The problem with estimates

How long do you think it will take to build each of these Lego kits?
Lego set A should be easy to do; let’s assume it will take around 20 minutes. Set B is perhaps more challenging; maybe an hour. Perhaps two?
Now let’s add some real-world complexity:
- You need to provide the estimate but somebody else is going to do the building
- You will only get the time requested to do the build
- Going beyond the time in the slightest will result in criticism and a sense of failure
- Some of the instructions are missing (you don’t know this yet)
- Every 5 minutes you need to provide a detailed written report on your progress of completion
- Parts for the last half will be provided at half way through (or maybe they won’t – whatever, it can’t affect your completion time)
- Approval to start will be five minutes late but your completion time will be fixed
- Failure to complete on time will result in a personal financial penalty
- You need to allow time to have your construction quality checked against the instructions
- Your competing against other people to see who can give the fastest estimate
- Oh and whatever time you estimate will be dropped by 10% to meet the quarterly Lego building target
Considering all of these conditions how would this effect your estimates?
Whilst I understand the need to estimate and to set plans, it is difficult to do, inaccurate, and complicated by commercial interests. So with all this in mind, perhaps its time to consider new approaches to commercial management of technical build projects.
The main point for me, is not to expect perfection of estimates. In fact, create plans for estimates to be wrong. Create plans that enable the builder and manager to collaborate in open discussions without risk of blame or fault finding.
Ultimately when we’re trying to build amazing things, we should work with positivity and energy to do a good job. If our motivation is only to get it done in the window of time that was – despite the pseudo-science – a guess, we can wave goodbye to any idea of fun and expect quality to suffer too.
For a final mental exercise, why not try to estimate the time required to build this Lego monstrosity: