Estimating time frame

The next thing I learned from the Time Management training is estimating time frame. When you need to estimate how long a project would take, you need to have 3 figures available:

Likely Time: Based on your past experience, how long do you thing it would take?
Shortest Time: If everything goes smooth, how long it would take?
Longest Time: If things go bad (touch wood), how long it would take?

I am bad at estimating time frame. Actually when my boss ask me for an estimation, I always give the shortest time, imaging me coding each line without debugging and revising. Maybe subconsciously I want to give a short time frame to impress boss. But the reality is: I can't finish it on time. It is not that I don't work hard, but there's always something going wrong. Of course.

When you have the 3 figures available, you need to do a calculation:
Shortest Possible Time = (4*Likely Time + Shortest Time + Longest Time)/6

The Shortest Possible Time is the figure that I should report, not the Shortest Time.

The other things in the Time Management training, besides Priority Matrix and Estimating Time Frame, are kind of boring:
Writing To-Do List: Action Oriented, Increment, Measurable, Scheduled.
Know which tasks you need to wait for other's input. Schedule them first, then put other tasks in your free time.
Figure out cons and pros of a proposal, appoint weights for them, and make decision base on the sum of them.
You can do things base on: First come, first serve; The one with early deadline first. The easiest first; The urgent one first.