Graph systems of linear inequalities and use the vertices of the feasible region to solve a linear programming (optimization) problem.
A linear inequality like y <= 2x+3 does not describe a line -- it describes an entire half of the plane, with the line itself as the boundary. Graph the boundary line (solid if the inequality includes equals, dashed if it does not), then test one point NOT on the line to decide which half to shade. A system of inequalities is the OVERLAP of every shaded region -- the feasible region -- since a point must satisfy every constraint at once.
Linear programming asks a very practical question: given constraints (limited hours, limited materials, limited budget), what mix of choices maximizes profit or minimizes cost? The key fact that makes this solvable by hand: the maximum or minimum of a linear objective function always occurs at a VERTEX (corner point) of the feasible region -- never in the middle. So the whole method is: find the region, find its corners, plug each corner into the objective function, and pick the best.
Mastery looks like: They correctly identify the feasible region of a system of inequalities, list its vertices, and evaluate a linear objective function to find the maximum or minimum.
Common stumbles: Shading the wrong side of a boundary line; checking only some of the vertices instead of every one; forgetting that a corner off the axes needs its own system to be solved.