[technicalbard]

This is true of all knowledge-work. Traditional engineering is fraught with the same problem - you don't know EXACTLY how you will solve the challenges, so how can one accurately estimate them? Even worse - if you don't have a clear set of requirements, or they are expected to change as you progress (a critical feature of AGILE), then estimating with any likelihood of achieving same becomes all but impossible.

The more uncertainty in the path, the less accuracy in the estimate. Kahnemann's latest book "Noise" provides some good background on why this happens.

Having multiple people do independent estimates and averaging them probably gives better results, and having a clear process to document assumptions and test sensitivity to those estimates can also help.