Many problems reduce to a system of linear equations, for example
2x+3yx−y=8=−1.
The inverse matrix method solves such a system by writing it as a single matrix equation and then undoing the coefficient matrix with its inverse — the matrix analogue of dividing.
Writing the system as AX=B
Collect the coefficients, the unknowns and the constants:
A=(213−1),X=(xy),B=(8−1),
so the whole system becomes AX=B.
The idea
For numbers, ax=b gives x=a−1b provided a=0. The same works for matrices: if A is invertible, multiply AX=B on the left by A−1:
A−1(AX)=A−1B⇒IX=A−1B⇒X=A−1B.
X=A−1B
Multiplying on the left matters — matrix products do not commute, so BA−1 would be wrong.
When it works
The inverse A−1 exists only when detA=0, so:
detA=0: the system is consistent with the unique solution X=A−1B.
detA=0: no inverse; the system is either inconsistent (no solution) or has infinitely many — handle it by another method.
Worked steps
For the system above, detA=(2)(−1)−(3)(1)=−5=0, and
Method: Finding A−1 First, Then Reusing It to Solve AX=B
Some problems ask for the inverse of a matrix AND the solution of a related system in the same question. Since the system's coefficient matrix is exactly A, you only need to compute A−1 once and reuse it — never invert twice.
Steps
Step 1: Compute det(A)
Expand along the row or column with the most zeros to minimise arithmetic. Confirm det(A)=0 before continuing — otherwise no inverse exists.
Step 2: Build the cofactor matrix, then transpose it to get adj(A)
Work through all nine cofactors Cij=(−1)i+jMij systematically (row by row), then transpose the resulting matrix.
Mistake 1: Re-deriving A (or re-inverting) from the system instead of reusing the already-computed inverse
Why it's wrong: when a question gives A (or asks you to find A−1) and then a "related" system, the coefficient matrix of that system IS A — recomputing it from scratch wastes time and risks a fresh arithmetic error. Correct approach: confirm the system's coefficients match A's rows, then plug your already-computed A−1 straight into X=A−1B.
Mistake 2: Losing track of a sign while transposing the cofactor matrix …