Conventional introduction to computer science presents individual algorithmic paradigms in the context of specific, prototypical problems. To complement this algorithm-centric instruction, this study additionally advocates problem-centric instruction. I present an original problem drawn from students' life that is simply stated but provides rich discussions of different approaches. It lends itself to a wide range of didactic means from individual or group study to whole class discussions under various levels of guidance by the instructor. I suggest diverse algorithms for solving it, covering some of the most important algorithmic paradigms. Some of these algorithms (greedy, divide-and-conquer) do not produce optimal solutions but may nevertheless have their merits in practice. The best algorithms are illustrative instances of some of the most sophisticated paradigms introduced in undergraduate curricula (dynamic programming, graphs). (Contains 2 tables and 1 figure.)