Backtracking is a general method for solving problems that require finding sets of solutions or optimal solutions under certain constraints. Constraints can be explicit, which are defined rules that limit values, or implicit, which describe how values relate to each other. Examples of problems solvable through backtracking include the n-queen problem, sum of subsets, graph coloring, and Hamiltonian cycles.