Why do basic feasible solutions correspond to corner points in linear programming?

Why Tutoring - California Graduate Tutor
What Is the Two‑Phase Simplex Method? Why It’s Needed & How It Works
Submit Homework

If you’re working through Linear Programming & Optimization tutoring, you’ll keep hearing the same rule: the simplex method searches corner points. That’s not a slogan — it’s built into the algebra of basic feasible solutions (BFS).

In practice, this shows up in Business Analytics tutoring and operations research courses when you’re asked to identify a starting solution, interpret a tableau, or explain why simplex moves “vertex to vertex.” If you want a tutor to check your setup or simplex steps, use the main tutoring services page and send the problem.

What Is a Basic Feasible Solution in Linear Programming?

A basic feasible solution (BFS) is a feasible point obtained by selecting a set of basic variables (as many as the number of constraints), setting all other variables to zero, and solving \(Ax=b\). If the resulting solution satisfies \(x \ge 0\), it is feasible — and it corresponds to a corner point (vertex) of the feasible region.

“Basic” refers to choosing a basis (a square, invertible set of columns in the constraint matrix). “Feasible” means it satisfies all constraints and nonnegativity.

Basic Feasible Solution at Corner Point Diagram

A basic feasible solution occurs where independent constraints intersect — at a corner.

Why Are Basic Feasible Solutions Corner Points?

The feasible set of an LP is an intersection of half-spaces, so it forms a convex polyhedron. A corner point (vertex) is where enough constraints bind simultaneously to “pin down” a unique intersection.

In standard form,

\[ Ax=b,\quad x\ge 0 \]

pick a set of columns \(B\) (a basis) and partition variables into \(x_B\) (basic) and \(x_N\) (nonbasic). Setting \(x_N=0\) forces the solution to lie on the intersection of constraint boundaries. Solving gives:

\[ x_B = B^{-1}b,\quad x_N=0 \]

If \(x_B \ge 0\), then the solution is feasible. Geometrically, you have moved onto the boundary by forcing some variables to zero and binding enough constraints to land at a vertex. That is exactly why simplex can move from BFS to BFS: each BFS is a corner point.

How to Find Basic Feasible Solutions

Method A: From Standard Form \(Ax=b,\, x\ge 0\)

  1. Put the LP in standard form. Use slack variables for “\(\le\)” constraints (and other conversions if needed).
  2. Choose a basis \(B\). Pick as many columns as constraints so \(B\) is invertible.
  3. Set nonbasic variables to zero. \(x_N=0\).
  4. Solve \(x_B = B^{-1}b\).
  5. Check feasibility. If \(x_B \ge 0\), you found a BFS. If any basic variable is negative, it’s a basic solution but not feasible.

Method B: In 2D (Quick Corner-Point Check)

  • Find intersections of binding constraint lines (including axes).
  • Keep only points that satisfy all inequalities.
  • Those feasible intersections are the corner points — i.e., BFS candidates.

Numerical Example: BFS and Corner Points Match

Maximize:

\[ z = 3x + 2y \]

Subject to:

\[ x + y \le 4 \]

\[ x \le 3 \]

\[ x \ge 0,\quad y \ge 0 \]

Step 1: Convert to Standard Form

Add slack variables \(s_1, s_2 \ge 0\):

\[ x + y + s_1 = 4 \]

\[ x + s_2 = 3 \]

Step 2: Pick a Basis and Compute a Basic Solution

Use \((s_1, s_2)\) as basic variables and set nonbasic variables \(x=y=0\).

\[ s_1 = 4,\quad s_2 = 3 \]

This is feasible (\(s_1,s_2\ge 0\)), so it is a BFS. In \((x,y)\)-space, it corresponds to the corner point \((0,0)\).

Step 3: Another BFS by Changing Which Variables Are Nonbasic

Set \(y=0\) and \(s_2=0\). Then from \(x + s_2 = 3\) we get \(x=3\). Plug into the first constraint:

\[ 3 + 0 + s_1 = 4 \Rightarrow s_1 = 1 \]

All variables are nonnegative, so this is a BFS. In \((x,y)\)-space, it is the corner point \((3,0)\).

Step 4: Corner Point Where Two Constraints Bind

If \(x=3\) and \(x+y=4\), then \(y=1\). This is another corner point \((3,1)\) and corresponds to a BFS (with \(s_2=0\) and \(s_1=0\)).

Common Mistakes

  • Mixing up “feasible” and “basic feasible.” Many feasible points exist on edges, but only vertices correspond to BFS.
  • Forgetting nonnegativity. A basic solution with any negative component is not feasible.
  • Assuming every intersection is feasible. Intersections must satisfy all inequalities (including \(x\ge 0, y\ge 0\)).
  • Confusing degeneracy with “not a BFS.” A degenerate BFS is still a BFS; it just has a basic variable equal to zero.

Why This Matters

The simplex method works because it only needs to evaluate a finite set of candidates: the basic feasible solutions. It moves from one BFS to another by pivoting the basis, which is the algebraic version of moving along edges to a new corner point.

This perspective also matters in applied optimization models used in Inventory & Supply Chain and Decision Analysis, where you must justify why checking vertices (or simplex pivots) is sufficient.

Related Topics

This idea connects directly to:

Speak Directly to a Tutor — Send Your Message Below

No call centers. No delays. Your message goes straight to the tutor.

Get help with linear programming, simplex method, corner points, BFS, shadow prices, sensitivity analysis, and optimization modeling.