Universities don't seem interested in going out of their way to teach any project management in their computer science curriculum. It seems like they expect it to evolve out of the fact that they give you a final project at the end of the semester, which is often poorly executed since you try to balance it with other class projects and finals. Disclaimer: I'm only a CS minor taking a few courses above & beyond, but my roommate is a CS major and confirms this as well as our classmates.
The points below are from what I've gleaned from observing effective leaders of teams I've worked with at game jams as well as longer term projects with a group of peers:
To succeed, you should try to follow a pretty tight routine that everyone is on board with and enthusiastic about.
Plan ahead, but don't commit
Have a brief period of prototyping as necessary if your gameplay mechanic ideas do not fit into a well established genre. Focus on having some playable demos up within a week and weed out what isn't fun.
Having prototyped, design a rough end vision (rough, because this will change), and set goals for a first milestone. Do some work thinking about what core systems will be required in your game and focus on getting them up and running. But don't overdesign these, use the simplest type of sprite or 3d model manager you can get away with. Set a due date for a week or so.
Stay motivated and focused
Have a regular cycle for milestones, and meet regularly. If you're roommates or live nearby, meet at the beginning of every day you know you'll be working. Or, if your schedules aren't aligned, schedule two times a week when you can have a scrum-style meeting.
Everyone should report on what they've done, what they will be doing, and what problems are preventing them from completing their work (this is not to say you should absolutely follow the scrum methodology, but this specific practice is very helpful).
The smallest tasks take the most time to complete
Remember to leave roughly 1/3 to 1/4 of your development time for polish. A game with less complexity but has really tight controls and presentation for its core mechanics is more impressive than a masterpiece of design that feels clunky.
Team members are like RTS AIs with their difficulty levels set all over the place
As a leader, you should be encouraging the team to move forward on goals. Some team members will intuitively know where to focus their efforts after they're done with each task. Others will need more guidance on where to go next. Furthermore, some may spend too much time perfecting a feature, in which case you need to tell them to move on. Conversely, somebody might write something in a way that is not farsighted enough to accomodate the full scope of the project and you'll need to help them toward a better solution.
Note although I say "you" a lot in the above paragraph, remember to delegate these leadership and mentorship tasks to others who are able to fulfill the role. It will free up your time to accomplish your tasks and at the same time help team cohesion.