3 ms·
> I'm not very good at predicting how long things will take Very few people are. Unless you've done another project that is very similar, any estimation is go
by JohnFen 2y ago
> I'm not very good at predicting how long things will take
Very few people are.
Unless you've done another project that is very similar, any estimation is going to come with a very large error bar, because you don't really have enough information to make an accurate one. There will be unforeseen difficulties, both technical and not, that weren't part of your estimation.
What I tend to do is to sandbag my estimates, to allow room for the unforeseen. So if I really think it's going to take two months, I'll estimate four and it will usually take about 3.5.
Which brings me to my tip: the reason I double my "honest" estimate is because I've paid attention to projects gone by, and observed that things tend to take about twice as long as I think they will. I'm guessing other people would have different correction factors. Pay attention to history to discover yours.
Also, I make it overt that my estimates are "man hours" and not calendar time. In other words, it's my estimate of how long it will take if I'm literally doing nothing else. In reality, I am doing other things as well but trying to estimate around that fact makes my estimates useless.