Why Estimates Go Wrong

An estimate is a guess about work you have not done yet. It is therefore wrong, and the useful question is how wrong and in which direction.

The reliable finding is that people underestimate, consistently, because they estimate the work they can imagine and not the work they cannot: the integration, the edge case, the thing that turns out to be someone else's bug.

Relative sizing is more reliable than hours. People are poor at 'this will take four hours' and much better at 'this is about twice that one'.

That is what story points are: a size relative to other work, not disguised hours. The moment somebody says a point is four hours, they have converted a comparison back into a prediction and lost the only advantage.

✦ The failure that poisons estimating

An estimate treated as a commitment. It is a guess offered to help decide what to attempt, and if it becomes a deadline the rational response is to inflate every future estimate, at which point they stop carrying information. If you are estimating for yourself, notice whether you are doing this to yourself.