A team is doing Scrum with 3 week sprints. Suppose at the end of a sprint they've got nothing to done. What should they do? There's a strong temptation to ask for more time. To make this sprint a 4 week sprint. Most of the work in progress is 90% done, they say. Another week and things will have got to done, they say. It seems reasonable.
Trying to run systems beyond their capacity is not a good idea. In this situation Scrum's fixed-duration time-box constraint has served it's purpose admirably. The problem is not the choice of 3 weeks. Changing 3 weeks into 4 weeks is not addressing the problem. The problem is the team planned to pull in an amount of work and get it to done in 3 weeks. But they're not yet in control of their process - they don't know what their capacity is. They pulled in more than 3 weeks worth of work. Probably a lot more. But we just don't know!
In The Toyota Way, Jeffrey Liker writes:
Taiichi Ohno considered the fundamental waste to be overproduction, since it causes most of the other wastes.
Advice from a genius with a lifetime's experience. Toyota manufactures cars. It makes cars. Its production line is an actual line. If manufacturers are prone to overproduction imagine how much more prone software developers are! The things we make are not even physical things. In software, things are mostly invisible. It's difficult to manage what you can't see. In Quality Software Management volume 2, First-Order Measurement, Jerry Weinberg writes:
Without visibility, control is not possible.
If you can't see, you can't steer.
Rather than asking for another week, the team should really be thinking about addressing their real problem. Their real problem is that they're pulling in too much work. They have to somehow learn to pull in less work. So they can start to be in control of their process rather than their process being in control of them.