We’ve all seen the victims of problem set night. Groggy, bleary-eyed kids, still wearing the sweat pants they slept in, downing full glasses of orange juice in the dining hall before stumbling off to class. If they make it there on time, it’s usually to toss some stapled, hand-written sheets, pockmarked with cross-outs and erasures, onto the problem set pile before slumping asleep in the back row.

Late-night problem sets, which often drag on until the early morning, take their toll on math and science students across the country, not just at Harvard. Interminable problem sets frustrate students, deprive them of sleep, and take their toll on a student’s self-esteem. That frustration may be driving students to other concentrations: Past studies have shown that up to two-thirds of students who enter Harvard hoping to concentrate in science end up switching majors.

The typical advice given to students with problem set woes is to start earlier. In the 1980s, when the terminals in the Science Center would run much faster with fewer users, students deliberately waited until after 10 p.m. to start working. Could it be that the pattern of working on a problem set only late at night is a vestigial remnant, carried over from the early days of computers?

A better explanation for why students start late is that problem sets in most advanced science and math courses are so tough that they require collaboration, and the only time when busy students can find time to work together in large groups is late at night. In addition, problem sets often require applying material that is only taught in the class before it is due, which makes it a waste of time to start early.

Short of making problem sets shorter or easier, there are many ways professors can make problem set nights less onerous. One pressing issue is that students underestimate how long problem sets will take, making it difficult for the organized ones to convince their classmates to start working with them before it’s too late.

Many professors provide a limited number of times that students can turn in a late problem set with no penalty. Although helpful in theory, taking a late day only drags out the process, and it frequently cuts into time that students had set aside for their other coursework.

It would be more helpful if students could be given an accurate sense of how many hours they should set aside for a given problem set beforehand. Many professors already poll students about how long they spend on the homework. Asking them to additionally answer how long they spent on each problem of the set would give them good estimates for future assignments.

Another quandary is that, intentionally or not, many classes will assign homework problems on subjects not covered by lecture. Harvard professors are undeniably brilliant and often very good teachers, but it can be difficult for them to know which topics need to be covered in order for their students to finish the problem set. Part of the problem comes from the fact that, at Harvard, there is bound to be at least one student per class who is so good at the material that he or she doesn’t need to be taught how to solve a particular problem. The professor may never learn which of their problems are too tough for most people to solve, as the other students quickly learn to seek out that one genius who can help them get through it.

One simple remedy might be to list the readings relevant to a particular problem. Knowing where to look to find the relevant problem solving technique would at least save students the time spent searching through unrelated chapters with a fine-toothed comb. Better yet, it would allow them to do more work on their own without seeking expert advice. It would also force the professor to only assign problems that students could reasonably be expected to complete given the materials available.

Physics 16, an honors introductory class for freshmen, offers an effective solution to many problem set woes in that the professor will sit with the students in Leverett dining hall the night before the problems are due, answering questions until the wee hours of the morning. With the professor on hand to clarify any misconceptions or confusion in the wording of prompts, students don’t waste time arguing over semantics or pouring over their textbooks looking for examples. However, the professor, Howard M. Georgi ’68, is the master of Leverett House. Most other professors—who live off campus—can’t realistically be expected to stay up until 2 a.m. in a dining hall with their students.

Interminable problem sets can be a science concentrator’s greatest frustration at Harvard. However, with a little more communication about length and topic, problem sets can be made much less painful and much better learning experiences overall.

Adam R. Gold ’11, a Crimson editorial writer, is a physics concentrator in Adams House. His column appears on alternate Fridays.