$ -column) swapped for $c$ .

PICTURE. Mirror of Step 4: this time the amber $y$ -column is evicted and the constants move in; the cyan $x$ -column stays.

Step 6 — Ride it through one real example

WHAT. Solve $2 x + 3 y = 8, 4 x - y = 6.$

$D = 24 3 - 1 = (2) (- 1) - (3) (4) = - 14$ $D_{x} = 86 3 - 1 = - 8 - 18 = - 26, D_{y} = 2486 = 12 - 32 = - 20$ $x = \frac{- 26}{- 14} = \frac{13}{7}, y = \frac{- 20}{- 14} = \frac{10}{7}$

WHY compute $D$ first? If $D$ were $0$ we'd stop — no unique answer to find.

PICTURE. Both lines drawn; the amber dot at $(\frac{13}{7}, \frac{10}{7})$ is the unique crossing.

Step 7 — The degenerate case: when $D = 0$

WHAT. $D = 0$ means $a_{1} b_{2} = a_{2} b_{1}$ , i.e. the two coefficient rows are proportional — the two lines are parallel. Then Step 4 gave us $0 \cdot x = D_{x}$ , which cannot be divided. Two sub-cases:

$D = 0$ and $D_{x}$ or $D_{y} \neq = 0$ → parallel distinct lines → no solution. Example: $x + 2 y = 3, 2 x + 4 y = 10$ gives $D = 0, D_{x} = 12 - 20 = - 8 \neq = 0$ .
$D = D_{x} = D_{y} = 0$ → the two lines are identical → infinitely many solutions. Example: $x + 2 y = 3, 2 x + 4 y = 6$ gives $D = D_{x} = D_{y} = 0$ ; every point on $x + 2 y = 3$ works.

WHY. $D$ is the signed area of the parallelogram spanned by the coefficient rows. Parallel rows fold that parallelogram flat — area $0$ — so "how sharply do the lines cross" has no answer.

PICTURE. Left: parallel-distinct (a visible gap, no meeting). Right: coincident (one line drawn twice, meeting everywhere).

See Consistency of linear systems for the full classification.

The one-picture summary

Everything at once: two lines meet at $(x, y)$ ; the coefficient grid on the side; the three swap-grids $D$ , $D_{x}$ , $D_{y}$ ; and the two little fractions that pop out the answers. Follow the arrows.

Recall Feynman retelling (say it to a 12-year-old)

Two friends each give you a straight-line rule about two secret numbers. Drawn on paper, the two lines usually cross at one dot — that dot's across-value is $x$ , its up-value is $y$ . To find $x$ without drawing, play a criss-cross game: take the coefficient grid, cover $x$ 's column, slide the answer numbers in, and do "down-diagonal times, minus up-diagonal times." That's the top number, $D_{x}$ . Do the same criss-cross on the untouched grid — that's the bottom number, $D$ . Divide: $x = D_{x} / D$ . Repeat with the other column covered to get $y$ . If the bottom number $D$ comes out $0$ , the two lines are secretly parallel — they either never meet (no answer) or are the same line (endless answers). Either way there's no single dot, so the division refuses to happen.

Recall checks

Bottom of both fractions is always which determinant?

D

, the coefficient determinant.

To get

D_{x}

, which column do you replace with the constants?

Column 1 — the

x

-column.

Rule for a

2 \times 2

determinant?

Main-diagonal product minus anti-diagonal product,

p s - q r

What does

D = 0

mean geometrically?

The coefficient rows are parallel, so their parallelogram has zero area — lines don't cross uniquely.

D = 0

with

D_{x} \neq = 0

No solution — parallel distinct lines.

D = D_{x} = D_{y} = 0

Infinitely many solutions — one line drawn twice.

Connections

Solving 2×2 systems using Cramer's rule — the parent this page unpacks visually.
Determinant of a 2×2 matrix — the down-minus-up engine used in every step.
Matrix form of linear equations Ax=b — Step 2's packaging.
Inverse of a 2×2 matrix — hides the same $1/ D$ factor.
Consistency of linear systems — Step 7's classification.
Cramer's rule for 3×3 systems — same column-swap idea, one dimension up.