How To Graph Inequalities

Graphing Inequalities When You Actually Have to Do It by Hand

Most people learn graphing inequalities in sophomore year algebra and then immediately forget how it works. That's fine until you have a test, or you're in an engineering class where they don't let you use a graphing utility, or you're reviewing your own work and something doesn't add up. The mechanics are simple enough. The places where things go wrong are much more specific than any textbook admits. I went back to this recently because a grad student was stuck on a constrained optimization problem and couldn't tell me which side of a line the feasible region lived on. We sat down and redid the basics. Here's what actually matters.

How To Graph Inequalities Step by Step

Start with the equation you get by replacing the inequality symbol with an equals sign. For y > 2x + 1, draw the line y = 2x + 1. That's your boundary. Nothing more, nothing less. The line itself tells you the edge of the region you're looking for, and that's it for the first part. Next, decide whether the boundary is included or excluded. If the inequality uses ≥ or ≤, draw the line as solid. If it uses > or <, draw it as dashed. This is where people lose points on exams. It takes two seconds to remember but everyone messes it up at least once because they're rushing. Now pick a test point. The origin (0, 0) is usually the easiest choice unless the boundary line passes through the origin, in which case pick any other convenient point. Plug it into the original inequality. If it's true, shade the side containing that point. If it's false, shade the opposite side.

For y > 2x + 1, test (0, 0): 0 > 2(0) + 1 gives 0 > 1, which is false. So shade away from the origin, above the line. Done. The same process applies to vertical and horizontal inequalities. x ≥ 3 is a solid vertical line at x equals 3 with everything to the right shaded. y < -2 is a dashed horizontal line at y equals negative 2 with everything below shaded. These are trivial but they show up on quizzes constantly and students still hesitate.

Systems of Inequalities Are Where It Gets Real

A system means you graph every inequality on the same coordinate plane and the solution is the overlapping region. The intersection of all shaded areas. If the regions don't overlap anywhere, you have an empty solution set. That's a perfectly valid answer, and it comes up more often than people expect. I ran into this exact situation last semester when I was checking a linear programming example for a student. The constraints were x + y ≤ 4, x - y ≥ 3, and y ≥ 2. I shaded all three, looked at the overlap, and there was nothing. The third constraint y ≥ 2 combined with the second constraint forced x ≥ 5, but the first constraint x + y ≤ 4 couldn't be satisfied if x was 5 or more and y was 2 or more. No point in the plane satisfied all three simultaneously. The student wanted to keep shading harder, as if the region would appear if she just colored more densely. It doesn't work that way. The answer was the empty set, period. This is the single most important thing to understand about graphing inequalities: the graph shows you feasibility, and sometimes feasibility doesn't exist. You don't need to find a workaround. You just need to recognize it.

Non-Linear Inequalities and Curve Boundaries

When the boundary is curved instead of linear, the test point method still works identically. The only difference is you have to actually know what the curve looks like before you start shading. A circle inequality like x² + y² < 9 has a dashed circle of radius 3 as its boundary, and since (0, 0) satisfies the inequality, you shade the interior. An inequality like y > x² has a parabola as the boundary, dashed, and you shade above it. The hard part with non-linear inequalities isn't the shading. It's correctly sketching the boundary curve the first time. If your parabola is too narrow or your circle center is off by even half a grid square, the shading direction might still be right but the whole thing looks wrong and you won't trust your answer.

Common Mistakes That Cost Points

Flipping the inequality sign when multiplying or dividing by a negative number is the classic error. -2x > 6 becomes x < -3, not x > -3. The sign flip is required. Everyone knows this rule until they're tired and applying it automatically without thinking. Another one: confusing the slope-intercept form with the inequality direction. The inequality y < -3x + 2 doesn't mean "shade below the line with slope negative three." It literally means shade below the line defined by y = -3x + 2. The slope and y-intercept determine the boundary. The inequality symbol determines which side. These are independent decisions and students combine them incorrectly all the time. For systems of inequalities, the mistake is usually failing to check the overlap. People shade each region individually and then pick whichever looks biggest, or they shade the union instead of the intersection. The solution to a system is always the intersection. Every shaded region must be satisfied simultaneously.

When Graphing Is the Wrong Tool

Hand-drawn graphs break down pretty quickly past two variables. You can't graph a system in three dimensions on paper without it becoming a mess of semi-transparent overlays and questionable perspective. If you're working with three or more decision variables, the graphical method stops being useful around the third variable and you need algebraic or computational approaches instead. Even in two variables, dense systems with five or six constraints become visually noisy. The overlapping shaded regions blend into a dark blob and you can no longer read the vertices of the feasible region accurately. For linear programming problems, this is why the simplex method exists. You don't need to see the region to find the optimum. You just need to evaluate the objective function at each vertex, and the vertices are the intersection points of the boundary lines. There's also the case of strict inequalities with irrational boundaries where the feasible region is open but bounded. Graphically it looks fine, but if you're solving an optimization problem over that region, the maximum or minimum might not actually exist because the boundary is excluded. The function can approach a value arbitrarily closely without ever reaching it. This comes up in actual homework problems and it's the kind of thing that sneaks past people who only graph to find area or vertices without thinking about whether those points are actually included in the solution set.

Quick Reference for Boundary Lines

Solid line for ≥ and ≤. Dashed line for > and <. Test point method for shading direction works for every linear inequality regardless of how it's written. If the inequality is already solved for y, the shading direction is immediate: greater than means above, less than means below. If it's not solved for y, solve it first or use the test point method to avoid mistakes with the inequality flip. For vertical lines given as x > a or x < a, shade right or left respectively. Horizontal lines y > b or y < b shade up or down. These don't require test points and you can do them from memory after about three repetitions. That's the practical version of how to graph inequalities. The theory is straightforward. The traps are in the details, and the details are what separate people who get full credit from people who lose points on things they already knew.