That's exactly what I meant when I said that splitting the remaining tasks was a BAD idea (unless also the same was done also for completed tasks - obvioulsy a waste of time).
I try to develop the example that another user posted above.
- Let's say we have 100 tasks. 84 are completed, so the progress is 84%
- 16 task are remaining, not homogeneous in terms of effort.
- 8 out of 16 tasks are huge, and have to be split each one in 4 subtasks -> 32 tasks resulting, representing the same work to be done
- 5 out of 16 tasks are big but not huge, and are split in 2 subtasks -> 10 more tasks resulting, representing the same work to be done
- 3 out of 16 are small tasks and don't need to be split
- Totally we have 32 + 10 + 3 = 45 tasks to be done
- Previously we had 84/100 done -> 84%. Now we have 84/(84+45) -> 84/129 -> 65.11%
The same amount of work remaining, but the percentage drops.